From 6c94f102dd91a538e5ea4a9e0ac12e0c29f14a3a Mon Sep 17 00:00:00 2001
From: Martin Petr <git@bodkan.net>
Date: Wed, 21 Feb 2024 14:30:41 +0100
Subject: [PATCH] Update website

---
 NEWS.md                                       |   8 +++---
 docs/404.html                                 |   2 +-
 docs/LICENSE-text.html                        |   2 +-
 docs/LICENSE.html                             |   2 +-
 docs/articles/index.html                      |   2 +-
 docs/articles/vignette-00-installation.html   |   2 +-
 docs/articles/vignette-01-tutorial.html       |  10 +++----
 .../figure-html/plot_model-1.png              | Bin 26066 -> 25258 bytes
 docs/articles/vignette-02-grid-model.html     |  12 ++++-----
 .../figure-html/model_america-1.png           | Bin 76087 -> 75951 bytes
 docs/articles/vignette-03-interactions.html   |   2 +-
 .../vignette-04-nonspatial-models.html        |  25 +++++++++++++++---
 .../non-spatial-graph_sampling-1.png          | Bin 0 -> 49838 bytes
 .../figure-html/non-spatial_graph-1.png       | Bin 23013 -> 23044 bytes
 docs/articles/vignette-05-tree-sequences.html |  14 +++++-----
 .../figure-html/introgression_graph-1.png     | Bin 27281 -> 27819 bytes
 docs/articles/vignette-06-locations.html      |   2 +-
 .../figure-html/demographic_model-1.png       | Bin 19722 -> 19842 bytes
 docs/articles/vignette-07-backends.html       |   2 +-
 .../figure-html/unnamed-chunk-5-1.png         | Bin 16826 -> 16916 bytes
 .../figure-html/unnamed-chunk-5-2.png         | Bin 20604 -> 20790 bytes
 .../articles/vignette-08-nonslendr-tskit.html |   6 ++---
 docs/articles/vignette-09-paper.html          |  14 +++++-----
 .../figure-html/figure_ex1-1.png              | Bin 60017 -> 60194 bytes
 .../figure-html/figure_ex3-1.png              | Bin 73627 -> 73670 bytes
 docs/articles/vignette-10-tracts.html         |   2 +-
 .../figure-html/introgression_model-1.png     | Bin 21366 -> 21844 bytes
 docs/authors.html                             |   6 ++---
 docs/index.html                               |   2 +-
 docs/news/index.html                          |  12 ++++-----
 docs/pkgdown.yml                              |   2 +-
 docs/reference/animate_model.html             |   2 +-
 docs/reference/area.html                      |   2 +-
 docs/reference/as.phylo.slendr_phylo.html     |   2 +-
 docs/reference/check_dependencies.html        |   2 +-
 docs/reference/check_env.html                 |   8 +++---
 docs/reference/clear_env.html                 |   2 +-
 docs/reference/compile_model.html             |   2 +-
 docs/reference/distance.html                  |   2 +-
 docs/reference/expand_range.html              |   2 +-
 docs/reference/explore_model.html             |   2 +-
 docs/reference/gene_flow.html                 |   2 +-
 docs/reference/get_env.html                   |   2 +-
 docs/reference/index.html                     |   2 +-
 docs/reference/init_env.html                  |   2 +-
 docs/reference/join.html                      |   4 +--
 docs/reference/move.html                      |   2 +-
 docs/reference/msprime.html                   |   2 +-
 docs/reference/overlap.html                   |   2 +-
 docs/reference/pipe.html                      |   2 +-
 docs/reference/plot_map.html                  |   2 +-
 docs/reference/plot_model.html                |   2 +-
 docs/reference/population.html                |   2 +-
 docs/reference/print.slendr_pop.html          |   2 +-
 docs/reference/print.slendr_ts.html           |   2 +-
 docs/reference/read_model-1.png               | Bin 31283 -> 31200 bytes
 docs/reference/read_model.html                |   2 +-
 docs/reference/region.html                    |   4 +--
 docs/reference/reproject.html                 |   2 +-
 docs/reference/resize.html                    |   2 +-
 docs/reference/schedule_sampling.html         |   2 +-
 docs/reference/set_dispersal.html             |   2 +-
 docs/reference/set_range.html                 |   2 +-
 docs/reference/setup_env.html                 |   2 +-
 docs/reference/shrink_range.html              |   2 +-
 docs/reference/slendr.html                    |   2 +-
 docs/reference/slim.html                      |   2 +-
 docs/reference/subtract.html                  |   2 +-
 docs/reference/summary.slendr_nodes.html      |   2 +-
 docs/reference/ts_afs.html                    |   2 +-
 docs/reference/ts_ancestors.html              |   2 +-
 docs/reference/ts_coalesced.html              |   2 +-
 docs/reference/ts_descendants.html            |   2 +-
 docs/reference/ts_divergence.html             |   2 +-
 docs/reference/ts_diversity.html              |   2 +-
 docs/reference/ts_draw.html                   |   2 +-
 docs/reference/ts_edges.html                  |   2 +-
 docs/reference/ts_eigenstrat.html             |   2 +-
 docs/reference/ts_f4ratio.html                |   2 +-
 docs/reference/ts_fst.html                    |   2 +-
 docs/reference/ts_genotypes.html              |   2 +-
 docs/reference/ts_ibd.html                    |   2 +-
 docs/reference/ts_load.html                   |   2 +-
 docs/reference/ts_metadata.html               |   2 +-
 docs/reference/ts_mutate.html                 |   2 +-
 docs/reference/ts_names.html                  |   2 +-
 docs/reference/ts_nodes.html                  |   2 +-
 docs/reference/ts_phylo.html                  |   2 +-
 docs/reference/ts_recapitate.html             |   2 +-
 docs/reference/ts_samples.html                |   2 +-
 docs/reference/ts_save.html                   |   2 +-
 docs/reference/ts_segregating.html            |   2 +-
 docs/reference/ts_simplify.html               |   2 +-
 docs/reference/ts_table.html                  |   2 +-
 docs/reference/ts_tajima.html                 |   2 +-
 docs/reference/ts_tree.html                   |   2 +-
 docs/reference/ts_vcf.html                    |   2 +-
 docs/reference/world.html                     |   4 +--
 docs/search.json                              |   2 +-
 docs/sitemap.xml                              |   3 +++
 100 files changed, 150 insertions(+), 128 deletions(-)
 create mode 100644 docs/articles/vignette-04-nonspatial-models_files/figure-html/non-spatial-graph_sampling-1.png

diff --git a/NEWS.md b/NEWS.md
index bfa8e539b..7f411cc13 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,12 +1,12 @@
 # slendr 0.9.1
 
-- A new helper function `get_env()` now returns the name of the built-in _slendr_ Python environment (without activating it). (#162ccc)
+- A new helper function `get_env()` now returns the name of the built-in _slendr_ Python environment (without activating it). ([#162ccc](https://github.com/bodkan/slendr/commit/162ccc))
 
-- `clear_env()` now has a new argument `all = (TRUE|FALSE)` which allows deleting all _slendr_ Python environments. Previously, the function always removed only the recent environment, which lead to the accumulation of potentially large number of _slendr_ environments over time. (#8707b9)
+- `clear_env()` now has a new argument `all = (TRUE|FALSE)` which allows deleting all _slendr_ Python environments. Previously, the function always removed only the recent environment, which lead to the accumulation of potentially large number of _slendr_ environments over time. ([#8707b9](https://github.com/bodkan/slendr/commit/8707b9))
 
-- `plot_model()` has a new optional argument `samples =` which will -- when set to a result of a sampling schedule created by `schedule_sampling()` -- visualize the counts of samples to be recorded at each given time-point. (#d72ac5)
+- `plot_model()` has a new optional argument `samples =` which will -- when set to a result of a sampling schedule created by `schedule_sampling()` -- visualize the counts of samples to be recorded at each given time-point. ([#d72ac5](https://github.com/bodkan/slendr/commit/d72ac5))
 
-- The _msprime_ dependency of _slendr_ has been updated to version 1.3.1. As a result, `setup_env()` will have to be re-run to update the internal _slendr_ Python environment. (#dcb83d)
+- The _msprime_ dependency of _slendr_ has been updated to version 1.3.1. As a result, `setup_env()` will have to be re-run to update the internal _slendr_ Python environment. ([#dcb83d](https://github.com/bodkan/slendr/commit/dcb83d))
 
 # slendr 0.9.0
 
diff --git a/docs/404.html b/docs/404.html
index 7454b8d65..c555382b0 100644
--- a/docs/404.html
+++ b/docs/404.html
@@ -31,7 +31,7 @@
     
     <a class="navbar-brand me-2" href="https://www.slendr.net/index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/LICENSE-text.html b/docs/LICENSE-text.html
index 7e41a0f6b..e1447c81c 100644
--- a/docs/LICENSE-text.html
+++ b/docs/LICENSE-text.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/LICENSE.html b/docs/LICENSE.html
index ac9b7a11a..a19cecc5a 100644
--- a/docs/LICENSE.html
+++ b/docs/LICENSE.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/articles/index.html b/docs/articles/index.html
index 7878c2d62..9e0cdde3a 100644
--- a/docs/articles/index.html
+++ b/docs/articles/index.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/articles/vignette-00-installation.html b/docs/articles/vignette-00-installation.html
index 6f4e7b5c1..2ffdc04f3 100644
--- a/docs/articles/vignette-00-installation.html
+++ b/docs/articles/vignette-00-installation.html
@@ -33,7 +33,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/articles/vignette-01-tutorial.html b/docs/articles/vignette-01-tutorial.html
index 561cf1721..ed5641e52 100644
--- a/docs/articles/vignette-01-tutorial.html
+++ b/docs/articles/vignette-01-tutorial.html
@@ -33,7 +33,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -775,7 +775,7 @@ <h2 id="running-the-simulation">Running the simulation<a class="anchor" aria-lab
 <pre><code><span><span class="co">#&gt; ╔═══════════════════════╗</span></span>
 <span><span class="co">#&gt; ║TreeSequence           ║</span></span>
 <span><span class="co">#&gt; ╠═══════════════╤═══════╣</span></span>
-<span><span class="co">#&gt; ║Trees          │    133║</span></span>
+<span><span class="co">#&gt; ║Trees          │    134║</span></span>
 <span><span class="co">#&gt; ╟───────────────┼───────╢</span></span>
 <span><span class="co">#&gt; ║Sequence Length│ 100000║</span></span>
 <span><span class="co">#&gt; ╟───────────────┼───────╢</span></span>
@@ -788,15 +788,15 @@ <h2 id="running-the-simulation">Running the simulation<a class="anchor" aria-lab
 <span><span class="co">#&gt; ╔═══════════╤═════╤═════════╤════════════╗</span></span>
 <span><span class="co">#&gt; ║Table      │Rows │Size     │Has Metadata║</span></span>
 <span><span class="co">#&gt; ╠═══════════╪═════╪═════════╪════════════╣</span></span>
-<span><span class="co">#&gt; ║Edges      │18682│583.8 KiB│          No║</span></span>
+<span><span class="co">#&gt; ║Edges      │18713│584.8 KiB│          No║</span></span>
 <span><span class="co">#&gt; ╟───────────┼─────┼─────────┼────────────╢</span></span>
-<span><span class="co">#&gt; ║Individuals│12838│  1.2 MiB│         Yes║</span></span>
+<span><span class="co">#&gt; ║Individuals│12836│  1.2 MiB│         Yes║</span></span>
 <span><span class="co">#&gt; ╟───────────┼─────┼─────────┼────────────╢</span></span>
 <span><span class="co">#&gt; ║Migrations │    0│  8 Bytes│          No║</span></span>
 <span><span class="co">#&gt; ╟───────────┼─────┼─────────┼────────────╢</span></span>
 <span><span class="co">#&gt; ║Mutations  │    0│  1.2 KiB│          No║</span></span>
 <span><span class="co">#&gt; ╟───────────┼─────┼─────────┼────────────╢</span></span>
-<span><span class="co">#&gt; ║Nodes      │18332│681.0 KiB│         Yes║</span></span>
+<span><span class="co">#&gt; ║Nodes      │18351│681.7 KiB│         Yes║</span></span>
 <span><span class="co">#&gt; ╟───────────┼─────┼─────────┼────────────╢</span></span>
 <span><span class="co">#&gt; ║Populations│    6│  2.6 KiB│         Yes║</span></span>
 <span><span class="co">#&gt; ╟───────────┼─────┼─────────┼────────────╢</span></span>
diff --git a/docs/articles/vignette-01-tutorial_files/figure-html/plot_model-1.png b/docs/articles/vignette-01-tutorial_files/figure-html/plot_model-1.png
index 0c5eef9c12dc97274243ff903c874c253f5f3fb3..d571f908cbc24ef9c0f34f5b6e2bdf82782afb1f 100644
GIT binary patch
literal 25258
zcmeFZcT|(<`ZgM5)Nxdp8Aq`J<A@+2pduh$hdm-7pdcU}rAb$+5R%}qm9c?>fYhK;
zq(-SygQ7xcL0W(iA`&{GBtQru$+=%=@89{Y^T&6-^T+wtch=di#X^#|Jn!>7ce}3Z
ze(zm0Gu-*Z{vS{%)J~)G|FA%zL>uAf{P$bnlke5;D#NcW*Gvrm0l(oN*Ps>$zqZ{t
zZySU{{Tz(^h>Unh-$J2&MH&6$?B!eOb2RibtB~jV0#5sJhb?0-elvZM6@F;<FK26Q
ztw$7;a@f0Nb{=Ei(oE@7slCqCO_U;Q_I;jP)}7d_^3$`c&+@N^#9F&0zdZWl)$Obw
z6@F0ok&`k=)Iaz6r;F-ySH;Vm2$O^4xq3YxLU@2w=50=7|9VlXMgC>v34#MF`oxf7
zQvckv1BE(s=x8JS@yq|jANEsE45dh!==l@^hUwE6y9tGA^tc#R)iqS%qZ%+$cg(5o
z?a(E-*uah-j>E4zStwBy>f-nRyZq6WGb94TQt%_iAE{<{YOA`nJnrr7y{u;9=;*jQ
z;2IaaFj8Njc`-@^6=ml3?$6DduEPWE5mV1~xWiv}P%%}@=pHevj`EV6BSCvqNF@B;
zv;e^@F}I<ip&<PJ1rgW{T6f&rlE$Yv4H}c;lIgqjHD&uUIDS4>mX-}`wY97ibDWlb
z-F%IAv-~j%GC3LV6NM|zn%F80&)zTf+LyVe@luX83>Jr|&DkL1JUi1h`(iDxAT!t{
zzcE@;8#vcUE0A<enDO=wT%H$YHR}uB3|;!c<nNpJ6<X&!k9KyZmYMij6Pz5Gd++4@
zy#{;U!D{_5KDb+v-7l<BHwxq%#*&&>X@wnY_QKgAo!V>T=jgl-`WOZ6m8mBNB^gf*
zlXd7GMV=VUh{laG%DSw@Q}bPAnD=E??Z_4H1hQR&N?|5u?d$@lopS?@k1Tf&e3Bwm
zkstW<Gm?mUnWGdl-!c`+n0+oNRMv50Bf*$TPnm1nJh!otc9Q#fFg)~Q>s)u))o6@F
z2Krn?*z9^uGkT^B-x*7*l{2y~vQPerV%AzGkvu^l6Cd-#M)Gy_Hrc=10dsM2Z%`)g
z#P%d5a!S)(M$cBs$1*7_O0z~y0MldGB)?=L(YP<CQk^zC6EDL*pnk0PGXC#T$m%Sg
zk7s<SVkYQDaLnQ(Ib_cG!>b1B1Dp<e?a2x?LircrY1ZOO4%a=Px-JAiVIi-SW8v-N
z&8mMnlALiMqI@jR#YDal?B+lRXG2*jPUkXnue+iWRy}++$~0CZ*_cb_d1RYc`R}Xi
ziOn|(9Cku8DmB@xwe2SC{)T#u0b=##i1C$kC9@c&o>ETttV&d0cl?A`n5QGNrbFl=
z`Hy6A%qg{i!Daly>KSBNhmI;m$SUD5<wGZ0t0MYl#msRYm#m9?lrrlZ*wHe!#4CMq
z(pVV@snu9@UUJaR8}cN)gkA=Re=E?emQM+v5AyWg*qPHE%A2&0SlC*+&+|^$^%CUD
zKa95WoKyO+%z{o~_`<{TI&WDG`Gx&oQe|Qq#ygDS$9*WdB!w=?c$mC_b9;_mc^t68
zwh8Q5yJr;tNrRIg_@c5Vq?oKd!j0WvYcynFIJByvn<xeE5X&I!#^THT!h90j`LrU<
z$*xGjXaq}pl(IAzRE_%_v`;YMSh<F-TaOs+o}DvX0Uz{wXM3iuwxQHg=R+OVoeP4q
z-Nil@Hk@k}t`%=Q{8-8Rs>WX(#^Oo(wW={v5!WjYi(GD}u_dx|F1Nlt#H)}e8U;?g
zGYIVO6_V6crCJv?a4v-n<oWuMr`)!r$t4M=_AeQs!v(IF!oiw+MTNnu`I;dNcKGD$
zSZ%qkdVAK0fk}7}QG%UsY3Mo86l7^xR`1N*+%*%87HgeR$te5$#$1f1k3U!=T&I`|
zdb|Q|)$z==!(UQ`5fl0LPviWW)EZG;onUtvF;b~1VF!C}wk;Gg%M9bYOWX1KL~xB+
zFMe@BUH5EDQZgZEgpgKGr3_VtnI;{xwP5Z=xmO3xC!|!obrsGjBpOruh+8urfrq9x
zuCC%Tm`W;V;?v9NdefhCT%E@?7e2aA)7c^OyDrpPeDhbMy@z;{+PDT4pR}|>N%xl?
z`oe4Twrjy7`Zvy0y!^$wS+Mdw>X#1%YNV(6BuAWarfvE%U0!p8Nz#uz?ltvhoF_}#
z-WLDH`8q~0uA0zvu3R`N^6D;U=fTGuy=tG>hMUF>YpbJ!FU<++Rk8(lE#z%q`-W#a
z)Er9|EbzIw<~m{%*&@tCz!1r5O(L8ZENw78x*M6AcH<@))06S{s${DKVSy`0j21K!
zi3@F5A<?MEB;im3E*~YDL9H6$m28j`wg-+CHM4qIyZd4jJgtjb49H3(2{GN%#DkdM
zx2()tG0)%0MTwqC<TtAo-&u{0k@LjpOb@BqQl2{S6Bzw;8DZ!jES1SbVnDYR`vzMj
z!AV<YD>HdxUZq&|J%;{8=&Y2Jdki=Fyl!o{W|jBlOpF!L`*Wz{FM7vsR7JXc@$ZX0
zyR}?bSidETIWnV0!p{d}bm;P)W(#s~DYLaQ!qc+VV<ilMEvePY$d5U(z+>finJXD*
zI>twCyd-tKE8e^&&XH}#d|3%nD|Spx>G|Z|_vb2a?o4w9xo4rRWFM^b#hc)bqZA@;
z+RjqITg#WL6c(C_>1C~O<0V?D9CsquV3yKE&#dz!^iS&7zYOGxxS(IaWA9EhE_T+a
zIA`L=Ok$?VC@GPYh-MM$GPP>5j<X42JA5*J!p;m9x9vnx59ghM<5!rp^^i}8i?>^a
z<fGk-s*hBMdCamUd6k_k|M>=^le~NVmPHASmw12vI_ugMAwPz(>##;fp$R{&l>Bbe
zvqc2$yW}Za>8y+6++uxI?TZ!Jorer0WFe`JpT!`H%h=GKrR&wKEKlT4FxJx&a&!AX
z7DhA#n+9m|Mz9n`dfv2J)rY!q3Q1nBZp1f`GFLbI*=u<0ODpbrR+CS}m)ui^BDw5K
zQLt+oZrBA5+1T0KY#1wXJLAf#^0eFTXAs9|^L$EN7z!z-2>bhGt2SOU)tZFMxP)qU
zf~L7FT6`#1dN-<|6^!qqdwuBn5@Bdy_f1l#wye4Gh0wW-^tDmO-31v;1njzphIWK~
zdMw%j<!38T4%-n&ZgSMQw@_VIpNVXV63cpScyeCaPC9Xb^Y}SQg+N;y*h;6aqoQ7%
z5Z!@d`VN-5hX=2252$n9YEcl1LS5eqo}|3&)2C0~hDWKa-8suMeY20kzKdf11$N6U
z@+4=5H$UIX7#-|0D-GYHY@Ece*F(=&dse(eo%zFxzyAweuv%3;SK2n@dqCCW!nsHL
z6J@--y<7jeZCBuAMp&*nUQd(DNC+CZ8dFTKs&5v_`nf?E?ZXzA)>iO*h&|D?N?m5g
z6foaZlQ{lUiNX!sn5F$Jhk*)Kwo++iPjs|>lyFYxuTleEjwL%|!`GLm6SnP=b?Zn~
z(-dy7d+O1ghkH+4OM57K<_P@fx2lZ|T;PZVxz(Mbo38ArPK&ZnwaM!(E&007+q)*a
z(j@LttS-NuB}1Z$)AT7$!svt8wb9)|0qUY3+%QgGa4S|{u=Xfc$-eSf7#$NPYo3{2
zQ&V&5)G0=<rFq9rjp4^X9X?miJB_;B{rYMXD@`b{G|sWGv$IpLrWfgFrfkO5*L<ZD
z=}|8O7k?rSOkK&pZ0zS(Z0n5Dswj~7LKLhJQcmBzCRl6M&qZ_QHJ3i@3R?bbn3tO?
zYp%z2+kfCdPlXR<g(trE_^&||?M6XCLA7y|zP<`ljYmi7Q{&><YbB1g{kWe|1HY-M
zsU6rj!1k`FQ1gnzUGw&F7V<Wo@wDmNR2Rw|Dzd9kh>;Av(y7OrI|h#<;T(B0GkhZD
z;Q<Zh2YZxS=W7>6CUkYts~5WM&QZ~I&!0c<7<hH*W$BI4_bUV?F-W&w^Iy&U;z22K
z44N&jnrIXizhG)=8o`;*2<p_7)1XHNrKY7N$T-%NGpAI@jSWTh&CY&PUC%Wcj|~D_
z4#hD10?_u>#da%Nc*xhIVzX!R&Jg>}w3$z{L$*{p6S21C*Ds|lB`VtF!sF{kEMMm=
zQP%sb8Y%?bhx=4dUVZ=1;`Jg(czw04GDI3UBWQLX0b9{g>6sZW3rlYnC#8Gx#+ODN
z!P?lYXLN#_V1v6)EBLy%xAzd{Al$L@Wx?jPC{eMNxxq3v3_ts_!uSGCh%R~i^?rI_
zt9*Wbz7bvzbJBO<^>A>HseGcbZB4LONgb;d`PQg=ehh9NEM=v;xdzsX`%WDDc($Zg
znZ7Zh(L-=1rr-K{PxI?vhx-?T1l&o_jkTqHx)H&lLtc65*!Af_*gsW+r{M}22By<1
z)|hJv9_`PbM2l%1^-RBc4jF?t7{spdwM<<5pF}6CrLolmi=#XAS*!P@C8i=%bBCW!
z=$jUX?$CO+-a|HHjzkwqx)#2Dq6C+`^S@bT@}J@SXG#7)zbz#TP{MH)97&>K@s2t*
zH8nMm7D91ve|AE1wy5l+-%z=q7WlC<lSmovR2ewl8p)&5J4<Tkhdg-W$<CR2n02@J
z|J-I<=3&~YFWi7)T=(#~hXb<Tyoi{j$Cfo7U5FLaS!`6NCMnq`)h;w>=cNZ4&&<qp
zcXd6sG}l$zC1*DL_wiDaFHq!XCk}&?->2q%IgR<>!vl58j=hgHgC?@8tE*|v&a?OA
zG7%rAie8;NcIEZ2w3?7YhBOqpDCR5hxdf+rEjn+6!-|&P_!S3Z9(t%`SAG<3k52PL
z0ziC`ZCNg@`N01DLk}|`zT3XNF}hBvWmekT;&+1|oym_75sZrCo~e>`RKZUVOx;RW
z_B)(poa3h@@NPaBbiwJ?=p7kAB}`aj(!yH_JP00LZ!TA$^w+0eB0g12ce(3PEN7OY
z0-?vd%x}Ck5pllYz=tp!6-bC$A$x*%G*PK9FCuS;SzbZTUThStUTEN*3}m#(7@3%u
zK&U+zdoZO<g{ULCRjhq1PCs*EVxqUN?}*U^8ION%5Cf3}o&->mDV>6v@B5A9GM?gv
z8<Qy~-HmfEq{PGZPepIp+4@x8vXtsyn*#~TK6T$K{x#FNPmObqIo6!AaSC@g(;R;)
z<HflN6Q~te;4HcUA+{pF*pEsg!@53{*4KM!7}tpevL17NT^>YI^#fZ0BM5pWUq4b{
z?%NH~E3oVA*71j`oh`d<94;|Wqk?}Er?ca17b%4b*kP=wUizrvh||*8rVR6*r)k|A
zd0t53Zi!nT%%`DmRJi&kirtiUfdO0(_}>M6`j3zQkB^59dFTI?ec&$frI652^BWKD
z&{$b+yZLQ(bvS>lU6p*n($X@`kASUPe7~K$oTpD4sRr}TI_9YEnJ1+i;Reob)|o<~
z5a$Jk(wOD`9Pm9+KTwM{MKUWWxj8vUj82Bm4LF3+Lh_4gYmA#h0Y7MMG{#2x5Zvl=
zqsQ(q3Fdm?d^UcAA-7M%-=TIha)H&0tG`{TqJ`R{8E{xsXoan6BVnbM&R0hA-ih{<
z1aLf$=2b&xdW(NRQ1ryhp&|`SpP&qCe>PV=nYmBr){PoS!=<X>%d8iz%8m53PZ~&7
zWH-GJuE;bVq@c)neSJfza5uCYJW^DsWvGbMiEIBzQgI%UvPDj#J#~yydHdzhNLB*A
z9YAX8VR87l{6Q%}L;#NVnSDUL8iu#p0kC$8WZ3YDii!iPQ$8itxxa}ZbM!OZmAKwV
zI2pONC_@70odB70?$_tYrMuxf+**_L&|&agsPaO%Bs@kG^RxBtFII)t?n_@kCbq&%
zpqROHmRSjI{YAF>v_pNt*EPr(K~WRGLrUi~gsb*<Tczs;o9!T#$HL`2U}~l*1ZTbF
z=i=jU$6`&OVu;Vq&zIjNd!Z#GY$OGOB!iJ)-gz}fvQofj>VAu;7gu*IzQ40wHF&CP
zuAVc3c<Z#?S%%>1N1I|i=ZCA)LW)9W`^VFRI!Q~-()cw08hNDjWJ5sJL!J;)aPD|=
z9;`NUpD-eib|F(o6Do#(d*-9HDCgkMf9%u8M#`F|KHX(@`cOzn$ZWH6H6wllim}ek
z5)Q9zn<H2xId)Q2-I1+Qw_Ko(=q~dlM=rmV?hqB%c`RdF>1#XRz?=V8I}{e07#J9s
z>8|u%NI!iON>B)S;ZHF?EHg+6$Y25jBYR(6LQ)6y*(+#fg^hCUT<HqI2GBL<jjs}#
z^KZxE%l9y#_zq?ch4A%6x9n(1*9__&9v+?_BPer>Sb6$<`9kYr<9HpY5~l~tCOt4_
zupkA5xlO}4Cdn(54gU*%etwXsw2B3Oy(b3+qFQ*H{Yboln5K-;?8>`u5uWd)oZ-`7
z<i%OKd-txoYqCS2(H>=ow}KE?SJy%OE?D!Kqr6|H63n#ZZ5%`TWSiwoQ_Hbist#(<
zf-U{)mOky+cT&cgH$>?VSm@62GsO9E{TG&O4uGlX@dn8gWq7`?8I16lZK>P80{>H8
z(a6y|5*}X~Z>OmH*+HNzKc%7J3DtWz)ViuSzkV3T@|Y<a0ZxNwkx7Q+2W(K&wK=w9
zrmqkwt&Z81zv{p$+LU~&IG3kB!<4MNiNpBv+51#I7>dqZ<A0AGz2$8Hy&**x9f~9&
z2Sa)<d3kw}7Dt<cU@a%#G)uKbi|ai*8@<J^|BZk3c#=c-c%olrMR|EP6rC$8EBj90
zaFN1oEWro~(s+Fw_O6&DOlfMUG!oL;uyN#!z-J(qUbpu3L5gOO2Vw*|E4`M;P4Dj5
zS31}n-mE?H#2IkaSwf*>eMCq}!}<)eZ%l!hK|qVEO?St0)BKI|FSmB*m;{X{+DtsP
zG&73-_f22{Y-VY-VMuXpk^fXK^VV<MC6wDCx2zdlz_HWl=?G)R9pN-2BWWMh&dB}=
zoO{z;LZ94VlgVpKW2zC$Q*-4#W}5uvo-)Xq{L8*2Ffpm!v&HM>+iaXMAq@aOOit(F
z&0rj#4W%=XO4ca;Wt^T?)kuU7I8y7S*OH|FQTAf0a=JQaedpSZyVBw4k8#_B^-L}O
zjoJ8IEA1_k7x>ejd7%sM`eNT$wd!qNPqOzX*)*^c9d5kaY!uI1AnFsj9RY#x^d7(q
zP4^YrH1KE();dRFSNlFHHNP#O;P{jaY3fSRTdUNCc)@bsAVBi$X<(g<0z1?wZUe8c
zBEXMg+AnZQ<?n;4jjEoV?T|i}30BOptc1v~yYN?pHDF}nu=<J#;9XD(*>jt8u6Jb>
zIaHrCVnIZ48>tKDe%Y%378}WBjeGg}7CHP5$GwUw-M*OOnVItWrQzhHzA!vibK)7;
zrdj~x%n33ob&J32ajMDV!VSF8aw!YWc&*mjuw`wojNG|A-4pbFn_PBfr8;r#ql$&I
zGd_Do(1_SHJb=x=kIob+DJ*QOpg5%=F1D}0in|!2Ux%FoKh=TQepLYd{@w4uB#W&=
zMX<J<d0Nl#aQepw5(K(B_{QFT|D{8SW2gCEvd>WSu|D^3?-Q`<<ylGqfAwATR>Z`1
zZ$}(@qD82P)7u+Yk<APqRogQ6f|u%;HYCzJr1f-yCOeC-pNGQ+?A%rU4Eg+rggb&>
zfPatg_xaN~lyoE+sKrCoU;f;F(yKeWHRn66Vpy$7*!Z0Y26n>r(`D_>2ykd7Foe$&
zo(InuE-m+%PtFt)%)lkVUG12sjqHTxZ|PW`dso(3AX{G>L69dU$W5}^ROpN~fSLmP
zZ9V<0j%Mw}Mz9&_-0HHjT$|Es{Xx~3K7661ypcshw*Sqb8eLjCNn=B1A%ap{Ai3>V
zqf!W~>DW%q>XlAs_vr+48KVeHk?SzFw6a=0+|62y79iYMUZ3qQv>|vKZW=f<xcRL~
zvNE7MX$J|##io{)J=S+NX%W48a>tidG7{Z|mFRAQPkfb!ZG}O7|8Sya=<-sEpqA(s
zr@Q(pbbr*wSoP$@#7DB{xEe)HCF6Pbb8KjmaWzkFSJ3dQ&s<usv$gil#x2NR$WAb8
zwX$)nAT1ds$RhENT5$2qWY~yCL#n&y0?UHS%v$wrDcyyh{JP*|!IVLTuMT>uj>l@3
zvgU&Nlj&l4xw1ZnI2~09-2|`~;*b|g9SM$iB%Gez7TxMGRPLq8EboCJkv`p%HwiY}
z@!{UCARe>I6E-#T$?mCtNQK|BW__$LlpNbxad0hnW9feEO>gy!zF;U9XaWN*e=f$>
z)YMvB_h?U^IT=9Af{ZO$k4m@Spf_NU0usRo4JAV^$Gfd9jvB?+uMR1S>#b#d6pzxd
zHny<0Wd0IR)GKG}gF^kS1@Q6fFGfB^*;e^G8n3K?@!O`()_Oo*259^1+i>Oh<psLF
zWu93F_`WiicYhW+)~Y{x^a!!MU_ID{ul@qdlha$ecS=poy*t}z`fpE2ld6t-!E0}K
zn{Ry-hY2Sz7Bk{6agN{ujIp8TyZKfo`ybZohx_}ndVVq@C!6hd=!qX*zrZS-jL)OU
zJfg+B^5mLV%xiqNs(wbL&EA<zxFqd^5mX4%4hjU7W71rg(uj~i0VRPWz0|sjBg$}B
zGDxQSChPN8Mox%S?gGoGxc2e>@U5{&whfWtXokr3zVz(;@+HGI6UEQOjR?yH%wTCQ
zV;4uxro()KU>sUxzNg4cK;KpT2imn6HQ+?U*mU>yuD`w6eRoVH4l_xQS0mJXc*)IS
z+6P{UV*X=r^V*OPq2uu(gTQy2fe6gW(LE_r_ciZuymS3a<``c)IqcyBPdSwm6TLc<
zz}e@dolWsnG1HMM%h_XWGTsY6hoai$KOH`OV!BT|oW3>Q(6gd*#J{~YkrY=#m^&e&
z^%k6J14J!VfH)(CJUV%W;*38Wr*qP=*8kRl1fztF&>z9?;Y>OEO(;h=DyB94vXgU9
zdh3I0f|64i+XGDRa?IB21TRiL=-0!)5yu26MmJ8NdE0x~{h_Dl9})>I9H|tVn4yRL
zE$?^li+ugD{ER&UrSQ;8Cs%tJS<wl#LOCuiAYshV|LubOSo}7MoZQ0;q4PE=uA72y
zfI|SZBnQ$%1Vo7Mw|rq!bmh{L3(6>d*8=tRcfzk-Hurtw(-(<;obVO-4)glp2>Rj(
zq0Vz@$Zf80-n`l%nUOo1_V9I)*B{vBWtFO<dcOB0ZIrJkZ>qb|F|=|0a~bo7Tjf28
zq{_}u4Vn2W{G#>7Cq`rTc6jskrB7`$a-WXq^SoFYICH@YHBnRc%vKx7<C&FI$cr!j
zs*hfsKNc|ZVUrD^&%$_Q%KHI@n#au%9t%aAKEQ?;h3-7wwXeKfMuYxz?zKTFuQ5AU
zHk@*mq}yZXc{5D1D^ytGn80pGjU1aL6~!n9wI`oSG_JlevVrjSR%-^81hV)olg^Po
zN7F1->I>?x^}Yb2Kq6%{Gdo+>rd*p16ij04=HP24X@bn8ZjX{fi?G=jG}bPyunpp+
zLI`z&wdH=>5t(Uv@yaaU#C}27A;E33q?qzE!Ouu<otU=!rHv!Gm*U?ScVWPn=A%(p
z{_5wN{(iT-2wYr(n`u6n(~O%>iq7iw*_lqGz-5!dQ^dMgeS}<Dybh*`b5PLWMy`^T
zz`U&ZV*T}*7yl-2xF#grzi&oq{b*mQo8^Njk&I7!_=>j{>N!yQoJdjiG)rSHU^%s_
z*bO8$>h0dP^|dVpMX8%mKUYDBpb>mZ?c?0u6}Y{p$DgRJbP`f88)MfPii*a>+tj-l
zxrJ$T*Tm%A;>AT>;j_(cgo3r3adT8TXDo-Rh#h8rj@zS~_9lZFF=ZbRKF;-dFwDTq
z=3y1`!#l+LQd1d!Zcpg>bi`cW3fW;Uu)|<P6ORqVNY75V20~u$2~{0HNhTAW8EHO*
zdUb$r0;JG$CywlW_3)9SXylq%F=x|=gTziVJ>A3H&-8d{l?z)MF8`Ilh|)M>UQ}?!
zA*$d?t$KBkw}pM$1^z;|!0G`Hvz|aoA0#;oQ|GjAH33)9&-m+bf>2;O9=omUY$9K=
zRwZ(Dd&=>vzss7XKNH7H8EAfq5<LkqJ^jY}ZCoxHm#IjoJBehF)WeDw|DklCTKq;(
z3S~|`)I?S)J$-Q^+dSEAbcW%mElZJldXPfI1d95aaOlZT?c!badMjValD0|v@FS?j
zX;@*k^S}++FrT<>T_TB)ij#FYjAY{QviGaJj8Hp0b_&_o>S)vk<*~-dq~~Gv2F&}{
zJF8Y^;?s3stt@%-YmZ1DWjDM>Np?`_DZ^@hW3J8E?3JZbXMsl(W~5h<i>Lc>zd1-U
zipcZ}XoJF3LvsW|(x-EUlOh2w)pyRJTD!Z`->%3sTFcj>Z~``-JPEneDL=QwI!JE=
zFthfWV-Hmoeiobh&|hAzyld#QTC%!>+xD#e^Z3k-)e`T#0Uf1|Vv3k(@NswiCO)LN
z2*1K5nPcN2J@XXuIp$D(WdZ|(_~#VuF#k^Ng}TT|ZMqk`K?S??VE6W&!JBx?7c!kP
z>T=K(33UzJLZ|A)pU<da7;A15(S2rb!ZRaJ4ho~Z^H4~Ru!a;B-abBM-hBlKT5#YY
zfC`hPZOT-a+D$vkmqrDoK6V0uNn5S_Lz0x5#JEd5lE-v{7)~k?jbASzEzFyZ@EW}H
z2FqiljivAj2ZL`^B%pk=<~G&6uaG-)B=UN)+v<JB_Of{KRg?8zImw)@<aKm+m0tDO
zAmgu$gy;w>?VlP=ICK06*0_o+hn99w2ebiaJ1=_KCF`u8pKaw6Yqhx(AB~brg=yQm
z<~Q6n6}*%sRikCZmCOaK-&t3-u8$8fAH?IXU+APO?(4nCJcS|*<ltCMKJm2z?eH7t
zO8p%xc2(F_7Oku^$b1|rks_lT_A2zZs7Dr2w`a7zm=?#yGKO^vGEgo(HKBg>8Ve_+
z3qpU6V#bbhAy6OvZu4HqPJc%fX8-laK*6Rum@`B?XYn^`;d@CW^Z#jf3StDsn7x{T
zZg<6{GpmCptxw6o4SkDC7epifdw!Z-QsU1aIFjrTf5_l27`t*ywo!sY^3QN^r3KUj
z33BG0VCm|bdfvUShHo}Y+V4h0g9^oe7tMh3no7T}R0mk%|NJ7d+emJ_ps-|!u?80l
zgtH(kg_B;D11i!0xRWZdj7V96*bi`?s+jd<9YD4bCfA?HUu{p*An29?Skkyhu@=Q)
zQZ)cJ=zM!~EU@>I;SnR`iDU-B*RZ}si+eyaynE+1xVnAe+@^FHK9580&9^KIp6*7V
zdml=PDHP_w11n}ASdYSsxlJ4Q%p-q;0jHeuvOo&Nk$QBMp)xYQSEUl~?A~1a11Zyw
zIXBelvf$G*wFuY&h{L{<Zs(N&Jt~KPhO)@<|ATZOl;%Q&oYhL^S2*zU?}M9f0pf#j
z`6rqWCS<QXo!YHveOYI3hB>%SEN}=DNZE6HtJE_CaozHvCl;2L2}qrRKyc@9PXG)F
zmIB`k7<;j9U>8XjYEtf(H;2nA+mypu(-LJho|Y4d1sMwRiK(UDt&cP(gPdFa^o3rl
zbes^9b`Xt`_RzLWySF8)klAE?{QCMjR5>b8uP3(tA!nhG$a=EtT+YH@WQb1q_%Lhj
zu5-sy`Mk)fw{hd2V-^tSP=W9+%&fUTUdJO^$Z$?AJ;SL!0;xs@Lnw|2&;E@ImryAQ
zK35q8U{OF)2{TETjy0}$>1!dZrzbR^F~36PlphM^W>yA(ammMQ<1wp)a^P!%kVzj7
zBI{NICmcBFo;(h-(UNsGy3~ynzuM9oIs>Sj(E?2@KlGtSuTSHl@Wtnb1(Gl`*X0nG
z))9bJ{f1$skRx+DRyp+x7){P_P|$Qv@<bYa!=nK{u{tJ0twI=%UjM$qRq8+d95H}Q
z6#nZXRq+T8W)9ekY%^JNGJOu9sc&R~q%kZXfI-vU?mYp8g@r*soO0)W{3C|J-K?sv
zwK~#+d#Wq8FGHq!x{g-rJ^13GVBT`Np{aA0!(H;^<$878e^@<N9nAc!UA4i!ZMz;!
zwM?lEo=y4mPP#=~PryI59B#tHfG5amQsC|X5*NuhoT3#nBijs2i~b}$$b$8q7gxYi
zy1VQAU&dMFk0<i&>^Rg#{_K#iN`>#Ox(Y?-_Jkf>lD!~&>RcQ+lVGfU@Xkd<nhUyC
zm6zupwc69y`+?aR&(F(jRl#xF(!)l=xy%3@Rmygn@yTJ;AuWYgyt1^QWK@-@Eo+q8
zU(})`)H&6LPfN=o>m)VQN?m-j0JH5Y4M-=#7V$ZBd>nT9CASL4t5SE}xpOCCza!kv
z*jsrvS~aPuN;Qg<+&CM1PAj`cIlpirQ)c>ep-`62CbHun;4XdM>-%B3M#leu7^+dd
z@(&ZXn55PzM5BVJ3)aTsJYmKM{i<jup>zh=prU)9{9i_Rx_k7`8~22EDMLyQXCuB=
zYv<(Vs^=t4JT+DV2CPHdTyVCAm`|6bmj%1C*2?UcnMa$-v}Fe049SCwP3`dB<GkLT
z%vul%7W^_Py}fRD%qj^0kDd~dL3RIilanDHyo2af>G+Ven_<r_OoEID|D^s1wQdMR
z+K-#AG=^RkhB>D2=_^g4rvpv35AH%$N~c*xHOEQg%`r2U`;NPu5sc{z4kPfig*|UQ
zuP}7BzxX#1`B;sbEpJUQaWY2iAuEFKu|!zRkod|x9XSGgak;fDNcfDZ5*~&P>ljF9
z;oB6Xv8tznO-F`n73eih;cv$1d-iLg+;@XsxlC|C3zEY+$QJB{HP4@q?xb4|eJDEU
z+j_uJw64u*p_BEIeUd}orOGgHHjCHct?9}q+F8)~ruhq%2Gp^7hrfKhy93B!GDiKy
zPMJwbxk`G)5k7P52}YTrA3K10$dPyX<oGCra#YrJ^}@{#5z{${7(*c75GO$x-BDaM
zk_ti|4ea5kS6wNng_|u?b)04CBSx$G>)XRO%y$JYbU2d?i0g5%?G@WHGZJJ#XZ4k$
z8=5w@J`OVTU%G)QPzGu9>n1T1(8`TTlL1?#mOMY23t8iEcXzpfYgQ}XwvGNQWvf5`
z+@8*LD}i9+!8DPBKFNrlbgvTCXWeb4U(LJC<;BaD%?)Z!_pDWF3`+x7>CmbtzHF$!
z{&{^80N>X8yA?~>or$$!iXVeEg!Q;G4zqSvBwT2#kD+1CVS~md`!m-9%%AQ8qIAlw
za_UoInMX(KUq2m2kW?(+sXC|q%e!bKO5MNnGU!Sz>tj*m)vFb*;14%LdGmk8h069X
zw>&wl3M4^$hIZA9LxT-UkyD-Rf~JbT*rmI6sT9B0@u&WcZvP?S_;hKk*>i5-jZwG9
zi6Rr_Y8ivOJ4k?^KStt*@)2r9v>~>zqoPi-@2lK=hcddpaOebN4WSkQWB@phFeIk2
zar@O2!XVOurwAGh|MJ%<f_+A_aCywza4`Fivb^zpLu|tBt+*mYkN@q|8_CL!P8=yB
zP_bU#GizlSUjsEYWKKH3>ud}EI+9ij8_CSik>Z_aoZp+S$#K9-Zaw(yqxo=}csd^f
z?D{9pTi`e<&E3f;d~ml>^-Mt_<Ou+99fXnC@4nxe_dLusR#zA|2!SQH#H8w}aYrAn
z{1m9Ol+g!IDl`k(qSLSveF9e`LS2G%L&E9S)h%KNpCMopV6P^i*8<gBCWOIL2sy&6
zqGyJ)Cn^lBD}&V3D!d}8KOsMTv85xA#j*KAKf%NK0X{hl$RDB}9{cpD?UJEnB5c0M
zTk&p$`qkEdq1RUb&WHYI;1nf=Min1#Z)1>dl{JZJX9GZ<*b2BJYPSv0T1JSLVaLJK
z$B`(uN(-X{j4U+`X*HsK+W;3^7<>+n9q@(#j)1lc;be5+)bp4L1kY_xQAMOwj86vw
zF2o~5R|-JpAES}vryt=F2-uZE#IJsOxUa{9ha?C{z7C)c7h(fGEuf+L8UkUu_&{Z{
zfzUVL>ml!&`usAeSuzX?6Sy<vUE3G^Yne%qbNB|<Raa}mNZA0|!E6~rhJ_F5x8<^u
ziRl*k<?ldY*NUWrKyQ`(HVS8FS*XwMN7zIyB_$=JcwkB<?kkqGIEK*@5r`N`#Jan?
z<C~>%ItW?RB(9%cSW;8qXaK?DOtpF;3Xe!2s{Co|6i7Fa(gK0a16~JUFc?IsfQY*i
zO}=GtUS&Tl&N{Ne3{lkRe=}%L{_na{e62*BD!i@ss5m*~AA?i7{3~Cf$f;_IpcXj~
ziOc%RY!#OGUlt6*iZX^WO2y&raf2AK4giZY<SlcvD_0W@9v;*@*%TvA9<B~j-2W{H
zwoVSD-}wJ?!hG$)4~uR8N|S9AkfkY41A8FC8mfk~7$9WFB`E%Go}m>|_$O?{Ge;8Z
ze^~Cban3+!5DqQO8S>r)$PN9M;ailX5g56Ur6P<+U<A+-UXRm3mQ1-dQ^W~y8jz+E
zS6g6XB0TA)$PYgmB1|ds$!7RusAEsYCVnQTPL3`VaDo=!ZOyZ<@;^We0A9>q@*6dg
z1KB;`=AOW|0<YCRJ71OG_Z;q<wI7m}Z-IX)4yS|XwtKR2HR|O-48(kbJd#2IoY-X`
z?hb%(n$8Lk$|$WDPz>w`F$>(k=fBa%z?%bYs)?Yp0AV7|^wO$)7dSja*$dAQmp#7N
zBu<VJ5w)<dy&Bgrk?N!H^N+|ap2PFdF&y^~fBj?>4_!gka?rSfqTT`9lol+;thONY
z!z`HdL5e{abh)i+BYwR%4-n6%2mIe&<+C*cn|5mWh9DBiS5G?>8UaopSnuY|Ou`4R
z>j#Q9ZOqJK9YLAgD5Q^MG?T2lxFbUAfazel>GzNSdFK50KHqIlTzmN2_f21qj{T+h
z?vU}@FNtS2|Dr}{JU{VXRcA^k==S#|b8dfT{d%&|L0Rkv|0&5EXBvOnw&{NNiF>q+
z>;$J-!B5?}Wnt$nGSWG`l^Gi=d{E7djr#!>Ni#&OvGaNhUIcmKN|HJ<ro(=hP@I*O
zosk*fV5XT&$#P#NL7zWu&TTQ&E3KbI2=YUNFATmh)0vr>w-;t-XEmz%)REOn8+A}h
zp?qEKe`%yR<1p(yUV;oVO+hdR*|{D~+HrC)-LS_ZsHWe;k<`XLn-?Ls)w&_<w-4y>
z%FC#vZ9yCy^J<ISyBa{Hs+vK=ihAwn_gU%-Gry`zWBZ<CEl3h;oD=UFt0{|Iy{D_|
z_-icaUJ;p;g>rJxOn%-(%i*|#%WP_pE}UL_w_xicNRXZYWy4sXC73%qJ0k*D2P}6S
zp)cbRGE6?X4e{cTAAQ52-@G{i)Byd^EUfE*+79lIQ-1k{g^J?3%NO-FM#b|=N|eD<
zwSZ`h^-`LD6rrdM4GkNF`~^>7|MS05ES5k<CvM)pM}`#Y40UW{$;$$fIw*_);m>*V
z=BCRaWGnG9ITrGusEPzk?DUP(!SNSTPsx*Y*cYJKtK>z%=tCu~bia!L{Hk4}Vd>G?
zHUXb5)D1vWsVA>}_^Z6K(u2hyMfYoHXvo>tb8evt0p8yG%jaXY=UZ6gZ3!}nz5$VW
z?c}O~6TK*DCET&NpGKnr3v6Es++RABa-wasGh5;yh-|?S>nQ{##I^PZDh;O>`2%lU
zHEIyCZByIp@4rY3*P5$gs|0pxPBKCWXj&ud=Gdp4ii(h~%8ZN*V3+o31~NjOi6w(&
zo}NJM8O2X#qLtxiUNM3OSD8C)CH}?UzxoKGGDrB~l@~#vLj$edRd`+td;2t*a3IZ*
zK%0A6!H=^ts|DKPE%W-&*e4o;1gGO0Rx2`BfXeDXVi^qbRy5k7Z9LEkP{YYes=rB&
zg%KK!Zdk1#xWhtdyM5aLu?6!Y0+2PPKrls=7fu1RKR0hjBy|WaTRmShd(!pO!-taE
z>L3pB5Yw8<y3MTQe4G89Ry)5`)r~756|Q=<DTJm@t=Xb<Fb*r*Z;Sf8Am$7VtTFbt
zzX2hUZ~d6xyTIfAa2DyZyZHU4Z-CK{jquSgf1<#GvVQyy@gsMB{m&@?q5WsZpn>;4
z8wTv<>cDq5JUu-T;Z$)|9k_dq;Tr~eYwyG-76kJPtR`u}jhGUk_FX6!ZI~j^V24PE
zk#+<~OJ|GmAf{;@4@6Y3B_$<@Wdpm~4zh%>mVcQtU9=9U*Y<QxRdG-tN6czzX|ebV
zIDLc{{#!+K5@bSvLAt~geY?%26_g%`T!JW$VdHoq7f-S0ATa#b5BI#TU*CI09qJ|>
zcDw5I>@0QuTc`-R%>Qvp&D*;_V5z^q9|Tz8U8MkC$r#z#P(q#Qy}hTd(~=vObIL0!
zju=7DU_3a`ozrR%^e*Z)$wq19miRCJm1yi9Q{qeo)kHGVvGRqNl$7+yA99|?#gj#%
zw}#)|^z!qQEH5f6Q-$bp*f4f@ewGsQakou&UY?v03lSl}Qwb|?rGS%}Lp5nd<$><^
z&UT4h0zn08N27Qs*$yEMn8wgi3hp}JGEZ#SxjN7t!e1rF9?oi1w^&JV@iVS==$r*c
zW;m4Umw#EoD<FVO0pAKU81%0hI#I+y$$hJQk#&Iltp#F6&_s&oL{_xe(A9s%&otCX
z0FkeLi%$co2I4%wX~NcmsOY}~^i|CFP+g;Z75~?}%HJI;V5zfyz%EBK)_vkZ-MCNx
zkR~^p<V;kRPb|>g*BE*j-1Pm4677WWWH}6`kSJtsCnXBgHw0lMoUnGLbBn<2@dp&@
zZ!N%UNzL!}qq5G|SRb(=^MW=?K2y^oUC*+*r3+Lg8pT}1kE0qp3oMKvW3WDEo<!WJ
zSmp0jvfpUVu0=P+QMw-Mthkjne~{#9p+bKI*LYD0gE5-CA95|^kJ8A&FE<>~gEd#F
zbRu(Ndcpjhx?gRx=w1&9Cyo6x^H|Mu$;G=$eC7jO2Ji322~rXjV+83H6{J^}NK>Cl
zAMO<{XvEoJw7r8*7KHQNJA9OUd@q>kYI4WE$|SyvbF@jWo2Z~NibeLu!WE-Ifp!$J
zK#0vqNVZ0&%aD+g6l529&?+_b^0i|AvRp<+baXKHeu&sUC&eoir%I37k<91%3x9p;
zJE4Vr-*&quw8;OfAeC_UC&I>s4hhHf$ccnP&Kb^fkr>_hND7-%*tFk8;z*Rp`S|7K
zWkwJFv_5Az+<Z@^UYd+%AS+F^;CW4qK|iH8hgcaeCtNOU=gmm4H5S&>3Obt>KhtO#
ztE;Q!RaJS##Y$kr2kMB=-EqD?CR$q7hV+Y1y{#9n`bi}MFx-Z=HYDsSJYcuV;!jNU
zy|MKpg!={d{MqLjjh&#77#M5kO<t_0-sW3gJMN6uzb&#?Ob<Qp*}XcZKT(XX)66&7
z)3D}?Wql=|WMlHI{c<4*)mP$MX5QLZN=J7YvQx@B$!puj*b=xU39GbwelIQZv&+hY
zT1IPDOB3yHt}Q~s2MMaG2Zz?9d6V$&UCqSt(c8S)V+Z=uA0DjM`ZH|v*e5y94UV_D
z$ZLabE+_;>ALbTovnRF%v=6*oZNWJT3VV0P)L&LXFMX5<`ub*}B|l*#*+icc$L$n0
z6b>EU?d0rC&#EWbmq}Cx(ApM#Zv*y_=b-$JE(G49b)4169O7tH7D0+@6c{x7B62O6
zIMN@m9o00v$ffQ?S+PItGG{z~vvqta(cE2-Se2XWaf>+$DvkO;+Ml!RdkOkUbmEu&
zIn?(BR)u$eDV!~FuJj#zj4uOP>eq9eK~MalJ3inzfP#XQu2ynnC+hz1DAC7IEgZA2
z&}pt`zP&0z^r;)&8PF2sr?P{}4b2t^1VG+Iw#4ipV2Ef}%5B;|p^M>Z+c@OIUgt+^
zCa@rZHYO}HOi_zJK&*sU0wfAfsjFF|A3lOOP<vsc1kK8QK7YqAAi&)HFS!8tQ1q-8
z$ShZ`g_vBcY-qr0*B?f`c7=2I{#ju@Qm!WF?+wR#dQLvZS3Q4zXgpAJtT~PhXw6&F
z723u2*!^nW$=t2qfj(^0-YBI~Bygmv`E->6F*oskyUK;q2*Lzi$fr+&gN#7xNF-&J
z!$-0jD-k6j(*80BBFSp*M4GC5D_*7QB)kd2A#&-Xwwt&slpqj%D%U@lGMP-I%wn-i
zG{bv<z0l6S;mBc_+Rq}wCIr+uVv@wbETHL%5k)oyAQM4^Nx7zJ)_t0&qf#O}8ux?)
zcEv*QC@6Y{zpSfvo<59#IxqfsPznt?2ZP#AAl);FK2--OC}cck(BFs{ce51QHxr;7
zU<7r5chTHfY^pkqRZKLtv$wCB`Ctm>L|9{(0{sIO-`$utQkr$ql_v;w4*|&3#>NIZ
z8!N2vLRu0M$`DoRUx#B`K>RiS!5TGpeA8Z+?aRZ|C-d_;eNwKq(-?2t(v*zUIKB1j
z2vfbY#J=v<O(a|+?T-x&dH^+jL&!kh$$&lBPR6d5Mb5cOw}IW10yi~@NEkprr3zwp
zkP<i}cdGa!sAS6aN#$xKg5ViMV1Oh9@$0_R;XM~s@MWs_x@kl)4GQgafGIFn>hB`!
zhcu<=1LQyP;(>C(>S@$3|3qyW*n}|;&Iw!U(yyv6n?m2T>LyMnDDXjU-vQ7P8!$ob
zj*gCMphAN_?ej#Gvp_rj10n0q#!Q<5Ga9GEIv;^XBbYOid}xmTb64rDrLRb9Cc=rZ
zV0n-ZTH04~F0>6jd?o#a7y;d)OZU>CcNSpl#rq@pj6S4+2sqLS8_;yWzE78dTCzg|
z1eFJ&ofNq_bMqFL7^o#SUotw9QLotZx&9cFCh0ZnZEnaZEe{D}=<m@@r7H<BH9WNX
zS{Z_+m%6th^-bisqBGK;6bMwW`{?^SNYgS4Xg1jxxONsEDqtgP=|8mr7)=Lc80iDZ
zHOojv&`-!RC{AcU$XitWmVB-k_!wK7ST?XT7JT^!F8yk*&kq*Uo_l@uKr+%31Tx&7
zp`lb=&T!e#80sPl*#%Gco5-tKady;T5*v=qIyGoH)F$N=7ao>GMm}>MZG#@FP0qm-
zhxn-2W2jF?1g6rHb#GQcAQ!D%tLi`e)*+}v?ORVR$TG7n@{2h8M%1W;LG~@^G(|w?
z16Ob$zL}_FcbWtQ1bC1$GoS@P7U?XT=txV3er4Z2^amF{pPyfNU__WtAbdjsDw<JQ
zzux8Gc8lD-_tWHD=9Aj`Qr*iiW&2qn&JfDf<N^wciybjX7Y71!pK=DL4mGJXtX-RL
z7%h4DipE+@2R1q$T)Zk+2q^%|PCGlRt)+$Vz~FQVoU;1*`tUEew<5yy2~d2JMYbM%
zj9{_A>sce(8&Eg-wh(rsT!Owk-ni%0WrgXfslBXyF_Z>+U!NLdQWCTPkkNu#{}f}F
z#?EnuDV;wb#!jGO9}jik<pk2v3-#YDW~+4M4VOU>8l@q;1W1s=JA9lHkgl)k>FIp~
zyEdTcod6Dq_JvJHv?2)V!GN-y1RWDkhK7dffN1Rid5We;UG$tna^*F}kdPz#kw?P?
z)ox!iNtX=eUY@f;D?rBvO7vrwy^;0u5yc1_y7GeKBrhaMY2iRQ%{kP}9ETG=&MdvG
zurvLiUVzFO{6HBEW#uBEBKR*}AShNQfh07yx3@<?WQRb~5x@)(MUdI)(SP1nU01d(
zmBFulkPGZF)hOm%0ZajDkN))NK(b6Y@a>sk!MX^|1A8kWtbW!}dwqO&|BdSM-BydW
z^dopZ6>d*=wuO;#4nxsKE}1VR6^fUv2BWBtP+RU2jpOG9eJ5%ni?2KGdez2@?;HvF
zEwuQt<49WYP9>)VnUEZ}%Z}X^$)e%|D(HoV$;YejW1+d`0oh+?SWQD?5+Pwd^?8GJ
zu<zRu4hP&hf`f}vhr=f`5J@(|>LB<H(xyMMoQpgH(n{9}!@a!#2+nXiPQc|tr$r_r
z^>-GqlEYejuHAANy07r+a=dV<nh}S)uufUu2njQ1Rcn%n!>JFi53i9$=RSOQ{51{S
z<4&!}HJk1D^O<Ws>uA^H05@?at06m}CwXcx>q&)CkY8!@UwwzP3)H<z?CX7dwW^}$
ziILQc33ZXn^ORhirXh2+#*g=D$6>Gp+riptgYrBuab)KnaHJ0mJVO`*5XrB|N+WMA
zif#ZqA=iboPr{3#l%${eT|4bCq+rYU{0G<=pnH~p7}5N%4<;2+ht*X!Zr)_gu?t;a
z6ZMVUS8^H^d~hoPw`|=zD*1)N+x|@VaPCt7F?v_2T3jEU*y(Dcew^H<Up4a3M}Kvw
z%w1Hqgg=nz#w}%}I0t;e-81eU$<*PxvF<iitx|q)xw5OV(F595{8sLz4N<8*pFf{i
zmMfO8KbTWNdixVjtheHDXmp#Cpdnv9U($)9E7>W9yz~i@wV)5JRhQ9EO=k+1bMx{v
zgzE(1`hC$?-|N%+`;!2Ub&XyahdtDX1Qw)~Ntok8EhskKqf)Z^3$k@Ql6xriCm&?;
zjbzQVD~De0*q&&71T-1+nOzzcRaFSiWq=(4S1|-a2n694*V0+LGxU*LWu;GD9vS#F
zbi1D5i_61fBTCls&g~M!2&)Oq_=7$v&zIJ#CB|sOTJkpWchJ*Ad^g@S-fay}iA3Ta
zSPxJ-Ky2PQ;?Jyk<j1~XW79zCf3sdj$g!}nWKZ&PlH&xg@{KI)_~L;+NtCDk&epAM
zRX#}s{*{~xY_7*5`^c=$f*twIeCf(WzU|~2%19`a@^od=w&EmrS<YtRCnuY*u)s!<
zG+xD~0k3T?q4EzhxtFjv__#LK5%PHU18gQMQ(JgX2jW_lmIGzjzk`$pv{pi3bvnF+
zV#o9cQy+hg5LWIi4i_e1k|dm`?VHrg5yAURZ+^;g?C%wZgRdZi$*#Pd#;N1cQUsqg
z8eXjB_p571#KuTtZ=J{Imwd`8My?4cr4l$a&X3pwPSt7l660v%aM#tw%{N=~Nmc%)
z5fOsAq6R%p)+@J}hQ74zF50^q=YEc;)n-psskA51>rduiKFNGm)?G)<_Kr&gXUeoq
zneVJD&(+oElDatWORl`R)?#iCeF-pnp!{?pv<V+WlAxGTcTb7KM9;S`H|0%k)n7Xs
zt2Oh=J5Kd{js+=?EDUzSbd?WdK}QwrC09x+J6>CE^)6%SNp`h-3eK|B+seWwd84B?
zL5Q`kkP{s!`R=%OeR3xDsD7P*=sqYd;n6$KW3<>({L7l~iiv{dqPg>7p7u31m0+7m
z^PQ`Guw)^XIiBw5n*JK%gy%X>utAtj_8(<K5~=`+DfD)bAHNUj@N;#2dB&v$^z!&&
z4!nN{Y3&uqE=Fs@E-u9G%XQ|@*Z>6UHUa&4>xzjwi(Q-z|A52FepWZ22+Wr@6NGvI
zVZ$nTgVQ^9z(qnfYS3#XSc!bNicC_GzJkerWFyU60==tHW##a>oxVL3s@jh)9cTYJ
zoqB`m6gFsgn_3n)g(TO|uyonXz*4h=FogahkgBa>bZiaK^Js=;tkDuuvbDtYHx`ke
zt?KFAv_R12=Xo-l4m<aA3QMwm6gyip-2Ncn-(Z;&J6|N?F2rBhrzRis5EBxg?Gab=
z^pm)NMnuj7ZhSlezkXQ&Hfy)xZ=me=M|>Y%d<5fA0fNP=8%k*z-O_l>X^u0OGHv(f
z)XL#zH;#Nv-piHgVGg-j{0C}7=5u^qe&toS{pTvP^WDj(E<}dxKUX3)aQBIl=!x11
zEWah`;jM66eyaP<jzaU3rLh!Kvrf&c_H_{h4G?+(=}v!C;v90xXI=OjyI1_9sEXj%
zGHHQyI9W<=q9+EUV`L%BOzOp-UMtJd7`Zj37$K$JkyDI1x?iN-Z2GgdiZSJWj!}GO
zq9nHohY4_TY-&mBjq@cv+qTu?gWc7d=TVN5RZ5~&D~IhZ?5jpq6HA#Jf3zLq;<skj
za(Bi&;iukx=h+%|sCho*GM1GoOceOiGTkaKn46nJ3)>au36#sgkBxIrW7igK*e?%s
zMruYX?o-7I4@P`VpW{r3=j4Wt$ziVEI{h}>Tg*R{Mt)IPxP;GQGD&jyc2L1vFi)r>
z#0tPdeNv)Pf8MAR4K;olhEJ`+t<%|=CYkLGxTPxF)n-q2(}B*Gj3DujGX2^kiGl|2
zxbH89Ab^j6w}5YG80C6?1;&BhZKC4-{-2|0t`aI#Y16X|yYR1tLCc1MBMa3G(i|t(
zvgh-9u++v7eS#`cFH7X=S6QQVq6&rp?Fs_wkOAkL=A5h((^yzoAm{0=<U`ffCgp+(
z-P??RV{(1a)ovtWV&mXyCwcr$t&pz6>)FiZThcpvH9B9dUPoDl^jX=ALt_~UbdSlz
z^+KXCRE}1Y-}J=*|GPor8_x7`PDAvSAyTOOf1t`tyavaz>Q>s*f(EpOdLf5X3w6cL
z^gWW&)k5BXBL%O=*+8m*)<nfZEiybK(q-lV;zG^2*G&al+fg46M-3jSn7etg!VtAr
z9UAmUpr>pS#BD(ph9Xe~EO_(U|7tQIq)!SpP^}9!NA>zlVGw_XB4>o8Bvv7rHuW=B
z--Lij<9<kp;IG1@1szU6wSc&XQKjhAexSKtugn?epBtO$z<h&>`jcH{;8NHDZMw-$
z5nf2MD6EJgXiQ7IgRK=3>z2Ds5EThx<de{>XABfc--LVtQ8<{x{i@QCG-B)=&q!IN
zGb-<dO$fq9o-{Q}T8~eoqJD#rZViQG2coBeSa2HZZ^G(n(N$pg5k)h1%1}QQ$w-j4
z<U(DG=Ns3iEvs1)&Zf8=oQrJ$E@0h6joT7EIsTc0Bi!#1LH!{Yb-Oqa2<%Bw>9t0v
zo+>aCs(4b2-tr-lx}BF55|RElD7d|ReA-^nb<-;KH{|^sC8wvRp#(7;s?qNLOnV|M
zo0^pO`*LDkk<YcFs@j%%dwo)G@QxvFK}G!oC2F4!?;1*w6Q`4d=RxfW<1b#_+w?mQ
zhl7H?yT4!k$2SlXLvb{BU0q$OTavK?m2vY+VyfoVNI~evTkp7c$1N>~y7WgyYJV2l
z;akytT=3xijZD8+fW4~kD@l7W2t0F+#Xh@R{k^@}4$4SFFA^0S`4F`(5q(>pWlofz
zdbn>*A#QKNiItaZUT1&%Tj9gAF=ynAO!H-Vu_^Ji1I_+zRa+|U<c*lh%ah-s?9ZV@
zeGUJWw-nrP45|gS@*({IklBKhTLlT{ZREo@JRcPwfB%sw{ojgeUa9eM1B+6ll<^8~
zt$`WY248cco?mYiUz1w%)S{IV*Kn<cAc}Ia|GBYIos+Q;DF<`DRsHR}zL2et<UyXG
zI89OW>CXy*Ub_yYeM_c#rQiPBi-Leu@95B=r-O)Nhp)!JfDtIm=>CMOk48pDR{jir
z6(i2!Z~_GP`agqI1t9Q)x4zs?d(;=QP*q;;28=v)ls0e@kLww(yE%LykcQ)%I5{lw
zE=Y4~+QXaVg3h!xSfy=4{rXd5W9SVid2?!O(Oso5l^Td(2@sSdng5QQI4n%!A)6nL
zWOpEZ<Q^rvr>=3j>PT@577evhHS(S;`-4QNO6(;|02;A<bM50};JGmvn8Q!8l4;0G
zy2?gt=1URCt+Fz-g{<uWGqrUUZKL;^UR$911FG@YO?wCC5!wPN0nBw*E<^Kgyv%<=
zBK|_?B|aw28)!!8U8E5lsd`I68I7=$&TH=_0>QueeP$!262K7#nN=VSf)oM4XucZU
zr8ubrtLZqS+5)`y53s+&zFiA?+Rl1D5<nAaDgu8{>4H$ua0da`ia;Uf0W(1qTicc2
z{wkM#<9ATyq$AQG<n1R?;fyEg5i|LsKh<j@iRr{M*SpJH3Q*~Dj|~b%Py<U)nw*`@
ztn;EZMN31<T{r8PrvG6DQj`uv`31S(4nY09t&nH_ci%FC79q{nNPn+;vSSjW97Ea$
zK`wq0Qt)A%kPn4n_3DrhJ4XfM^}yKxkV`}i?AJHD71$r}x8p^=IC&!aD?PB$-i)w%
z)T&|vKr?nVMNJ=i7OFls6ABBdFHjAPc#}YYO7jrH-PYB0EUS%$D4r<}L8f9bwMY#L
z+l~y{gQOlaZ9kVZWP`vP3Y-um2d3>C@Br}yDS-+y7q_9JGGb63luP+;qvpaR;wq=*
zXo}QQsuO`U&#QYgWa*a(s4#T4`dPg#vHb=hK$ARxp@mK>1s6bY^M6!#u0c(mc^E%X
zcj<PEmtmYmYr$LV76nAD(S|m&wAGC+!OE>lF|^7>DT#t2;gZnW;<hfLqA(N{QG)~|
zM34Xh3Ed8I36MfWFlg`sM2!&*H-R91{y4K=cJ|YL*x9el;mo<b=Y9Y0|M@*n(sM-E
z!mF!p)d#^zFQ~YevSLSdTxmBC`^Lt{=>afhzq!4$yE_~1`-Sj8f90mOS{CfBwV?=^
zn&Bd0D53zUbCY7-ZbJ>gX1Cl@GoWQ}=N6%}nq4-WDta)XK6WL&XG@vL<yEyg+BCBN
zpi){@>grS7bvFB@qR*pSS9;HMS+EJP9=cDfLYq!53LZSrGMRh2`VnOZgq2UH&Y+it
zffxI$g&qs_69&<nnz<r@ESvB)!6qGct-g{nAMmV}HqyqN{j_2UstFOc{Mp*1r!kOp
zl3f=k->UR?85>R$HfsfwyHxD%RpP1g-h5rL29;3F`NL+Dd6V|1xZu&))tdMK<)sAw
zjU_i1)kG#HIcT@}`mWw<-w7LLWJ9=<0NuUVWkMAUYO9b!F`NP1zD<O1pG<01VcaI7
zp%IM>G@jV(w#~hk&1RQIw4JB7&b5q`t3NF75hoCuK`DMmX8tlgt(Q6kT^TALd*0&Q
zG9Z2~ltnpaO@&|YTmPcM=b~Q$S6dKz=-aENzP|=@_H?)3xmD8|6*L|nU@tnWSIyZS
z;S#<}<3vTonl-0<IVJDdd-weVf3!Hcit2|_b5i4wf>3os8iS@Zb<KKLgJz01;R_r@
zKd_I|1At&3B$9)yV?0jMq&A8Vk3PZhb~TVgVrbGdQmzp1qp{w^F%91!$K0D7jInrW
z$a<ltbyOX{X?#MZLs{-C(poJ&MwCbHT)cU|ahBMUmnsu3IFphWef@9i7TPZfYF}A$
zIwjBYc*4E@$=~xV=M+(&o0Q6$ed%das>a58UY9<wdJx~>qsNc+567Vo5|7^>_U~<h
zK+h!%Q-|04-sJe<PN}+>&BgP{F!*Z%4@c+(^U5Y}a1KAxA6M}d@*3Amxw3bLQ+IAH
zkQG@9^_92dO+^GsQa5N%8br{bPnc5QcvBY3*v95$hKmz-#s|1=4t|hdJ^msl`q@6?
z#V>bh`kj*lpOhgC;N0eU_AJU_MTLh}9uwjyla0WUKL4jCk*-JVu}pGrtomtijIV6S
zXgs*%q_EYnY-+6dPnQxx6UsK;?gRRr(}c_oT~80+{Mp)v1kDg~zXvpa?MmE}?EN+7
zj83)5Ioe{((U=VHvt)jj!GY!$R`@_V-&!mw7>TR8%7T(XRyGSbwZuaC`s!<_G;ZUA
zq0jW@gcYbqUH|Ne*gr&iy*TzjtMSEF)yE(asWQ-F_`Iqb!phpIXS1e`VHymDv<%Sb
zsq=f-v-)fe49q0ViF_GU6&19;1`QY&CdEJ-R2^(-`MEx)YAjV$H-G;8XmU;5KjO}d
zjEo!~PZ5upqjn?hGRspGy&=sh7;*vQcy`9hs*4x}WgGu|67|Lt*zi&HGt62p^V-U#
z2L>F^tmqk8#wlEX?#pb>yG*SfGT_HsN5e8RZJAhJtF1j?!T&>4fi&KHCCUnrj~wC~
zv0zhb$pS2H9ZnF|lcO#_v+=|f55<X&L1YCP|J%`mf!G<CK;pDA*UQ=VPFve?Nb3_6
zfdQkO4xk_0yP>FbI-KxuaGqantp5QKD7?RmH15dnKboPMt|5J#u3pCK&28`KI0d!q
zFD5tD!}C}U)^SUaQ%YblJ8TR7WycjSyS*oFbl8jk=yIxQLwOPjK6ukJ6V-x+1I>ng
z5z#lk={WmU=gls~SQNvg9MU@&j_a4C$wwUDTrIT-I-Ks!UA~`5!xUIlRCFlgKx{1d
zs?gqs<EE&oRAJ)Ig~ue%kMO6}8*H1(KMuQN*uav%v}Kap^V=Hp+b~F!m0otwC^HP&
z(a`X)6fqB63v{qvAei1VzwN?_vI`Z^6wz5^e;q(qo#mA?qwoti<8ZMg39_*gUj<}Z
zV<dlKJF+jfncHws*Y!89*{$P}MT-oFSspt#P$fknYj$t%-%?hONg?E|<%tXn@uQbZ
zTbs6ppYb<?hiAHuzVQkgs9W^Eqo{=6V6~@4ZqQXTa@9_vxxePK{xOEJeV2YCKw$g<
zSL*0uQ~f&w@K+;Irj(@7B1j;F4k&Hc{SfeAA@#FsgYJIkz-z!BlU7o1!6<1!<bk4Q
z4^_VGh!1cs<#&w1))_43zhLxPApGnFy3Wo_v>#bqN*2jO?YZ$N2+P{A*YCP+0amwM
zUa^=PZ4eoyYPIBAlpIT`CwfNoxQ}{T8s7eJG0vU62tNbBf7!?HhkZ#%g&*z04&B6J
z-baEbHVa4vboblsg@f0A%FOziGpXYc8%wJeltB-L6{E8Ce%%qe2f#0tKRV$3DLicB
zUMI$$+TXO_>ja3rv~ct#EUNrwxD740v|?q`t3#zIwq+x;+39_nOd~|iW}Ho@)1^~r
z2DT6luI0B+`100E^$$R1D(2Y6EvML03WWH5jf8@KoUM25uyFaGhUGEYdroaou3F8Z
z5LhzGE^U-p>m|K-55*Qg+{M>FY8m18=(2vXuWl|WZJ_bR&oLn=929mh=Eh-4EdhoK
s+*7W<fa4YOUI8k5r*W14n%}eXTmSjPiI^Zm2hPoKKjnQQ_X_#;H+4h;Z~y=R

literal 26066
zcmeFZ2UJsSw=NoN*Z><U{ZpENf>H&f`0*(Q1e7W@DowglrG#JuMFpiwM`_YKNC{C<
z>Ai#=luqanAOy&n%kSU&Kl_Zc$G&IWJI=jl+}|-YLRQv#*SqFB=kq-CdGq3irYh5Z
z_WdXniV1z~FKrZR=V$nPZSNlV&0Z;gQFz&NS6%flc!fVyl?oiZ?7M%>&;^A$;)?ub
z7_#Ddh(i5=LjQGH_hIbJpm%JyTO6G}(RiOT@X0mFz{lf{nBVL&GIlIU>4`8dqp|Ys
zJL`5urm)=Y>C#%0>);?WcVCsAZQ|8~N1jEoyY~!y7!eL+EKol5JZuN+A=#!z3hfCe
z*Y1(KIWG^4yAVi@E-~bdN+p-Y@=u;exqX-oOw)CRcf+vCKW!MXAj9;2f`aiU9@=#t
zet7nW9|H<?W*5Uj6iWHj-rMl2{g+WYQK%bxcbr6_{Qf-j*?Y2-u;sw4w7K9jE|HK<
zIRnSc>|v-3C~@eI7;A{Mp6x3x=7vl74gS|BgU=p&{J(bn|9c-P@A85Z_^BC}I-1UX
z8VP!J^1^91H#e!cl=tsX1#r3F@}f-Zb;ut24EKMcz_u$RR^C(lxYB~M{-&qHRW7e<
zd+;m#QgJrk#>NpjrjZguGD0Wl{e%?nwK=J{v|3~mTlwvK@=lqPHpj1X$zDps<Nmbu
z9w<#}I(LrRqlF7$iFeVga^Ip9bEoB*?1XuC-ACLc;>HF8Dld)r^OR|QcM4gV>po|&
zH5jw>lD_>QLe$iV{9wO2@)%6Nw&0dRD_tf}6?2<MvQJNoueP3|j)~z6$v6M)h=`_`
zj!(fDOn;g&w#2?yn|$H#cW&%;d<>7}Sbc)DLY%=S@fBYk6aBvP_TnV3To&OM!D>`1
zCVptDMR7@tr}py@-eu<1_@$z`MP@r}xkrn!KXcufVEmiZ$d&Zw1=m1cWQN}R&^*)&
z@!9A4`}>@!PRFZm<HyA8<Y`{{*!Y2RDH;0q&sXsmGOMd*<T)82%JwIPjW!Sr-|h5T
zt2Ykd4k=?fSBH!yl49fYrOHn9lFXLZ>cci7Z6u(%jsfq*kCt{I#b1z!F5Il6P`5_3
zM!Oi9vwj#S9{u}t=2IhDR$czDqr`3hYL*?BLWg(~a1YaH-Sa=Jg`_$aVsr<m$sL*R
zPoc#>Ek3)fF)O<RCj8isMn>NlKH1rD)v-L^tq?xU24zszqx1QR8QPLaE*d*n$htXp
zRO?>wxHYG1R}j`nJ*6mIEj>50AzTp8&nUf`^Q+>v*1p@YY@T#uHYFVh*u-x0u<W8$
ztvV6^8E1U4Zar=Ye`d>jwQ=DQW3p(SkSb?{ainNVwN_S($etDNFWcEE?S(BK&ZhIr
zYc?X`W}AC%IXKvu-nT9xI+3<yN`fa>*XGp*FIQ}D4e}uCF!4!ybWQszEx=acDz(Jk
zw{7fMaCTklto+x)`4fa_zFC>fjqxXhrYkjs_to2DCOb{kmdIVf<L5}RF@=t_JKLK-
zT|BmTrnLLgXvSX4#)TZ#&*nRR2jH=<QfrsIb!k)&x9ygS<0-SUs?MQHISe`K&I6al
zv_k9$O3s~|+kw||-gOeS8lsn}KjuPT)1hvx7@|W`BiC{l|Kg&|v?(~P$H<J*edi9y
zZcHZGDKINvaj>yzlHu_=!{p}XxtTPPGJ$c_59iAz__`n8nrbPtZc32Y7!&Is!Hkp<
z4@8<-DA}q)PBqK3s0)c*FCo;W3o5=X!)SP|USnsE&!XJovLB4tv}((iopMlW7?>Y;
zI}#+H)6L7E+ZZQekdm!iKYBQK`VavvKIX>jJhVw~CZ14u(<X+qOY@CWp;bwW$k;bl
zYh8-^xszqG)9$7x<h0etI9YmQtmpIyHRdL5?!qH8g-{tDgZc(Wg(#L-S;aM_PiNiS
zXtXXflzipZ!kylw7YE$SSGV8hHq%tA%D1S2I9YOT)sTX9Tz<zaA;Ec^yQyaAC;M&G
z$AWXlC8LiR=UO;|dOb$g%7Wz1qHR?4b4*~dq7pwquF&oEXX0pIOBpOs;=0b&nk!1~
z&d#KqL|K-Z_XeIdt8k(dGIfP#4RGA9QoqN#H9Idy>#szioi{Rr)aD&&t9tPb2?KAw
zV9iEd9;XW<dE0fLb<jh%s&#TjVRmtGchXE}NKHbcn7&dpfkQ62JLy<=)m95H>c*j%
zQi0`hK_#l*Qcs`7!spSaL_zw>YI6Qs{Co_bmi;vWtZU7PmL@ugM`vqo=ZynSY;t@~
zW)JTFXpWBz#{0T5lS!r3sz<y2Ob^TI%ex@b?NmEP_tx4b)!|B8cv9x-H=B1Vzjhc1
z+g$D(tz$x)nu@P~9&J++EObOWQXe=ETA~S&5rlaO-&F#-0mJOyxK)MPQ(n?SpuVQg
zvu28WFU~L3ypFL|(B$qf5@Ol(G*$E(ATMn#&e!GnV(k>P-@_XGm@kIIs<u{bJsquF
z6HxHt>o*P|JE}QSZ!XOn2+}tem)11gIX%}`*Fvkr&_SMhOC6+{i>EIQ=WcFgW_O0X
z*1GgYkY0)Ra^mc<b|Y0<5tyHqGAr;oL9)-Txy@6gME77q(eUStLotJeBlqU0M?>du
z)X*~T($NG#zEV10ne~IEmGR^%-?cUDxbCD<4tl#t)N{Tf<ZZRbu7#F)HBC?Q%a*PJ
z2kac?yljkt-aLLMEhmW^tIpi~+$flVbwc14>V}GzW6vUXbiw+`DjO4}j%-#qvRE!y
z-@nGDv_TD;7pzqD96Kz}jU%ttH?g2-N8wRRL2Rs7WqKEN66brX(Cyt31Dn;!TISBz
zblQ`=m;7(a2koUVYQ@EHpjK5iAMNpDXK!_?S-V0Qe2|pk$kdWkT*B`$NElgU*y+wW
zHaLh4wW#Bj*UZ-p@rWxNp-IS8Z!ETiuEmv;%gWtwHbGv4ov$7?-SOTW^4l^<NKlNu
zM7@<=;}>}JGYlW^C6KEmK&9a8iVH&vYvwQF%TlDue)gl)8-ijCDAWci{Md|lzgGN~
zn^FAQ+jlbb^HJ;ZC*-3Ccp14bNoxe)bC#UgdnSqp<OQ=2e|A-XfVZ7VvgWtO+SZuH
zI?>VM$wJg>bNR<5`5sGkE)67){#7m_y_+Z~gUcR~s-`c|`5W{F)D8Glnfo)HSW&k1
zQZO)cEGA@N;o0blPRLG{Yx-H5%#)F^t3jhb#00iz5Pcs=bDyl%zscYi{0->~C0ryY
zf=h0t0;<RQA`}an-;bfXQC#1m-V1CnSCkL#M>BC~dz=e{x*-{a+Nm7w0JWawoae9H
z`A;lD2)OvT1D=dtcW0+A`EY&7105kOZFP$*zvHv-U&!&4sSZJ>i*r(*_tDvcy$YE`
zK92b5_-YoV%I+kT@|CVl_F+0TfnLXIVA6J{|LMM?;*JBK_|WR=>Q>FMQSx)2c(-V4
zEo|1YZ%2Zas<0&6+?~q&K4X~@OIq}Y+Ra>ai>r8BB9D@;9l<!l#1ml11U)8x!;7xT
ztW4BiMwwOlc#7gz+AF8^a{Hz>^bAbHx3_54<Bd_h<!*MVnsL!)crVMwcNe@ZxlqD)
z;6G*53VK{HZl==Y!Pb+hZU^G-f$nntl#*H2g!ef)#?uv)p{c&4V2A4=mD`^TtbZ_J
zn)V6C-dM!&plq|p88#QHStbVOeb+RJ()K-X<mW3TTqc?|R8&;hIXGO#qpZh&a5%wb
z%rmN(QQ?xdogX(RYY1xTx@yM8{i18en+-yS6dGB0o6N%QtZlYdGC%8@tt*^TRW)=I
zgDNkIW?Y!Z(P@p*QX+)SF|m$JgWOp6J{vZ9cl+t->4Ll810|)TQdEPvXK5^K;(Y3n
zq8fH~c3ngcc6PbOayN2!n40|D%QA<4X&f4LJGir}3qRcQZXzQ$S6yoRd3r8x5px1n
zULL(?0RV1+=khowZgWwVzPVg@r`q4dYbwQaCcj49b7?fouuvd?+v{FH!c<3^RtVPx
zqgl$2A3vDwMt;0*nC{B#PpFXHi$Z<gPD%|p@4Yhl*4VL(KNmY|r(<9c>ASTt_V|Ek
zEnw@}qLLB;G>oigvW4>z_G&RP!EUtnt?X)#xeHusJUrLOQrc-~^YlD3AXzNBN$Zhs
z)uWw^0=*A*xwj@>HY~K#2%w}T_{ssWof@q0=prV}FDKA_Xd_o%9=C)i9p{yc0?@7?
zrLe<?XzDT6k+AI=;1H~^dTWD1jvT^!TTiyW_gw!O&>y@<u&zE_M8bKr7A8gy%LP}=
zNlIndj3!nJs7IWw!><j+Xs3sz#d`@GV6biinz8T7%ga~hNlHmqU)RUVx@ZfB7r@P#
zxHL&cU3_piMEBp$k*2&h-6ewAnl^CRq7Wy3y%cLsy!^zu&n9im!O7d0rl^yq*|aj%
zA>q5ZevtFxdFuqMdywMBB)$Z0TbX#nz$?E(F<<@V@pJVTYB)n(S`=iE5oFgeBp&cw
zI6;rqBM?e{<;vfyl^5Um)-SwWu1xF@chWS%6$4ny3!d^_rKT%6Kg_O~JjP>j;6LR%
zFjN0~Da!v<BKEI={p-2>FERK39Rbsmwqsp}K-Ey(XbX1l>27ZiA?Felj|yC`^_+-{
z+<oBqOqUysquk%In0uH_OKHosYT=pK=<pMkxxB56%p5iP_LhXpc%%JbxfCP-ja)Ol
zxK-mje8KNUpVz|4uw8FZ6+*09E#@jgYOlwTr!|sw{QdnUrKQIkC0ZIG5wF(E!BtFa
z6c2xf1fgbXYWm|f2k{OqkX59;vonla!K=uLL)_vTaO>#q?(U3SB@%D&r4WAg2%sMP
zhQ+peUaK>PbtFXD$@n;)rtH*Pz;9sYX<4<=l{?OYTUQAvma^*}ZB=2J&CAwGP(+7h
z8<*;1@%j1r)qT%Sn7!-DG$>V48hxWu|2x@sYBud$x5LnkQm;paK!%-n#T%Cf@kXen
z;LEp71QUFm;~)Nd!B+=qf2qmI*Cmp+1BJ?SO1D8Nt+c8b5`3{iQ2XO@SlA`c*mY-3
z<;_l|C1}EZd9BTvQYaJhP>}k0#Egp0L}GkB4d+ilqV9uFlf38gWwf~pm+AC00d!wq
zALL_a`2RS}b$#o5m1U^TX~o4-OiWDZkZ=(LE2uQ-nVFdzzvoX?!|Acx##Guw1!>E1
zyeTFU%HP;hgIG)6ooaEXp-N8Y+EY^`0`BT-_HbF3@nl_L%dHJc*?6>__=8{H_|pGz
zmZn?=-<*w&jZUJ6{ZhJ4n$^b26q}N-&mLyZcZDsA@~S-86*FT-Ah~uQMeseQ2|X!L
zsm9)!2i28L#K&ajR+V>;pjw|@T7QXa&LClo7l~Ywqg-s*Uk<D|(?qgOl#F0D!RQA0
zLQjpW5^b}SN4B!t+_heDQT8)Dz#9(#rd`1g|H>KvpUWAD*!HjI^1rTv{@KgmGw4Rt
zv*3aG>xpp}@iPX<Oh7VBL_`F0e>{!y?WophtV?tH>Hr*aBf_zwP-lJ!UkH?Bbpntf
zu5GI(h^xOW9o|j+3Ma5`eD}7bq=c_786D~juq>_0cY-mPk%c?<sOH7@&PO3BDes=q
zM<qP?!8RN0R6dn@J^oUe?M@Jlm8HV4N3)Eb@Wt#PccYZ`X1fm$Ptc~caM4aUTEAhr
ztBsMdu}68!CUsd}3dCfTu68fOCGXXl>+jB=3gGd5=+D9v51Col91+q#ek|Cz?vg+=
z0hN~7!wZ7;HbHw;>a&PJPU8!~_*<_HkulrWGY?-d|6u_TyXzeyYiFd+pyH(+jE3mj
zYu5cG4qfjct)MbK%-H$is12C$2xG@OKZc#ozg`HA`|V`DhTS?Kzd^>(y{_!ph3LX(
z<o$39Pk1a=Cfm><HQe535mfv=0PfpQO$6TCUQRGDhgyXCm{E|6NSc0IH`8`JTpn*S
ztZ;Wov4V;4%MhX+o8&XBeE2{>BRZ^NzDgbqSvQ2QI)J4GYSU6V5${-vv1^K!qV^H1
zB53mjr6+rs>Phr1(nOe!Cg>!lA;~Ez{1Vp9ja(iBZ-Jjpypm|HU990DZ7+A3{Z5o;
zXnPlcR3ONl({uD~l#G*EfV^^o*K|6Yf|qLqNKeMO#qgX>?tj_sxmH1ngZw;t0JHw5
z?`mPf7^BkqWtyUashb{@1aYXhK2z$L`hhYR7vKZ4#ebJ*cmN^lqIH15I|c?~*D5Cz
zZ0_B=hpCt;Xh3R|L`SMde>Tx5|LzrLxyi)J)#6+Y7&_$kP!BxWv}|q2JCu(@+P*f<
zV<_syQNdJD*k8YXExX((Aps@qA0_?fs!%KN!6qQ0#SN4PBR|F}mz&|Va=)JQI)7(1
zC{iL7SgKd8;pfMYxhdd%6B8ZXo}?PQiuFIp7LuYKnOa<2JWXzSP2;0><@$y<#mESv
zxnW9R+}5Q-g7B=04ct*4Agyghlj72IbG3j>_dY>LB0_nQv2%LQ<fGk9czh<`2e|jl
zM@<gE8$b~`IOMZT-Fqz!@m_AU<<!*F0@CmEA%g~j<&}e9?K91hk&!9o`EsrX*p4`a
zGtyE>pf+0fmBKai+Y)$4U1qqPx=_Ih6`qY7iYxC~#)>*L8=%s=mx;mXXb}a$D2!{{
z6*f87d&ke-dQ;deCz73=ZTRW_9d575_Xx8s9dtL25W|?Q{sElM#@@aW5dhGk0+tOC
z=h!6g-RRCT?6^btA2=TmeUayHOpvhAnO5(c02k}L+N<ivc<|&KfI6eFj%v|@abd9H
zUvVqQS;s!O>I<PVRz;g@fC}$UE_SS<jw95Ks{|%FB_~G|l19rvVM8$%J60dgChKeg
zR2$*%K+JBsqXg@QzdVt^ZEs$pj)s=ew`neGgB}yXxJ~njvxXsud9Ti_k}H2QIn@G?
zs$M1AmfH7*kaH?T=e>v(?X9hE5J95z<IN}uYfZHj?e1Jmf?ZCTz)?ZX>meDr&;@CA
z3&ZYTaiNfjVgNVP({<|+@CJx$0&wKeYyz^<<6EFTRkl<wQs_F>eh^SoTCB6$5VhH(
zWM0B;y7PL1qQXc`aLeaMJ6%?LttTdvf`{zLFhQ<NZYd`!$Q`049)oVw8Ed5&KqrzZ
z+Ct_vK~2HFD@lBHJ^AJ?4d+Uety|(*dE7At_u4*ix-?#pT^n$5p>!D#y*N?jZy0fl
zEudSNlI}ZeF|jWNy;58(2_%TNQi>X5Lc1~NK2lE3n3pGDeW!yt&BzzzQf71#onW?o
zrgakPi)=bV_~3L}a6O)}XuF)c_7wGwQYR9!Of%LKVVtdOmI*tyoan1ghIzLy`?HX4
zZj&~;OPwKt$06gRLwvT#n3j(>lf>a}>ot+4W!#O`!^23nQRQ;er7g!emJ7{$lW2Vj
zYqR|_BKlb$8=*&ux^c>qaRHwG(n^OGEF@SNXE9ZmkQClelyMw*2dki719S`&>kW65
zR+KnDGL0St+F-2DQr%QrlByYfgD8kb;H=MjHBU30szguw^#rLL)?rwt1?ogX!m|S>
z#ues0wWi7BW&Au=45CV_U}X7G0IX=pmlgkh_^wZoG$gDVYXN)~0{E;P;hKjG1i?~(
zJl2!v6F;c@K65zN=axx`04Onq?z4TX4!o*C*RfmN1U?w4ZWDLc^snzP;Y9C%Bv^q+
zGTR>j)dt3)Z+{Vv_!w5O^C4xr3kl_*OG~wlF*3Q;he<kG3Q~}2nhPu&gyoPkx$5`I
zd?_CdQ5VN<tXz+l6silddz|o~*D})b+w;RwP<WzaVnk&4TP+=cFd*E-v2yVT6u9@@
zBSqf54Q3Hgtz6e}oUrUm;gWL$Wo6SDneJtZpz_7S@p0_?&-goGy&Y41);H~EW9*GK
zCk9HW#A0qCw030j9S1uv#lGk=nV~S5s>RpE-1|^|`=Mr4xRE2Y(~BA?fU=8<id=2a
zB$ACGE}FN$dG;>M2M9dz;`6*B-Rbpl!g$qnNOpVBaJ_>xnpe%7Ns)xx((Waq5_~tN
z%2xaAJj#Y3-Ro$|rs$|Jn38Ig)~4()(k*c>NZ8JwuY2zs5ll8x8f+EzAt3+}^-{F_
zj7k7=)8pL-oVyLJQ-DnzI3e1o!lDocQq)*UzhlGFSVQIok=AFMt6#p?)dXcxPU7_X
z?J&_8GrUu5G!sgBB8=<v&OHZffq@}XTC8h38XW?$B!~{W`0M$A2MNn2ZGY=Zz^?K{
z?U0RfBdtXu<zvh0tg)Q}RDC28|2^kqVF-$u`q}lM)*)06!d0yW!2qm~m<aROc}V{1
z@Q^xkF@zN|F_+zi;mX<<TBxPgyw7Z?DxcuN>puqLE_H?YO!OoZT^w!BQ{zWl`a_&v
zqc=9k%wLt1@RweS*c@)DOEL~kjjxku6ihp)Sw73C`hfv8SybTkxY(v6wJ}~no<qhl
z>&GROe35{S(ClXBteD3K+&DGIca@F9C}eFg)MWN@Eg+@8y#N*BhGnv{iQWaa>PvL)
z-ot(J-b_G~W1%#(yt}{)s$~--1*9M*S5?jU@9Dvk*h5$PlQN~^-KIiD>t^fZ%_3rD
z&Q2K;c)jT6HqH~}qZPsY^wA^CpxEWi5cJGIs$`S0)va0z&mHyWaIA&w|BhS1Z2f{j
zQ&STw+o3a`nzuRnH)iV;$f4Yny4?Z%OgP!WGGo)FHUqvS8vj-WVNyM-clDuw!u<z=
zjE@G_HUy84v+6w@yrSh$A5JjUer8~WlV0`Pj=K2pMs^LVM+DH>_m?Nq0Lvf@H-szg
zFQ>ku^j}XH_|Tdav}cDm@gwm}y?mZ6boYUYvNTEhqL{~h?!o!q1jCARQLZh8w8~^-
zaqAUP)t=S@A=1MYnzy9l%!{4zW_sZ$GrXH|RsOh^-b2e(Yev_30l!CIVJSG_Dc!b>
z4999cwq32IOXhy<F=!#`tM$p{YmqT`KTRwPZ<M{AVZWXD+3}&hhNJNsD1beJJ{Gl?
zW$CvgzbPAZryl^UU>ph3Fw)ldL{_aLc0<=zH1U8jX-T(F4VP_8zw431usxTLp7*y-
zAb9)?N~(k&^F+}6S{7-sx7rQOJo9#Qn5nL$WYm;iVQjwQFc~^WiK{%$=B|^wal*_<
zcb~#ESv;?T9u#U}mwryGcxId)Q2t5K*l5{8-3}R2mYq#HWv_DhY+J%(s6_?>*WbMX
zxZ~clohpMD0*SfhrYe`-+n*?RGW&4(1oqX<Ethzf>9?Hi@`lB-QPcp7uV3VI=ng{s
zK$vH^Qs{xks@MxS`Qdki2^?}izhE|9^c^ZeBK7{|w@n?5F~ckp2YImOO$p+4lvdO;
z9d{H7N8dtpXNfyspPIOJ>B)kg(cax10kFgJ$166Z&$|rLfab29;rWGjep)9#jTHKM
z5c#V^U5@jU)>W38y6Rd9yB>s8j;GY|#)wD*vZv($u1w<Dx`*^LB$_WKRj<C~RsCHK
z5IE%n@U%%oW>7c()T15!y%c=5rCDKrr$iK`oJ-XYS1xLZ=l2|8sFMv4U~#;@z5G^9
z;{fWjV%7|5`@5L$+j$(0=%J!O&MI8`!s43;8Elkrb(4avlCYZSw+{09SyD9^vK%nG
z?v?XJA+(hH)*RHlnSlE3r%eX1GI}9vh!TNDb##u-d49&^x4pg!+^PZsIoM~Em%0$3
z6dMsYz@MutJcdZ>%?fJ+fd?5~?-=*H4}^PSh(=yNHA>(UV&Q;LphFW6y?}|-7{~hc
z#gXM{1A-hNV8EuT2v9t0TymC&zFxBo1JAE+!OocTWcIWwpSZ7^Fxqym?lRt1q3)$|
z8MQZ9dk3%|eRmWNz=0YTiv{^V7Id}k-DCuhL91x-@$8P1lfzt3Ldp6939<zvHOFVw
zAyvt2pOPiF8fonS+6E~$M%p3c2d8q|=2k-pe@sZQsOFI3nJJP&uYugdMV&goxw=`L
z+K~Hg=j$XRtzCbA1`=oF=-28HMri?p=V|Jo;!3bxX*PO0T`0=yFY&DW?hXVq>^S)`
zn^@^uL#%>hZ54n9>D^d0O@!CmI6gmk`H!BW_c{!fP|mYVD@wErAvr5|&D=zZnN?mi
z|M3cKZf?GZiM_K5yT0%f0m0w>0rUB(3Ca5z<eJ4H!BWc`+z6l7#?WWfWB*Bc^xICs
zl@bf<w>gK^Ma%e*H3CF7gb&oMS-VJ6cf+f%PQHcooKjX+=B$GN2f=XcE~E-aoxk%{
zlvwkcBeLI#fQPVw_fh~<43@hAzU-qM=q5rMLn`7w-uMsjAE9kioj2e{ZrE62cKWnk
zIbjJc6a<+y%aOC7p=0&)>oZTlX5zqn2Hg7{451AtWd*g%PXV4Q%gOV9Al;PFx=@#P
zv>-qFYA|;kpDaRwNZ2*Yp1eD(Tu0#Sa+xzI&5$oo7_GNAsfZ}3pKTQG&|i|0o-U-J
zp%EUBIJ$&C%%FTpw4q240VHK)(l7+yDPk^0Y8)Z4fx<L_TM!2IGrfg{z_x4~VQx`+
zP#cjl!+Z5rEtDBp;`*<n;P;8Y>J@iyw5y9X0g!~c;lnR8a9>69(&9JZ$0NIGQZh0I
z7Rqqo$EN^h7Jq%Wv9YPE6HJJ;s1a&If^Gdaym<o+h%&o$$mq2!{FIonI%)nUJ|8ZA
zX?uNS3dB%&+F^5p6$ZBdApTB|!}*XB9y7F(x&qj=+MlW63nNb~hyxZ&qjkTTwlzZ6
zHAR301Ces>ys2tA(~}>i;O)+=u+oaC$4FxcX&-?ZE9+MN>79sis1mN9rwWVtrC8NP
z%=*_8zmr7uYH8)zxw3_?2VAIQk-s}p;Ra@g`Tqze=3#_7zg6Y#9oyJ_!sHDA8(qYU
zm8T^{V;=_DH`Tb*)WZcDnou=}>F<$yY=>*$)&p?0Aa;0^ZG*rd=j6K_BQJFf-qr2n
zHr?q1?X`&!9{NKNrXvk1f3X6$u0^te51mG8sZrWq+-#w5YW;UH!X+n(P=X;!j>j0H
z7072v0hR^kM~xv+V*;7>RxFLlhcxxuD(v>Ydq?F64+k^qvl&E}U|sE7-a;Sh(pkd-
zJ{G0T`z4Np(JIWceDJuh1#>G1EAY;AE-4AdlyrOTU;UO~7BcE*q{V<zZUdzU_v>gv
zEpD?0=lakD`jkmQa>59G&||cif43ZCvAtCszr&Bulfsc(mq}VL7Or$QWi>mBTIH!=
z{fw+XM#Y6k2chqZ*<OlVE^EPxQYd1+^sIOVZ+#2pNnv(xKCyWNvd}7%)?8be;0?*U
ztSDiJ(rtD*C^?l7a`w2Po}P6BTq<}SsJgAYxhPAcDuP{Jf!OL-N&?H=SrnMkV@-~`
z7setn@rxrh6X|L3_3|WY^MG^1IhfRstM)MGH!i?<U@<p<h?2nP3A`u9qiQ)$j|(A2
z!n5DSMh9Ow`HXPP2FpNhEw&@-yZ!<1eDp8ac@ceF%q$jgdvbpM1rTa%Ns2z}3ujF}
z<=w6RoedzdLese!CfNnT#A>M0lWT4d7DRR{SS5xf_J)94Ay}J*Y1P+6%SJ;G?inhe
zp%%g{Q(X>Xium1c&#%Wwp92^c9`k#dr_b<yMnn;4|Nl)527GC=Kn;+wX3wDJy0iOV
z`u=c&N?>GUB=>hv8G#yzNcDYoxr1X~E-o&bvSVSIOo%8Jt`EZ52~*db{>)s_(1RJx
zDr{DRCirntlXuXuK2vDbRPRtah`2Wm%=O^|J?5}tK_)AJZU_i`vVw80+97{yV!S~Y
zPXn9BfN|;H(=GEbAKRse_s3fj5`dFicV*~79W{;yod#SGQ&c!%;_1~hehF>R4uUm2
z$}i@L-gEHejr4!yn5upHA!l#p{eiHEM~vL>l7bcGetR+3=Hd`$Nv-DDzA?oHlbzTB
z3(I@1Ob|ya#l?X8OH_vB<1$B}Zx2AD16lqdqtMZ4fu6O9BH~PSz1o|mjN9l)Xa?m?
z9Ab*OA4BUST5YZ`+BiBkt12$N^`Z5Xa(dpt!L5dtJPg~tExr+uoMxS|a8TLSYX3i0
zEi%EirvfZM_Np*(a&SO~+VjulzT7M?masW0s3RP1mZuM07K9YmhY7K=A#$OxWoZsW
z)m%9lflqu$>dvX!n1;iZ>S!eTHs^zHhi&Za8o&nvpvZb0vRz{_(agB(J#y0~5W6SL
z9^NT&A`NX<0wzU*N|DyvQgxdId2?cOxuvDoI>EF|2st_wbX?3#GzPBP!vl;PEUBoO
z;Qy+t9JhZMlrzM<8y@{zNVN*yfn;<x39IX1)R4H>nz%ZPD41WL9T1TQ{)*C7>t(0}
zyX-4YS#ecvPYa|!w6S50`PHeT)il|bRA}E@5XcNH5f#^!J;3wnzhlj4P<dLbKv?@9
zXfqrOR8-Y(qB$;R-%)`XXY9bI2mkRA3_!vSQ%f-=-2Oi;`2RsT|A}|#fF>J3b1ifp
zt2Z4w>;*MFjvWkqDB<*(3S|sne#FEw2WlKr&p^!XpMtg?YDT?YamQvZeIpl<)|SDE
zk;=XdzgN}^11tkM&a*4iOz)ricqx9?0TH96{uZZulXMb6HKI#nsPzVj7A>may*e=E
z3yW`d<vyI}h7c@%Sc1`z1+Bh5-6vK~%9&#ZHi{|+p8dNXfNC`412TD|I%cCCF^2q8
zvWG=c3rc{y=d=Ve)|d5F`?>)<6absDZ3F~@(gR+BsGViQj}H`pK>cqB1SHi`lCL+y
z3dYCBWAA_6#|91BX)>m#4Figji1;7yVO^x>0qF<&6c^ATOdQhd*KfnGR}a9DTR@x#
z;s-zK=|R4O*F2@ZY~Ig2>z^6hA@guJRpxIkVq8XE%eZ^}E&2@ZA0}<hdCa!a25OMw
z2|d{;-#Irz4x;Se#_7UoKo2*)vI=MQ!`qbZ_yjWOmL05W06NfvEyHMCS_J3uM67KK
zGC`KT**b}^KI3NlOG}lY(?WYWr{q4<)2x9kDb^BWhD!i&i;9c+(MW3mX^m4T@=6ac
zJzN+@?l?T=5F96byL87o2>$nQJ{=k|mVkKlI1s@t<NZv`?LjYccGo)uv}M?q2B8de
z5NB9ZbU4rULV!@XnrDOYPzVUazYS01|5*p&=R{`NZ=v?aGOFrVph!E&iZGkJpg1FP
z>2C-Wny_gRGWuV(<8Qz-3on8vySb94)C5So9Tw-xOs|A>yr-2fv`VZ&p6<=NqcR7Y
z{`J_47XXJN;&qgDd=yAS{OB$sOzfnbE&FF+-0Fb)A)RzpWrMJa4gnhxg4+Pr&Zemd
z*1b8Uf=GS<rS<>mAmVXty^JLN00+QJ!4fdj$khR2*S!zm?z@9W8uRX)3OTD@aPENO
zd7lY+PpHcwe8Gwv_mB#->J1M%QQFp*X+Z+kR|G`fB}%6bV(;ol<ogHTw9`9(dV&Ru
zj7TQGU0f`j7w`Y0J>M(1WBmkS{Yp;7T)EGr3J>B$>Zrp0n+XexNNTV=Qat8|^f0(8
zE#IZkUg&0oAZW4<Tmn7mYY;Y{hjf-vj!=!TYlD%wzLZR{=gsd`0>*y^@MuDnW@l#)
z;X~S905HHRW&quO?-34C&-TN%4_9Dr5GXpIIi-=u4aPJC^pvgtdg<Zytr$n&Mg$))
ze1(EV3o7#jq9lKNadho(cZ8}&Fp9ofN@z;eh(TP2C$gX=C4%=}Ywk?fkpO}76_1i4
z;O0iCEpd<#%CQ7+8$y|s>Xh^&t)$9~K{w5_4eJG7HX)r<3s@FXh9I>;7k^5q6(TgR
z^nbd81m5{Sv<vS?C*b9SgFjT!FK!R<##p$_oZhup^^d*J6Q4gOoPK;p)vha0F!A0a
zo)V+t$J!}=kN>p$Qlhl`a`yiEQ~URF@9w;N^B8M(V&b(cJ05@F<9kON7^H6H^Nl{X
zPzmX^FoUfZ?|SDOn|-X%=FqC}A)rv|e;`x?g*y8Pg;WMq-mV=;Jx6^$viCFc`mY22
zH4jJ_{Qo{4p62cQOr=uGspB!3MKm8p1xVrLKm9pY5Pj0P#NMWJ7MUl%@h>*+4E}e{
z2nwpcJRufAn$Io30`OY$!W&ijlm!L`N;>pOUb%859b}W&1>kpue0S;H(9X)c-}eu?
zc4+h!+318j^!E080Ss}2`ds((e(kKg9RO9N6P{#`^cql-JgJO`F=3KI#1H}BuL<`v
zb6!mR_U*4SlbnJQW-hsVfDV>uTTE4dU&7~|?*Uh#3cT%cEe#DlZEfvmCGXyyL~Gx?
zdGm>rN5vp(iNfNyqxFrAVSJVr7H3YKI+bh(q!NXCCuqN8;l;%`R&~jPNjFO)*Sd|J
zrl3P>nXACvvTqH+9jK>~dCZ^p?c1lOp^@Dhk&qx3)08?nIk|^q)@vTDkQx$l1l<mV
z+!f3uT{V^nNJFt-zQs7K87C*i&ws_2fME_$32|K*zD$`LG?tFZ{!(fX5gTh#Vs_(3
z;)f3({I?N!f<jTgtz6F2{6VHn54f}d3(T}Q>0eM#umYV#GH9T-VgGmpyjk9U{5CL?
zMwJiq!8?`$c7<uEC2i2vZoAiyKsZ#BO3$w04lN~oAJxNDG}5tRR(cS0v#zS#_ocrx
zVst=W)oym=P``~o5%UZRb!!I$;wwSj-nZ*E^6JNb2$3F9s0ii%7Y@)Y1XD4}Z=d2r
zjMtpu6m34G+adX_%(Y0&sa|_4gFPrFe<Y{hmdYgvH{o=bXistp@@JZ86)V+i?M^sr
zM+a1^-zVf2KdS}qPkFW+96-`mOjv2(KeZ4%-nkp~+K>A-Tp+dneDc>ng~SZ%6;}0p
z@#Q5eyxI>QP-<Mtl5^eB4o5lG<WPxkf6qkYuQhe5XZTm!_^=qA3!fsKGCA#cKHA6>
zJoM(nAjRVnYOe$`bAIK*qKe0_x(M$kwo`2TmX?ADX^DTVxX5v8+h-j+A;>SVhiHEH
zNkh*?A?Z|G<LhN5AJ5n{UYojB&X@EnKTdNhtB^dF#Nc=TFJ#)gEp{buZU#P*l*w&B
zS3S1Fj#|>b<-`9zK}KBo<W=KR+2c1F0=h!8B7MtFYhjbe*$R`r=b2ZPBg1w*$-Q$;
zXP#w`9qL{x<as|~!0{^(TSBK!T}GqPsoh6ToJfP-%o%a<_D;D&zJV`ahKGlzWM!ST
zw6YSGkSOe<O!oy@SgG~)erY|iKSrUan*FFqv8`K{wN#Gwt?ZoH9<S@y&)Y}OW)-HS
zh&)cZJim2SZJIN7hfsQGZ#aYUll@N;;JWv3-n#V>EG)Lb{rf6Baygj;z!_oHF@52n
ziqh6fsQcVNdYQ|FEtn@$yB=>?p7dW@URGhw8SCFaeV=yU!XmlSEccgjv9|ABTHvGN
z73rS^6g7wZ3)-%(=Xc34?#oWuN06%6kGiJ|)2|!_CMrEJ@N@#VA3GRemSGP0BDq0~
zU~<l=sCD6Ib(1J#$fhsu*$}aO<?DusR<+K72M-=V3Y+dL&b-7OTxkY&)y<6;eXut9
z&<)xhE95~^=UUAU)LxgL%l%BK&$_hL*^brO{$leQ=HA(#UDaJxSZ_G{HPEF_qllqj
zzkYr5)pl%FR#x^@LdTSA#Iv^MIQf8MCV04yk&&_c7ukn9V8QK%1@{NKuf!n+%s5Bf
zf0lNF%{L8LU6kq7TYGoysBdawqDK5&B7YZ*W?bbPa4dz1kaL|pq<#OKDWtW6xanaH
z6e@NHpc>^6UC4Xx-uEsR^xEIyXjaklTu8m)eW$RXz~bq}hoMbYA_tORym+Da)z;yw
z{a=6mwG(?9GA%O#h)$hRdZ=T%$FaV?Ufb8#w|2s<%mn7ccCfs-++*IiJ<rJW9duqC
zq;blCPjir1GBV7+`wYZagZB=AkfRl`<;k~m$G6r<xxikoI`kE_fnp^5s;|;{Oid&H
zlA&q2Ya93gI@3EAr}SZ;Oa?F*DK(ZP(!7DYyL%iXa_x|oi38NilqWK7Q|<i8!$xoe
z<$YDy3agg*>{qW2ntIO`r@3b7jJa1x-D~}WGTBz$2|n9y6<8*n^+8<npuGz)abC>p
zn+3=84G_@7A|g^BKi*;Lv-mym_3JLnNYhkc>-T^}sG-p}jf|pU+CRW=YOM;VpmCu`
zs>0fF`5riMfPCq`j+NE85X%r?`81<JaZb@EGGk+7Q{ypC`nI+iQ0R1T-MW=Ffg|f&
zz3Pu7VTPT1?tOW@`=I@9)ICXv!;jBGn0g@**t)YZ)t5>c1S%}Lr1Y0M5rODs3`WGq
z=YU?AnV6`2`0(Mt=g(U~$&hlJ)<yhM8>69Gt~Fd9ur9T^<lS|_#qGK>d9xUHBL`nv
zdJZP18=34NQL=JzaV-^>4LW*WiZA#Q^Jko58jRQ#ICna5KqJ`5a?7;3F-k(}&evUA
z)D~R#D>jKnQAX75!^rq{(<84er37pbSm_+OCKS;uySN&xw0#RmZ>LCZ<p-#r8VQP~
zFq3Zv9f@FzPR}u|(Ep@S3VVT&i1krSfbA}6&<?VJMw6PB77;?c>%)};&`{9;nx0N6
zOn0v^eZt7XY%?qnT*DjL;FiIo6&m*T?R${Q$iUv2GZ#eAP#fX0pMbiiK(nvoz(Lf#
zV=%nDy(<1o*xmJ_%{TQQl!(792s4dMYW!tydBU2!#Z#J5-D{eGb-Husy|c5kq%AGw
zC1^W;djUGnC;$houdQu;ewa5bB;*71AMSxK9@wYBv$cnKGxQhZ`>=uNJvgZV30$tL
ztLyvpAjw`j=J+|IWWo4{AG(Rd?iG}|iurD;L|!z+$0S&S_VwRqgLDZ^3K1zOJrMdN
zDmYPoyP+^79=E?K-8(QIlt`A@tQTr#+C%)nX?cA26FaM$B?W;6VS2GiS<e0SrMn}0
zmv(C^KJ18lB%kRJ92l5!epd6(KOe_qR}F!=s~1|gMBF*47yt9zAfagD0d!>8LBZ->
z1)=6%dop^7>Ju$)8HgPHnp?qWd7{O@!ong>kyrQF-d&@v{%g2t-qWW?*Jw?mJK-7s
z+`?Qz?LGR|qgdPZr4ISMFtWx|ThbebHJAf<ABx;p&z19(Pgsfg87J@IFR7s!z|6&m
z1`s&inI5`2Pr}^sc6TqCaj*FG>sLl4abg)BwFV-%uYlVt1HJ)3&THkPnJ?9{DOOeo
zJdpdSMt~M<0ZjNAoj<4o81Gc@MH65A$B$=5m#?a-HjpV9ppu<p88`-63k*}Ouw*j^
zOYT(fX^N9Kh>>v;wS)DMU{vauy}tf1&aK;UW2UeL$!gt4EA5D)Ol;!Geam8o1%K(J
zUVqD%jUkxX5p%+;f3`{yO_CV7ANrvNu0=O=>VHi7aN4H9yZw<<$~X7+#4k{Gd-b3i
zoks1<+mmo=jGA%@b3cllxl*oeoJ08P8#%hB^T(mRQ(=J3><^`a#{Tq}Nwvd5Dzpm8
z&=BZ?JSSz}a~`VV^lHBo5t@INg}t4r8XQbKik0?%ksvv3ZLblCU^C0)6j4&bfWD=6
z?b;hYQ*e#II(!W$NCO3dTrZMqKhsc{bw8?azrgyo_Y_NLazc=rKR}vxodYSdxX|Vq
z8+SKU>ut!fUlUHHUp#a$sch^|I~(ODiaNDQ?No`$n<fhS%KK*q=#UP9?{fUXI)fE)
zIYHTd&pHoqoH%i!{oMDTX9WbVVz88O2w_Fj2j5xM)YRG`kx9W`umY?ZmZFDb<i*9s
z>DKp8xUnlYDf1*keLg7AX;7sj@bMQPd=Suz&w|Z{I$#9r4AoNXJwtrD@D2uj{@~Pm
z#$;aAWaYAn+?VRfjHzm3Sz)J>G&g-w_rAiY*<*ne+@~!Ng)nNEZQWuS>9t)7MP<Hq
zxL;HotuuY6zF$l(dHdxL4)UA3v0qAh)Ef4aYa&D|TxaG6(`X9mm(Oi<&NFs=;}aCL
zgGC`Yz|G2<0!}kqXaNf<AliyZWf0^`Xbnn%N75_(Xz>YV9b@#Dgu{vnT91^K{qN}B
z(kJT2I7rR9EL<xunpZx5aL<mAgF?4=VIr}+Gn^#U)|htDl?|#iYJv><%uJ1^@^?H^
ztC7w~pXi}wbl%$T8MD)ol8R6&IW;zZ&6|)uj`{kgpdi__B6)eEo6^4@^_89Br1EPh
z0@F|@^z|bO3JX<~(g)qohRG#AtsP+~_IaLNGTUQUs^a{VSef>-d15&|G4b<^xMn>p
zBhZ*BNBpuDwi_F-7elt}YY{g|EkRaONN6Z|TjSBj*!^o$Q&t8H2*!P7Rf91}EW+{#
z<<}f4RNr?1*cUivn|I1sT3V*(<((g0j$j_@%rTXxKwN&E{ldsK<(0G$8oKu5s%*U%
zrtKnFm-z9EKRE6T9yJEWf$SNyifl_h$RVB8w+fM1KjoU?M5=<@2^n;ZY_iI9HzYa;
z_E|gKd^=Rv$gD`SEkwK17Tx;uE0kv3(Ivw?qiX%LB^eIVT*eV4kxCKS1Ku;NErNP-
zPGv5-j+xrQ`1bsUdE@P9-QxJKJEmAX7`>m0h=@j%yJb$n>_wm5aXvb{2+uRQHL-0_
z#zwOJ_Nrynfw;oBv_SQ7W6nuwC_gxKjE6@omzXK-(DzjGzRuK=#S<BYwQ4zhKDjmB
zSMsAZbVpgpvgQ)v;-iqP1C%y@hE)d}(j?uvAT=>fMcGQEmIMX|6M;u{=9||Xs^ZWb
zEa;TP7gm(_`<=VHa&l@cTpL}tSn_<Y&F~*WWiLJ57?gJ&Ady`C4y^|+7jK16^-)$Y
z)t&Dg`PF}{<HM6)OSJzf4Ej;vvP5{sSy(lb_&iTf&r=D~r%(TBmTQo0RIE+RF;$6I
zOp|3AdfG=O5eSe*bJs{X4D>z4D`NK_tZqdR5<m@8_${G#Op;BI<5^UC_F2cTPRS=r
zi+s{`Z=I2iYtPLI@%4`R4WUn?lIrBaJnrB*hYNS%yON(RXE*%m3dNmPcxRWc_M7Zq
zo7Y)SSa2CzWn5Ge`E=Hyp+y&cIU=w=JF-e}SZSL`39$2oCh#p!yqVkFfGpsu@>fuL
zC<D35LIKt9Phfm*f|EV`Y7l4o<tKZh6q!1R7+m(`<h^Xe!m9vIqGg=o(<zdI*BvY$
zFb7DE6Wd?QpxM~kZO#RgBA8H3CmBv^=%a7ulw?O`48}axafp>2)uiB8#R8wdI@ez_
zFCu&(=}P!x8=L#*at8E8ha-+XRYc%uWW#KD!>nGT4Oc^Vf63D1<9Mln-u3nMyecNv
zrhY@~xD?pmrVG}+u4pMcUG44FTJufJ3G@}Wli|kNX6Sf#`pWC}{Xy44<i5M#$c?a8
z3z>0<D=5xc#}izqw3)c%bODXGLtDF<$9?*+j`3yf1?sT=K4e+F_9qpz@p1K*>Tyh%
z=6tA33Va@9kg16ck1JPo+QQy?cAG4=54#``AYznrFL)r{-I+gL$iVgd=kgB<mUl{A
z6;Agh)tWmTPI1V2H~a^0_zF5?j@^vMfJO2wge<7p<fr`ytdoTo%)4%L25JNh5yP@s
zrlw0Gr&uNigwTPX`jVbE2rC+wLxWgHjzu#lJsXA6S3zA3E!yg06RwvRMhgqob_dh3
zY4$N9aPYn0U;VBI7G<}MO#PVrjkS8kPxtW#6mGm#yNUB1pGu*V74yfrs%xj_&Oc18
zgooq%MzMTDW4CT;yhW!b8vO!_Xd|Ekzr4$E5y@d9y_K<r<a~CWJ<w@T4=)yz(INK9
z(7ombBlhycNw(0e_zR6R8#hgP`aaux74Ea`pzb${jwA!<1M=2u35x3DHs)5o@5O4}
zDY*60oOrytsjqvIbI?pj#Dw5vR}{W@KJ<j;u1%}Ry{J?UDci7e_gP_r(j}Y#Gzq4G
zg#j6S-J61(2=z*!Ddu779{TuhQw8rx&3UD(z3JXX)F}>1{j0(vx{^qfC=2R3C?X<E
zC(kQ`1taeuZ>6IQ%dQ>e6Y<`Xjsq9HK&W`_PitvC`>=1d^-AekqO8HCL@x4nGL|Q|
zl?8*(9$i+AGG#}-vjvRK4*M6Om!E;z{B^>)VlcIP86<ikF)=M6At5>1Orey^_;npJ
z1M1cRhLcT!X|rtQlM$A36M?j`SFbJr>N{_0ViF5LFb(=O)8JCmzkK;oI%RZxTv$|8
z6O1$M(9trtJ%>V(Poa_wmx>gsXO5u$u<x;#HihjZ<<Pln6a{5TXYt_w<6`vvXj4#4
zP=DS&*^cv2e;FUa$vlPIh%vqSW>C~kGh$=qt^G<qYOfX`Y>+-qiy1^q=?IHP_->#B
zU%d5EEX>@$L`{w~P}5T4Z`v;G6#g<3h)Ud#a71OD#ZuK|#R+thQ<%MUVP^VJYH12v
zN6It$UJ^UQt>bHHa9zpWHa6|s*q>a!{5G}cg+w*`C@H<+C}q~auQxo*kX{o7@2XQK
zg|HQ*8R36F`JuhL81c6IW1DZ|QHQJmM<9RxmhUA6mAJzFF`S!yo0?s-gmTGGvvD{B
zd)%nlm(_$If8Xn<d#6wCX8$Y%wuu-pWqgF7&<yD$5XzQs#=|z1a<GU?`ffgyjoW-K
z(O^fadOxvU@h54FtDU-U!nVHbXgFu2g}b4jR3)u>G5xKX*G+aG^!Hgs%mUfT1$bRb
zZSCy}lE*Kha9cS6l;q(}_q~bDu$?3ac3s^9%-;R!!AMwGn1$`6<Utki#dkvkx5%oA
z9lyOnb}lnX?K0O}!!6};(|_4e4-)unoVtCp)jah$a&Ki;?VEUK!qwWy<`24*&=7AN
z(ZI_?CvUCfyN%vjQ%hvTxwno~hb^w7sIQy=IpUI2Q*FUH+6R>sV6ua|tfid$%tzRa
zG6}k67kCTYKvs=04N38Nup-h#`^B^WWDO%>aktO@o|%oNade+ASE=R>rs>u-ZKI48
z@`Eqb$D~0%e5oP(1Gg@$6T8Sc%xP1%;!ffBzCrs?cY;s{6E%@QM_R2}U!6Q*Q&KJ>
z-R!up01$uo?%jxDU}crsUPVf4<iTkpgh#y~TrVv9A5=}rCLz6ipV5x1${6c8y{kWO
z7QgyH+<Wi5Q`KP;W;Ox-k^?z}^*<?hDC7C<eO>J&?ziKc#s9pvwvJAC*s-b>SLjXN
zOixd@iadOL$0I+$Tj5brhR5%!M+zF{4(LO66DvL`K0S8vb9#BYhTUM#1|L5D&Sdsu
z``6TAtY!SY0A*#PLbbrx;URLnCk}9*zp-&&J?}&*HXx8zVbYG*NeSS7Q`M)ZRC7t-
zt5t{jVEJVCmz~o-c&j{_XR;9tu}5~C_meApwkWp$r-|`LfJ&O|CU*M^LEsU8)%rLM
za1)UnUNdsDCJ3q-s62N)Dvn`}@CJ)i%`|YZ<*q-!xb-F><0glo?x)>GAv4ky<=;HC
zGe1U9sBI`WtI*MVca;Qf4_agB2Eu0#=Fm5L{?JR=X|P4su-QDPN0L~-o8B<hp1{3-
z&%nk-t1_>cKI?;tR9@6h<(=XORlttvwjWT_*LIAe${|)(ekek)iz6KL$FUdP4W-||
zeT!WP-up~Kn{_>ZgZrnVoH2ncWkz4Oec|(HSgcO7LPxrHrE$3CPuPQW>fu(`2svEU
z#n-b~at_CDoWgeALE2{Hpv~7?Q#Cx(6Yo%me!%*i9<<BFr$S&<^*JQqw##70cRH(x
z{yk9gSxXQ~2mmdTFWaO{J5*3hKT_1R$i5dg)3{bz-vxpRB|YNw>C@~p>}ph>D&a?x
zo&5njA1OBpW_%l2tQ}eWh<qcQe^))3J6QE?N=Nj&+m7@oIi;VM&TY>hkPBOTvcbR7
z*10^VFQav$!*)kB{@2~9Y#IIvQ?F%`HZ>ubz$P}kb+xZhRW3=jYt3EOlwI!n##%qc
z<2q`u4kROrBWyT<be+?R@J~eOQ1IG?yV;j-9sr0l8NLrh-;)laJ~HT(<m5Bw&S|Ll
zA8do)SQPfI%$AZ6#9Yiw0J;o-H7Y8qzM<g&bpzqw6t6C5QpZ8>T51om-T|aV=Zl=D
z5@bwzEIKG@m~^rhp?yXhMP9d$0clrVEcc`?|5S$EW0&n5@2zil+<qyvr)qpL?<8Je
z_~YPVKL*K9v9X-5Ic2pGMmMuMobCMYDiR%-SsUzed|z98%?(xv-iy)>+dh!%D=gLW
z8i9c}yEz5~mv`^py&=DT@!~C!te2)4XNzsPPc4BSh85+YxSznV^@(fq_UXR;V}$*_
z;a(%N8MJTT)nkS-X%`q0UqB1>-hGdO2LZ4%R{Kup<bNs>84!H8koFIdVw0(<slGCo
zG^3(>?}6difJ4IaRJnj=nhwCt2iIB%KWb`Z>a)90hcZB)3)fwX6-6a>LsVrTs%3pC
zBLOl^`BHtf(}O{f1i+=4^y<~Cf|9Ra{~Q|~y$8G7z_y%*NJ)r@15FmtA8vvdQ}XV&
zz5IN9XJlm!VgE>2lU2uuD=$}I8*k?ZJ1DzDAg{`qDxsdnAkfm2PW?6`Zu|ZCg3Oev
zEJ_ly>$ar)Zo>e|R;1I%{%0diqTh8vjO-dKBet(mx3-URlcrp|g;6c>3dZ31)7I71
zH8h5Qp4oUFIdWtLw2BOHSc6_v0+B80K6C3emwbtR!hfpX$hdtLMBO`EI^>lZe+zFE
z<$stHFZ=5I%+^(fAHmt4s;GMwFd`oR_hG@9)^1jh^iO~2oa;+S(T{79smPE6@%-7d
zXUM{LaAT{in{us(&Jk>l)NyuRLMM1)X|uNRr!s*?tF|c7E89Cd>;_6fe+hZ~L?86R
z{r!ODkWg<Bt+)<;j*s92PKC`{(*T&e5zWxLIrdF1tDr}L=abjwbi3y+@spL8beOI|
z<6zpuO3T5FR-l%X?VG%H|F$=-@QAas>(6H=qD}XsbOB-(zDb}yQ8e+&tm-e*40?Vh
z4mZea8|4~rLpga>nSVFvqP5x&d{co+QO4e%J-3y=dZ#2gdBam*O8}$`So3jmZu@PT
zp?`8?kJH9BZQgER0xZzl5MiC5BJZDtjj@PgS`04uf)WtS)4@gqY%kY27K-6Luu_C1
zgWh5XG94_wa`;AtmCrj_4*8MWI5>t{pWtHCwW`*G`O;MyR`ooDrITB&-7|V^(eUjE
zIH1>msHljteJEQV&%n}bZFii#+qK*>HDO?y9p*s4sXwo{|I=5CMu^*~YL-!r>$;h|
zszr+lxkRtZJ*a!v;gLD`zjsnJ`Jkmggpjl1Ij{G7USqP>r3ba9@_aFG)zqr(#w?i}
z_;(u$i#A4NOx1af^Gy`>&sw^C0*|8N=-U7VZmk}sp(5KZVWerQ*9Tk3vmo8-Lr{0Z
z=EAV3sC0xx0{F-|@_BiAIUI!Ur%!jG+fyRVGT@svkSuHjF3D`i>Z`Mr>tFXb#VwuK
z8nMO^dzHvG?y{`wI~4Kn3iMB`@noG#@J)n3cxQU#b|5IAlc64@Hs~m>OqDnY3k!=^
z@S(Le9do3Ae3GIKDcu8KzOx*a^{Dd0Uuj39vnrEFQ`MHG(=@T6H`T1!bT-OIFXk=0
zy>x}zeiz|g?H|lf72AGn(tY5opJJK1$^YyWapeBh`mt5vz7GBP66;#{A5fXd{>*Eo
z251L`4NJn&;H#V*9Q|32%K)cx@vG5u@GVKg{de!)UFkBwOo5-T1Ga3uPfq63h?CPB
zsWHGhdV^^=8onN>myC3o!0}ZMc2X(l(Q68@gO<sQOaG>>ZnDzQvZ@37YT%K<pC?4)
z+|wNAq_wHrJDQ@m-yByorhRGk8~<TdXanvecFB8hdQ)^4CVC5-L>W-Fe*tH5y-gGG
zTH{G#T{zJ&wpu^ykTyt<&*dUpWq$fY{p#?yCo*x@{?d&#am)p#7An%mYfxk$S}OC&
z_iW?Uay;wf?974o!-w74N7jHVo-`whIC|Mj_oKZ9K=UaE@tZvu*;l(i7b=J(Zri6R
zL5Ut__wB*&%33<j8u7EVlB3)t>E-s7TqQ78bAW`B>{N1A|1;(+TDZ9-Ecd+3$-f5A
z9qSjSK0JSpM_Xif_i(#NWPj50j}Loit`iRh!q<}sU$}4+T!Iz7DT&;pqoW_6HRvn{
z3B2_9E+nQznh^((fAk7+uWjcVfQT{{tyS|23pKu#5Z1S5u25G97~(f@U8f%*amHd-
zdj+|Kdh7uJ0fVxNe6*)EmYWqI`|FoKsjYlv17;M$;)B;dqkbDEvO4~UTdi}OTc?HM
zC{!hD$YTvNM`G|`L{<gHwSO1;-HXFIrS7vvreJIYY3gj{_H4|OsHXnB-V@>LxT3Lz
z8u;nI^>&?6O=Vj+Sb!yBnX#c#ltigPDGCZW7Qle0gGi)CQ6MBJ43eNhNYoJ=0ns32
zlp+WSf*=AR5FkMj7;5MdAyf+`peQ|*x3BYN&0Fiezw_=&eq`Nq&$;K^bI;j(f8V#$
zw)kApr7hm>9`<wJg{vEP4oA%oXPCGYrIb`p;&f`&$ctL@p*fQT@4kA#;7=~fmJ*ws
zMmOx6s+KYE9pwyN{&1{n*LXWwk4>}m2ps!JsKdUnhY;ttmQ2UJBG?%yk1iJ^ZVFCB
z7VP?1qto%q<|3!Y!B0b5g*f&0?x7N8Q!VmiE+yJ>iD_LC<g#2Rj&d%d$HC92W!Gq4
zpH8(Z3H-f^x@8AK!(LRs<819pXPiqz_eARSkl%*FO+~-aZXj;3yy^4eT;zCb$G1Fk
zf09PE_1>cA?%u#gqkVZL)0$+l8na`^Bh!&`P)?&$o1cN$;tVWzC&9F|&P*-L?hw=5
zza(*|Co~Oak`!hIY;W?^iGVzlz*H>pG7tHptL`OMQcY8uw>u6nCzrC_yAMR!-OFlg
z%uL_feD3nV6YA2u!4Pe3a-a{oV?t)nP#<;UP<Kcdxn&r*0d2_o-RB$7`l<Tikdc>b
z-TwXug-%!#$V7)&ZD5jrm<lnSr3O}4RsC#Z!vuKXSZ0#ZKK?}E=C*U$T?;|Ernru*
z#g4VH!NlGKi|Wbt7&f8%gHn^SATI6Ih%~W|M;~gP-!k42&(HET=JyJ9_OEJfE@BZU
zXHI>Pn;u_?P#*N&SPS8L6hj)KBDwE<1M0A^Y+{e6zT<d%Qx%A3<)rIRN3dpxD>(Fw
zvKdeAj+qs4y(13!^S8WRqS4OuSy{?mD<;~6`>l$!G#l;QFnLYHnK<}i=Ig|H7gBbn
zHD8(sdVsTT#*jleE$Z&ek%=Ogev07%Yv#C6PD!KV<K8)1Uya$<f<8(xC*#T1``qGY
zAH~{v7nlil0mY;7%@`cGG_9!?rl$`1X=2JL#4m^dzRe6lBj073RN|PkoWa!m(Kfe{
zjtH@Bybi&>yq)y2Ahb9%lD$xBj%&Je-lQz5-29E2dj8G)x#^c(vzH>x{TAHX&pJR&
z@(5=w%^=le`h8f<XnMnfn9hpIWZhljOzLHYuG^LV7)w6#i9i}zpp)jWhi`rt@0kH_
za2Yae$8m}LqnbQ>)sVWxr2bP6{A>K(2qlh_Hpg*Urh@65cT+B%3y0=+j_ARNP%dTU
zZYlJI(J%nZ7~JT&GMN>dq{0WOaoZKw2LyYM7}+uNrujOylCH}R@fdYOZ5zq&Z-TW!
zI&wxS>8GVAlC}{Rca?JupE$BnF~rtW(_l1gi#~ic@>0SjTrZPG(bLxfjz?%Gu4qBr
z;gs3x{|J15)Z47al;BBKV7qP#ZpT1@9e1GX4LU(%a&b{tp0>NSs3m<#GyY}FodpQf
zBQ*r={`8p6TOl@^-9dibo)X)l*R9nYg-0p@fciIN2;Tecn|R-s>L-#cX2{FzO)te0
zsW`Wl3YMBjg#_M}V=p}Bt(ui$FHv#G6|U=uC<fYdS`b2>*rGEmm|Db5@h0v8a8irK
zG5yr$Wy}hn_z9f-hf8F-@Yjqv&}#;F8?=L{q%1OoC0<C=DES+X3I7;p1}HHM_WSwq
zk_Acqk9Ul;aWAzaYq1tg>#>VmbshzuD~B}QPzo#GI^GlGzyk3ex2<oV)++mJYG4n>
z)t9^z=Umtnq&R;m1okQHqwHgi@%ip7gb0;zeGQ{gt&4uCxcFiKvEf%tPW$W6^^y{0
zY-2~nq9yBu<S2FjA_q#M9wmLT+D@eE(mOL_RV99-w?kg*&RN<BrOvMxax-|{g<as?
zz@dNd%Q01=o$e~HQ?3zK=$JaO&GB^Sh;7x=)EFd!h(Zxnz*s6^@Q;9nm|K%}Uv;~s
z6#ciq&mm<Pm@`3n%8v!9fR+z~3CI)<w?ei{eV?A6a%*xP&=&@;L4PdR9ayg0L#99(
zgL9yGTdcp3N^0q*pS1PL0Klst<JL8^tK8m~s=j6c<3l$;sSQ_(*5B<93;WqMV%reC
zzuz~Y!=SSqX;0QMdUYv4tKiSKg6{q_7!H7Ys%vOiLaq$GBK;cK>A;=E6K{7dshpCt
z<r7K?Hi@D%cRm<;if?&#AyZy$1N=p`BGo7H=`v7XgYB7pC=U;hVUd4#E09pMLdv#r
zIx;~a<2l%rvh3`SonPMGUtv;|I#h90;ZB&5wT+KbA6{9EjO_!;G?`7L+m*y2HT~3l
zcF=<X^y&Y>gWF-}ZaraQ#;hAP<c<17%fDH#*!|Q+fH1uhJhS~)`v7Nw$z*EeK`7uT
z6T%mOnP_^&>^>h1egOstyEsXB*8?})9O}=>`r(FmkS>85AmUl$k9n!b)L~V7Sm&03
z<4bw<`>i{7tkNQUk%^|iJ9<?q^Sz(lob1Xn!(QHgsu-AaA=NJURexBw>jYTc<nFC~
ziQiyZvv~Sjka?-jVamYO1F!h@ANm!{6Dp*#u>%@)#e2qoAgNZL{$>g__LlCRWBIP!
z3h$Dv-X3qcTq{F2Xyd=RcN!A@c$1Zz)C_En4WP8r_Zw(1f5uKMH(Sof^u4J<p_U&<
z{v%-XjY{oyFN4qcX1&t-+dcf>_ys*zlFt1>uu*%~r!BB4^kB#z_LP`5&GMbnrZ&tw
z?A89mwAaXWd&&{-IyUUs`X2cRO{-q4%b<C{@b9M#B}1%lg+QWYeB~sNy`wW9z1g6s
zsLknsPbOg9%a>f&kR*(fXwn4~p&SM$F?IR5Ejs#Z;!4$o+i8YoM=I=$T|8Ct+aENf
z@X^nbcUfF#yim$O7<e#OxH515=Ktx_#))n)bs$$=8%&>zRL^KzkzQ%lr-&v9aYv#j
zW@k(++qq7Uvu2Wbb3({^tN22kSHeu@;TLC;E3<q$BRfB?Lgv6sj^FBRv~|oLYkPb9
z%@B=Wn65nV7iaJs)js;dvN^ZQv(bT3!<iQ!`ekL&2IMHC0daG@Lm!k(()Ng?J>PyN
z2p=A%6Z*%_J5rg+ISmM*y3|2Sv{Va+Wm!Rrofw2=m}~X&`C7dL852CjRkaG@%qB~m
zOvrflh)^)U=Rpo!gwofG{w!r)risM@ZWRCL`<`vMWY^|S@a^*Y>n!T9N#@rXFa@=P
z{~Q?fT1I2f0>~r=YZve_&gS7?o2eOOR{i%4;m^gZ2h!oCy$*X3l|QvK*UBg{R#ko9
z_zcO}U&MB2!IOkiXM)yf=^KMdC2T8BCb;|({w_8!+yTA%o~T#1NDCL-WMj(I)U}j3
z;E<dmSUe0IGW{Y%1&5OV6%}Uw3naQB3+qp?ug%EK3MTiS?)%m{_f-0$rg~!q#cG^o
z6hRdCE356^-6&=v7wM}&kHRbakpRsls>1r`PGiN$?gr}~8nHUy`2Wz+eYC--1*N-?
z)C~%>VbGSzJFP{USFv5z<4^!-RSU<P5D1ZDgV6wlJOltLdsQt`f>Ci0m36DPECeAc
z0qeqOFbH@WZ9t`IExhbO0TS;0^zU@i7HO%14It6&P6M$9ZLEAF8<HH5rp6@m;Fmo?
zdxC=oj(E8JcBIw>X=vBSa4QY+@uER(V{N0dCM`z*p_)7()P-+0umstJm-vV88>xW`
zO4x*9IX!_SM&_X6YK9~y8eTQEwJosmt_N-ai09rUulQbe=V8>US0H!7o0*%FAj$|q
zIZg6oFj`@iGjOY3hrwVTp4o=nLeCM|(rp7xDWM3@0c4gYkXDO%qlAx8k$_;A068B8
z@Bdw?tKQ18T}8K_qCu_=!-%KRxL_B+Dv%81rlInUgccZ51fqh&XUVdjeP_6S&noUJ
zI!?~MmeXm0ltI84c6U9EoM%LCLkSP6tZ4v<C`pGuXTuzaw;Ks2?!Yfc3K%NHOJyxB
z5@kX5v}3@Hg6E;wi|E!+k5DwmiC6F}5FpGb?s%@%_dd5U{*pFay8e&Rc;^mm4sx@t
UU`_+5g+k%XjvvW7boR!-065hrQvd(}

diff --git a/docs/articles/vignette-02-grid-model.html b/docs/articles/vignette-02-grid-model.html
index 8d7517ca6..0b522b379 100644
--- a/docs/articles/vignette-02-grid-model.html
+++ b/docs/articles/vignette-02-grid-model.html
@@ -33,7 +33,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -279,20 +279,20 @@ <h2 id="simple-two-dimensional-grid-model">Simple two-dimensional grid model<a c
 <span><span class="co">#&gt; ╟───────────────┼───────╢</span></span>
 <span><span class="co">#&gt; ║Sample Nodes   │   5000║</span></span>
 <span><span class="co">#&gt; ╟───────────────┼───────╢</span></span>
-<span><span class="co">#&gt; ║Total Size     │1.3 MiB║</span></span>
+<span><span class="co">#&gt; ║Total Size     │1.4 MiB║</span></span>
 <span><span class="co">#&gt; ╚═══════════════╧═══════╝</span></span>
 <span><span class="co">#&gt; ╔═══════════╤════╤═════════╤════════════╗</span></span>
 <span><span class="co">#&gt; ║Table      │Rows│Size     │Has Metadata║</span></span>
 <span><span class="co">#&gt; ╠═══════════╪════╪═════════╪════════════╣</span></span>
-<span><span class="co">#&gt; ║Edges      │9065│283.3 KiB│          No║</span></span>
+<span><span class="co">#&gt; ║Edges      │9118│284.9 KiB│          No║</span></span>
 <span><span class="co">#&gt; ╟───────────┼────┼─────────┼────────────╢</span></span>
-<span><span class="co">#&gt; ║Individuals│6316│618.4 KiB│         Yes║</span></span>
+<span><span class="co">#&gt; ║Individuals│6357│622.4 KiB│         Yes║</span></span>
 <span><span class="co">#&gt; ╟───────────┼────┼─────────┼────────────╢</span></span>
 <span><span class="co">#&gt; ║Migrations │   0│  8 Bytes│          No║</span></span>
 <span><span class="co">#&gt; ╟───────────┼────┼─────────┼────────────╢</span></span>
 <span><span class="co">#&gt; ║Mutations  │   0│  1.2 KiB│          No║</span></span>
 <span><span class="co">#&gt; ╟───────────┼────┼─────────┼────────────╢</span></span>
-<span><span class="co">#&gt; ║Nodes      │9090│338.0 KiB│         Yes║</span></span>
+<span><span class="co">#&gt; ║Nodes      │9143│340.0 KiB│         Yes║</span></span>
 <span><span class="co">#&gt; ╟───────────┼────┼─────────┼────────────╢</span></span>
 <span><span class="co">#&gt; ║Populations│  25│  5.7 KiB│         Yes║</span></span>
 <span><span class="co">#&gt; ╟───────────┼────┼─────────┼────────────╢</span></span>
@@ -382,7 +382,7 @@ <h2 id="more-customized-spatial-model">More customized spatial model<a class="an
 <span></span>
 <span><span class="va">map</span> <span class="op">&lt;-</span> <span class="fu"><a href="../reference/world.html">world</a></span><span class="op">(</span>xrange <span class="op">=</span> <span class="va">xrange</span>, yrange <span class="op">=</span> <span class="va">yrange</span>, crs <span class="op">=</span> <span class="st">"EPSG:31970"</span><span class="op">)</span></span></code></pre></div>
 <pre><code><span><span class="co">#&gt; Reading layer `ne_110m_land' from data source </span></span>
-<span><span class="co">#&gt;   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpI0CWYj/naturalearth/ne_110m_land.shp' </span></span>
+<span><span class="co">#&gt;   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpmEC5ek/naturalearth/ne_110m_land.shp' </span></span>
 <span><span class="co">#&gt;   using driver `ESRI Shapefile'</span></span>
 <span><span class="co">#&gt; Simple feature collection with 127 features and 3 fields</span></span>
 <span><span class="co">#&gt; Geometry type: POLYGON</span></span>
diff --git a/docs/articles/vignette-02-grid-model_files/figure-html/model_america-1.png b/docs/articles/vignette-02-grid-model_files/figure-html/model_america-1.png
index 422092a921c9106428f287744bfd91c306f95ae5..3c897041f6636f6fe3fbc29f06360dc873e2440c 100644
GIT binary patch
literal 75951
zcmdpeRZtvJ8zt@%B)EhSEV#qq5G+^-?hb>ydk7Fbf#B{g!QC~uTX1(Bd|;dZuicle
zecF9ms;0VY=()H1$|vVLr^CJ|N@JptpuxbvV9LsT`U(RBHvt0!yN-ecd=uHHT?QPG
zjOC<10Y~8PUZTYxIG{SnXgR~c;JUniU{}p4Jz!wo!N`6RQ}swYUUB!qpIz#_SWB?+
zqDDYz`>yy7wGA&{=4~E2%11N;ReVObKJTXa)e6GO05tj^F8F!D`UrR^>7p1kqVL-1
zS_IB0;pO0pP32;5?=O>;7Sb2_$4z-HGWngyh@*1o(Zo@xzP}Ow@pTdef_Xhdq56m?
zn-%<e@|_<3|L!cJ%5DWj!fQA3bbm!o9{!&;i$p&JvWQDbA+fWw=M@xyK=}XZRXI=w
zj`=^`e<J?BIol^E4*(vgB{43p=WjeCv524G$+vGr4rguASjF;O+L)Lr2pm?|;3p`Q
zoILuC`0wx!6XGb!vgrcu(MHh6>EaaLTW&iUMh!Y`DN1)2O*3GK3bBs!w_46FGw-j6
z`M^JOC2+-m;L5(<Zv4+h8}pGWqgKMHaIq*-5Gc*bJI~b_N$>$h;6+eorP=jAo#4wo
zh@w+{qH$U8ge@_tz#ZZgg|}y=u%3;iBxG@Mp$Bf@{p(2Q;D$#>7erMy<n~mRyc01P
z%i!DCXZ!vp=Tq!|exFB=28!yRG}4S$%<Z$`;Rj&=bM))=0ba-Ow^%&oComU@RxKEZ
zl-I6zU?3C~k9Hzg8e2M)=D#yr{4rD1|M`CJ?(VJ%Jo(<q$;soSZr;JsF<T-OD;i7T
zKkcZBt2LubTUmYF*w~Oy<HJc;3{6cXRVk1+8j7X)Ul)B<Yeod|@FWBTz+fN+?KCyJ
z9^sOcM-7>Z7fx$sXy6A&Mq&(arKOS8E`9X%qxxQy!*O|81X&cMMXWOg8=6Gx%dO9=
zSa@*MlE;11{<@SMI?q|-VKj{8C-TT^heQ{G#IspWp;vUe8$M}Mu;L!Euh+pMbTqdy
zn(jTso$Mj_;fF@E+o!_Gp9!hMN6WWWPo}`^%3Dpx&NsVu_pF;vZxM>4=#);OUeA_l
z6dQE-1@8S@S%Kvib8~jC5wzz>-k&a13beJJudux$3Vzc~a5-YGo+&(F0exVrl9iCE
zL&@7|rIZ;T-Wi&3qiJ6)qj!NCF%W*8F51~pl(M5(lcJ!xJalzaB*E-l%)@_HRNVQq
z{!URTvs?NR`{Ln7sP??H|J+<*{}MW$rtrCI+EBD`^86yOW5MtFeG4Ut^-&LSgQ+E#
z3T0ltUCs!S%4jQ!xswxh)*8c8MatCGH8m$VId8-lKCOj79GV?z%74Z1BBe+3x4baT
zTC?Wo!=Gz6L_a~dsdZD>-v(^W<DuDTtSv|UOnC+ez($e6I2<pzzOU$Y=3xl#3XE#J
zJl^URx+2m$L~z*uEa-S}pbiqJ`tA_<aRu(=>}(tuBy_>V<m51^+^*T#y3Q+5dNgbM
zUH7HO!BjqzLHj=zM>MMM%tWK+!!M)Z-zIHnsTk1_KOgrij}OODhGH4^6@x>VbfPj#
z<+7L6#&25{b7_yl;rt8p;Mj7s@P>V68KtnOXaqtJ8)iFz1;40zK(LR5N#byGtR-Kd
zgLjoF;-67(yPQ*O%=-@5!OP`Q(Nqiq$N|j$2AyT}GZ)wNYLxrrd#sZd7fw(AbhS9W
zh|oi^=v5Z$gl0~iNXPneV9?lH7Ok|$9s`+j|1n9TC^xAc5P&HuDaFObHS2ANxB8=~
zxw+##uXfp;cjcu+X}`bW*~^>#dK+PHxp~xc(E9<#-2F(kWL{rJ4n8X|k&q-RWG%-8
z<_mo*>Aw>Y!c6acU-Ay!cop-_d$014QsQJm;w3D53yC2C;D*u4J~xQMx85vS=jG+G
zxgBfjd+xj~P)z@PwL3~qo&ZdwE-;bFB6%XRi;FW?m)pb7MuPt;T#NMuN+eZiYwfzO
zF;h{YcI-f}(Xe%cBb>1bsXCiSm>_b0A&$UbqwqsKB<v|~S5Fs%vCmGlR(o*8Q9?_n
zm?vP6f&vew3psrsT%<$ae}ChaKiRpEFiNyst(9zWR5pFFvhL@@+nBl2B&I+@ut58m
zPmD;>P@-Sn=?qJ8%GB;}m^@4%{v*8NFWsvxeA_oL;QZjVsn@Z(V_<;aWam*s`y}Wy
zp?9W(LRA9#ub~2YoWCnoO#7A%J=3(9IJmPCCBkfLi|~P-ln=ZWs~Gyes%2OG{(I5g
zhd`j|&)}Qm#WaDcjtPc?v|4Qn+xdPRqxj)d#(A(nL1k_aFnqXb_hR2)d#|&c4k3w3
zJwM2%&QBxEHD2~*k5>Cbn9Su_ohyO!YI)IX--xSzjYwd%*ep^m(`=8ngXLxe%YpHf
z@!574>G)vMDL(%=SQEJA;49+xw=l1{yj{CsDOe5zeun>FOS*G%GP<OY%I9`Wq^zt=
z6xF|JV8~^x`yA){4c?`#dt#<S=a>iYkk|a-`oQI6>212AMd2V|h<*P1I2Eo(bG3GQ
z;rHpzPJ{dxi6`*QU6%WAYMCR>B7___KM6=ue}6Ma?fd`8kN&^$#{YXOZMpDo%nIqz
z!otG%1O#+^*D5M1_FX`wXAQOTw0N?zvW<<6nywurQAoVJyj&zvYPUKB#KgqgyStm)
z+orQn5I_IocFh;<@SLB&e}9jQi<|ZDU}0g|heVeajF@u*&8oi$!Zo(E2wY!j#?#Qz
z<<447T9q$LZ66=oKswf{zu*Ud|BjIAL$UJjU)l5H$B5Qe!PvMsM_1RnL%%w0X5fCX
zpgMALay>RYI?$F`LpDP;qO#`k%>|BI`mkK*xNaFZy)aWVvx{Bp8qI1`cE<~b{aO-|
z`HT9c=ePJ%HS*W{lj$)cF9}|dLoM4i8h-B^CZ!RN?+P7fFz>Q{#i-p@CgG1rF~<l#
z(Vw;Q4^+gDExB5I-qdM?g@KR)l(s_DS33B42`O3ag}rWtq@u_yXMG6=z+9JO$j<i<
z9c!^fgy;cEJp}5@9Xa$#)l-(s_2g^57dXt?n)YinO+NP=H5nOw)w9EFDUn<ylW*wS
zAi|vCi}D5R5y(<PoyafGFxc4Gr6-RsjFQTCm&4<C`?>yaNJ+6pdywYJrK@WcxU6P=
zj0?X|k_))i9=1osrKOETCYo0p-=lHFeRaIK=&Vb5Z4z#GL>CEzsMBBgA?<LsROj_!
zowiX_GQBfI8f;-Z3zbz$PYu|K9mO}A3NyB6wOAJ*AB-+GX<D!5$i~uyutam4=gNDW
zE~|@_bCF)2@8BP9PON8r_YzZ6hac_RhHtM39fXCk0tjHEY}8jz@bRadI1Z*;uqpi@
z46v}UZj3E_v7#Ed<~7MiL8sYaDj!HQ#`|YsvS5bPnFmZm>Tps~juw9jAxD47mBK7N
z?O@v*&n&diw9r<P*`7Se;=$LvZG^ex*tqR8z~kO!v(~Vym5b;dA5LIDDPPQM2Vs)T
zA=)^a{`A`5h-rs{hKh^jeeaA&e>)WZQ7Ndt2v3M~r|wcvP%ZPb3GdDrgt#2>{($Q6
zPSqB>cIcwV{r=r6v~p(T6*O@FG#N8#X)}HJ%XVb9<_Bvd>#=b6ZHPc^e%nV6Nc`R+
z&NI3F=;xk+sR~1ANBve|G}~pMAJYp28{~Ru3W2V%(9wmm++Keu$YpF2!9=1X=NFBx
zL+41*syk)!zd8In36JB(Uo7BtP2U#!Pw7(Stef#lH204zROU18%hN9TSZ{CDxro7X
z&UXQy=f%X47cXMg7)@yfgJqi3Jr-$>lV4#zHxS(r3}PDylrg;f`!ADXM=`3y_VEiR
zry4Kfe50=;K9p_2RTV;uq!)DU?d@1$)FKD)da`9&HAub>j^>!H$%4xvis(Jdsl;o-
zzpz)md)yFrt`13Sz&E0}m@Nzo9EGk<LD4Fl{H_N|fY&u|`DVvP5+w-TM9*u#O{4Lw
zPxgY%YZr-2T=hr?HtTW%A{n1UKsip~kjJ)v!uBF97YPcn_~l^Jh{wI_a(vu4-OIyw
z<vI#Orz=?=xeQp{YnE2Br}q%&ql+sgk>#?q7i?n_ll%KUL_uC5|4SRv@{}Gt?85Vr
zNNHQ03Cx$&7ji;Ut3mQ*vNRhHRGrE`zE%be38A~~h;bVYMd4|SwojMWyi~QwJmpVq
zzOtDRoYIDq?S(%k^^AKVGc&5Y&s!4u<SqRLKy1AgV5m>r7H*O0M<J<P*8XyS?hvb}
zyFH8*Bf#W|G+|jQI8{rj0_SzLf&Aj6Np>#_rzdOk;QKa6yKt$zRqW()q1iG>@9$Ex
zEn}2d(#cl>5|sVek^a$+#nz%YIwf&Y%%Tm8t7@2Yi1>6ZRVX%Py{9rMuPs%6kLJWe
zA(}EM(r_(9$c@4J>MYeWYXha_HzHfk7}Q3qqt9+fO>()xzEHw}Leb5t*M1Uz%-~1e
z@X%p+l=aZQZHL_4V%t%bSKP^ww4^d0*#)7?z8TQd#DUBiPe8dFSdiag9j2W0rgNn)
zBntf=;SO=r{G5h|XYTNEwb}@83$8gg+c<KbNT&l?hu6-@g(%>|M7)%j6^H&DxsEau
ztA@<*V6B{1zaX`jK(s7dqr_*c*wJBGB_(8R_teC=%+1r!+S)0=H;Ut;f6RujPF%c3
zc))rL@?KJ)RU}s6N!Cw}6|_rE)@n+agoMa75_MY=4t7hZIj3)2O|_*!9t-v(NEAqQ
zr`7B=i;+@$rVVipFuAF8L&f#>KmHC5c8)3zT`5ph2=jQZO{sjnU|+cPF5?qM!k(I9
zl=5XrKmC;bvA?MxN_8iw^luGFb<(6XMOqD__Lo?d<8mg`PEpeT{NsjEytbLo_~+7(
zlsdw156%>E_g_MdE4o_C&%56hogVCi|BffYdq1PIdv}f!YGZy=4?0<@G7?y03N)~5
zNh6se5%!s7zt9|R#x^>z4zBdJMn-V^V=t)wJ2thOvBjc>#azf$@bKJcuqM(Hv@~fH
zhP%DnjV)*NXb0(@4DdGEhbc1Tv{-2~nO{$iI(*THM6w@H@U`SN$76!BmOaB>dlB6t
zVE9#i{ty>d?CA3{P<kxTj;6#bg5>TO``9x6skFS6a&Vd<G0A(IP<Wp$IHkf0-B(+m
zVEim|;4(?igZQ6E!7Oydh7-<~#sixs@Y(;9@;i2TeO<Uv6Qd;|HR{0CYhm5y$rD~0
zyc<9CCgL|7V}BgA`W7|WQqY9mAGiJvCczi9cF#90dEh5u5|2uGv37(m)Ly})eFpW_
zmMg`Y{5AU;fr=Hah`*PaDQ5cn;U!Z3Zo}e7Sa7lu9}i0@8^igk;)Hb<xbTAZ=6vz!
z{BcSrM#LZfTG3&RXrRv&(Yz+iq8)>ZMj}0$os%y9o>cz>a6t84b9HbrE6j3WW(*Fk
z_aA(=F7(J>A(DezxxJm|b2!0=8EPUP0iBD~l3;Q!FW8g)T`57cb;-fUM@(sh(OXrG
z+bel5q_2>ngBc>N(Y60lG%juLZlkM?=S$I#6>=i7qAgrEZiCOm?Zd;QY{z_s+qpt0
z)dugX4FWZ{xY0ke8c&aLRxD~(2)QMzV3Q9_uMb&!>q~bD`!-50($JI+IS;(i1W^+B
zA+n5BIkTGT__p&QYdYdud3XCNkdcQ$@?Nj`LZ|7Aum0&<sldo%Xf8ElLvJnLSB}kt
zNpr5!)nSuK>kDuls2iKr%>`^$kd+0sW8+Q~9DF7XyHp_LsBm0gVrZTN%g+5)dlGV^
zf;p37aum%oqhp>>6IEyVJ{IzeJlmd>&z0CO;(Mr9H$3@lsTK)l+P&t=^NE4c#cAwq
zDN&8AP<>~87kqfrewRqU&7F}T*AY4<NMdu%!H%$&0&etFt>NJA4MIq<9?x&mD8@kF
za8}OKB}JaIniDQny6FBm3nt?)(r#E7e)y+Hm_ne=rmHGj%1E*-;ljy)MU1c^2X0sj
zKUCaUqiazj5sd3#gZ^PIbI!AcTN{@ZqaNxY^2D0}ZyaZiJ=0EAgMA^yr8R8A9K$sZ
zA~tw0g{G^^lAO-M+|CA`NcxBEcf(-Q_5OHufsh>C;jxcSdIjupA?R?PRrh>I++>3n
zy<261Ehz)bg8f0G1CH+xKEUAV2>#G7Eet6Y2_nHVJvcD?U`D4rny;Px7!@UFbSk(!
z&SA_(>N&NW|F7C&g}uT~vDTCSmq2brS~sHrGOa?Wx0iU~mmh5?Ew?@EesAVfpnS9+
z#M?k@jUs#pT|+AAmeiL;4;M`G`q5#S=(*<(4JU8ow(sP#e!yBTN;UcuV}LPM4#ng-
zz7KT&*k#KKTgv2TAZ_PIkdx|U9gtHqcl%87NZ6}+LWE|*lghU9c9J9`9Tgiv;zfC)
zj<PrH8~M@3^U#F<b`)!E?qNOE%KGTQy8Q3v(?6Rn4Q7)0v3FOJjFbvu2grC|!ujft
z|6%lZoDH{StL*=~^fRvf^yg}yBg8Ga3;1RN%ys95w7#QGUalUho-0|Ya^|;KdUBV8
zKFTT4t0wd3a=q^kBkO$JBZQrFjdFl3-_qdJej>Q2w?m7dCXTo~ggsx&fQ$E;A*N*I
zP3F=Sw?U?+g(vjl1LX6{%r3mcBr$khIY{w&lpolD#h%Gn={%KO9|9}hN7CL=qpeb)
z`f7J&88%vAnQj+_x%f4I)puvx%g1{epq|6N;rWcCamnnMw0_n$VGEi$f{F6(oEo5F
z)ltY>j{4g0xW)!4TqV}NX7$Tt*Rs&`I+4E*VwLjY*rqQJU+n~cOZV>{k<PM4`R*Eg
zF)pr-CCCOdWgmHtz2<(z{i1`!PPj2j9DK@w%-oT56jYq#R4&G^{N$xV@zjJ%@r)*p
zB$3E|u8TLQsDXFz^sEaz(9k9FKK!K7PW9SGrWR@@M0ustS251H_Xn`rFQ@E@&yYv<
z)(Z!5ze7<eeI)J`E4ZrNf7;D+hvM4$B&5&kkLk3#`01kavd(5p2BT%(?}8`5PF;R@
zkrT|cq_%b;qr**J454jx5&HHVUC$Edccx%Uj1_?{yqcM9uP-7fHW-;Z){XKCK|KRm
z3UUt2n-z<Sw4Aq+y7?(JwicTnBDvqurobV3gzL$Q6Zu%zmN0l4VhPF$c`t2*Qgz=`
z@-t<-Xc!a6T9gtvFE^9%IYG|sE3%PN3vD<VeuSV<u`@BYCTvr)&hw$44r(i>p-T2`
z<`8TL=*Lz_=6m-eubYpvJknZw@50w{NFxf8DmbK6D7<6JO&=yI_ZNKFgueBkFGwp4
z{#CF5&lB(0dSwvZymj`c7>=esB)b?>v=f|gR-X*sB{G^NKCip9;VO*FGN1I<^)#nu
zG`T%jYoH_6t;91Sxy>1>mX_f)XV0Cjf?GbcD_z^-n2kGXK5aPlSsZD>Las=EPs5wr
zO*b(WoQrm_`lQ}tDkS1M24#J5u#unpN=IZDoqk@1s4evApW9iOv_`2OpL|nQd-t^y
zz0tc*p;<ci+9xDdSM%Kl_N}V$c&DL@N?r9d=fjCvv)u_oKTXycT24QI5ow%s5|8T7
z;_VDyemF~xJU=m+sAc3rsrQd`FIHL+=ChnNvEHAB37YHdz=Eb$HMcoVf!NRXU(QoC
zUQLkfkOFMgfr1vo7jeR5H(NVQlrk*|a_4%R+4SW;{QzQC325{AUuPbaX@Vu>B4{cs
zoZ!sMiw6RhLGWPdxr4w!zk7qYcac55sHkt9|JCOB?0{76L(Q$>-3ix|)0<YG**t;i
zo9XlTvb$yxzN*Z8vnKVr{(V;@ztxht7}Vm#q{l1GQ&i6%@n^f2TcsMnW^cXnb#y0j
z`Ipt==PK?vRdS|cGOkTj`|#DZr*W~I`6Zw$^`&OIEPv-cs>@sBzr6J$Wkn4;ihqm9
zD0^d*uE<k!0*t>MtI9+M#*xm71(9HS%m-@QzPpFpRt+eAA0PXXTcuqcttO{EjKT6y
z3`A=9@6G-)W-SRMgz3cY8oL66oKvA+=S5@WaI5wRQRJsIkGK>rhWFz0MSReZ&(Xq2
zIB~A=2Py#^r64-*f#c1LW}|&B5N$(RZ)eCY?E0B@ThR(CBYklS=Zbo0lDw6i^K;Br
z%9_!x&Hg;Hun*zCW4>Fzrld>q?>R*=Do!)WyiRs*SDJ8)_Sa(Mb!5gWWO&wE_#V?E
zyoUoHAVj|QPePc<S+t9*vsv@+AWbfhD=ua9pHbC1_zlQ&=*{KLsf&~6mGOf`luA~}
z9_syx+Qv{RV^V7Ed4A&U-Tq84dt^~IGaI=@v{lEC3=cb6ET)NmEIKJzH9dCPEq~;A
zO0U35yXH|np0oJLLZ<e2H}_{O;%Y_J<L{K1*~_L!Vqcdr;h9X+<@Q>brKXjDO&tv5
zTM0!FpLb_IO5L+dB^EJi)e&ReY!l>5oz6hr^r7{&q^=htc+Z`+W_hum7ySL6Yol2W
ze_6ft@m-!-K%oFx<kvN>7pue6dJ$ax7b0^(gj}Q&*{d+HgFB*9jJW$=!Q^>JX7tac
zdgye<Ubc~>EGfBmIqkq{`<pshC|sF(!#eV14C`#>s5e1-jE}C1l(4PsqrjKZ*+d`r
zrnTo)Q71(X;W1&GmPYa#<AQo`+s&yhkB>e`(4*U}`L1D^VFhZiBML<DXk$b5&Kv3H
z7wR}n6S;i?6GX4vU)ia)`jLIv`ue(+c8_?WCf0)MpD&il?$D<=XBZB#*7lc=HBROR
zE(#4c5@*l5NUdJ>Vu>)d-6GL@r-$o#M`VPw<t5#wPtkuSpl2ZQd1wjoz4)LGo~)pb
zeE!4C*d)x7=JBsyMK)!GcH~H|@nekzUS459C@<Mk{v6e;b+Ns69O=r8+Y^$lK&gq;
z^HRPWeCe*iHp^8r*E<s{Y(|Qq^y-HEPFB=DyPC6JtErmUa|Z~`doR3j<AP>)JphU_
zD`_6wS3uxC44V8sJV^v;(bQ4p<<$@pB!(L;|L1!)tM*3YiMG__HuiU1bY%^4Dho7N
zOS$Mszpb~0u>QIjxpT@{Ac=@b%1fWeyMkh1cnI?h&E+Z~2LP0U_gyFMF^deyjc4Nc
zGFY-%9sbmMVax`V;p+VP>}aV|8Y+p`#R&72rFc9zGI+^$Rsj7JN)TP4jAvD&tWaQo
zLqQ)I>J3ja=im7m`AZ|C<Ku+YJYmN#MdTie(B``JrzmZ!zcLVuM8k83AXdMCxBj9}
zu7jK<sCD({;j)FZ)7fsU4@ft4o4@`YU}wt6uozLN<H=uzrUW|NOAXJ%+plWd^K@-s
zT75fF_E=kgN~`rMs5^1L6-%7b2#LibXj4~Ozhwp=ul(Q~nuKYc@gt}cvWiQK>-I+8
zJ?{ylY)W$s)^%n<e}GmlKxeNZM!K3qbT~T+aS~(Z%h<iG#l9^CV4eCup8V{EFQ((O
zn;m5kP|_tn!pwKM#M!-m3=MzU4<$pKouribJ}V8G-;}6KimtuL+nRauwCiF;T6Q*e
zw8HOGAle@5ghgAZgm$&<idN`B@Ect|9WUe9orlv^KKOS8P=4OVeemK?L)tIGaZ`?L
z<{XNe34yL6tS>FFx`KtmycfifLVuXrBQ$Hmu2YeBygIg?e(N|8S4oEwv(C11+dX^N
zFh}N#hj#tF-tX8?s6_;W`A1ezJ{I|M%9eVPuEf%UKTJ>X^^z>_nK=W22&}ozFKd4}
z$eP<q<hEH|nCI5GUB=#F>hQuqFj};7Debc|GPs@i(lD&j8i<$bXG#7%cf%e+uipSU
zOA<b3rA912I3Hd5?NklF&)vH`b8%Y(4|}f2gX8_8vRpfqYflQN>8>71FTr9g{25}y
znfZyLmK4Dj5Ao2jiNzZe%JN(j9S<mjj9h5S>_%`g8=t)1-Zp}W4+6Y5kaKWTdORx=
zd1jiJu$0$BH-;!@)%On}8<Y0$Y!1sI2fQqjJXU8(cYilByo{c1Gb&}rWD6>!^@LCF
zsj%zPejs6zhE`e_M<Bn>Z$Wz?6M5tJs%Bp@^2-<Rlcmm@=^3CoKQ^e*{<4mqJI9)b
zaa{5GR|Xf90F9Q?@(J#LM&=eo_#S(1ym)i5xbxvTzDgeVCUtL{rEEQJe2MsO<{tR3
zX20_bsMtm(vE}w^@qG;c+4<bc_8|SbL2rx4Rc`$yUa+j7`v<!NiQ{BcwFfY<Qx)qQ
zhr{2!*No<p!ui{DUO?GT3*H2>O$Rf#=X<HAN}P4~0bkqP2G;%A8O|wUou)5boXiwM
zhgv7h^P;Hhu`B~CK69~sP}K{+>PK-cwn@4Enz-C_-`y5pihjaOg86S*K@V+f557`f
zBk{8|{7QeRoKfr3EPoaw6()%kl5po>r$D8#2Wjhl&aH*d<Rd5YC))6wNzw6&L#LM0
z^x6EWDe!tm$l=$tLV*%FvuxPzkP<l~j4?K-t81|-7YSzDQZ%cRRr3S4_SM*LOt<L6
zQ=rJ&&wJ2wN9!BDa(Y!Uecx)`b;e;IBp`B~(TCj@k!+OrEtUZM-*Wn2QSB_oFiJoD
zC@PF*LJqaEIK`6+<?(T^oy8ix^N08x=^*R$tO1pEt?6*YrGxeN)YP)8)#$4}hhy>I
z9~Eu?Cl|m=Xe0#<mb*;5pkGm?!slt6e<FwOI>l%2a~o{IGvO1dR>KzoavfC{d<JqA
z^z@oV*`r>%HsPE=3*XXs4gTeI#)OxrJrObyJLB!20`vR_v!^875fmD38pqr%lbNaR
zV)GdjML*jsy~>yLsI&HAOU%p%uuYeC-rIq5csre)v($x-2#S95vBs*E8FXZD3v}%K
zd2DYrN!PbnFWz{`PpEV~8pyHxpsJq<y=B}aftaf13Lxu~)OtNH`yxdNS}*1ucP||B
zorM6HRBs;gz{Fr<*@UYXc_ia@wW()sp{T3KGS>U1q0yD#v|5`qbmGUXS+R&I*Zhxl
zBOfqfT@SV03xiS6(R=C?VS-oZ6)S_XZ?)Ux6~sB7pVi0k(bylOc<%<-gT?LNW;uWJ
zrf9GW<R|9ij~7hpBT3=5kxKY;-hu<NOi!A+q$*8t=ifqmcC~%UP&4DS7&D#hrVA7(
zbe}oM@=tOdR>7B%<J~;4T<!wHq<gt<i7LZ4U27A44QG(}`<@EVUtlCHr-h%Pf&FYu
z%->50<Ku3Hx4Ud6ky82ko6@seH<N*AGk=c_-kiy!w8qNZ_#^3e>=2Gw+w*oI-YqS-
z8$w~avN-keEoLO`wvo+VGum~7HshH@AdrqtdY-9ya++xBtnI45-eO_Kj?4fjRu_oy
zM{OM)a~AdeMz`P<;52SLL1y5xo{O|vU3IuD8y$vQ|A&}XK{r_tbz&iubU7hGg!PoO
zTEC3NmD;}hoE^-?y3{#6E;QO|vbL1Z6lc8YPuF46+xVG}J{yi$*oTj9d|&3z^mRwe
zjo98;`mxj3+bPcRty<8MFmn&uzD9=8-y&2udJHKcE@B@1RW?)CgJQmur>H=9+st#w
z@nT%>wVbZvk9ton7-jYikB3I((gcXkH%iZAWdgk=vHD`;f2~Im`w$FD-oKMZFj>kU
zzj}KhR=XU0?T;cIpv&~>b{<!YTvk?~oJ}y1&Q0QflQL8LEPAMm_ranwre*!^t%e(!
zVsBqQzVpun6>V)21qo=;+cfEj=k-k!sQ1U@%Xjy<PxnIChtui#2B+0dsKGkz)&jW$
z(aMDdM`Rc*`AM-QO!$b^)l{2&Pxi!XEO*hx(1;eDc*wuo%}^U%2%{h$;b8-OdORx_
zXp9rFo>kP{f#$;Bw&wn~5sas&CkA*iGh__Y<2{(9ZpAvcX?bWS;-|^q4e=$(XU~0u
znG%DD`(;8+*B42`G>NfP)}=zGPRqYrZpr;v>orOl>sv7$MwfhG0*|ALpgCinVwFE<
ztJtLV-R)3+l@34nBZ99elFCKAjY5eWct*ygMLR~&zarxpVOr-`>`&F0x+`AS>gazk
zf$FSEsfqjVcIx%jjYFUDxr!%er+2#_MjnxK3kAEY7dgvyMK=;yAA&40yu%sbpJl&n
zv8B1pl|R1ilUG+H2~~U<n>yz&d$i>yZ(cB8T68z>7)qeAzPCt~tu3BgkJfQo)~*;D
ztUJaR7OF{5u^W1#Xh#j|GMpzuqt@g5`0F;~bXW;J=gtLQSHAa<8JB&2hO|=t6tz-h
zs+(nqGoY}L4%=5frd)39&xynluhmC$nKldvNOS@viErq}8FdYRMi^h%=WRk+aOQmR
z6`#&@Fe%;=)-k5D!TY?__{U59+_T)8g(B>4u9V&JoQJGSDJ+@lEkXZYK0Z?(=V^1Z
z9-VfeC@Ly4W!6W&RHO`WE?YPW9W~Dm|4`(Ys{{#oRxR?;3JVS$KEAwww}0rH()#u+
zVOE{f9Vv+sx^GD__aJp6S4u!1=8ljJ$fBeq#+za;udpN5k0czJryqiK<%4P4*%}$r
z>0+woE-_x~RsTBf;;%E?0~-p(65<(OMqFfCkyajhwkn;*@7|kVRRvb*Lp^P~(bSLc
z2qSnqf``&9W}ybn6|3wvhJ^<{Y_6oX+RLY=r{(--xKqXO&cgSxHPU0*raQ&Vb~jnG
z4xoeg3Wq$~#TJ>n(-!8cKemYhbVikngb2TkrGEF=bd6J<j*ywkWHvQdmEYB>D<Xy;
zgtiP85Y@AhH`&wqNa})=fs)dXCpK-PpHQdM@FHS8S?l7CQ)UB@VlXgN33U?lYwTAl
zFRnwR^;Hu;78s4b%b<kj0z{*5qg!l~1!&I<a!1<tDE1qM#tg#N*%e{4_+QsR?f#V#
zl!NhF_Wu9>?4vwfdUEnmo^0Hy$10es+F?VyKZbG~Pf4(=La(K)S^q!$s2|l60w5Sk
zWf?g+ljo=VB>4h?x&I@nN?pr9E#I^YKC`nA;S2t`Q?REBRqZbTQ0~C%=*sZSN}3%|
ztYCAaX!H;OK=->%Z_PU5NU4b<Ls3?b14n6THNSqFq(yvOmXa7(ucRlM>-61)#<F*p
z4>(d4lnP(naA919j;hUirJnSl<YdY8;X?>#&!o}m2sOkQNJ)ZNNH?x7rs@u~IOaA1
zt)y@B6wxVH(QU0kJS^5fB7)CELbh<ej8Y;+h6I|3<vO~U?$nzOFPE)6@+b3KdNkd$
z7q2~DpRvtO8pZ1v+ky1vMz(FN$im@-k=*C5y(1{<v;{!5PR<_a*ReEmNqmm)bsFs2
zO!WQ_<gwSmFxwF?oZonIWn;|BqMb462oNeHC)c|t)5d7&Xo?qv-!|*YkrmD8>8_5p
zSPR-Ao!KS^Fht0*;gZcjznZ=jwY*~oa9L6$n?H!jf7!@LOX8Ccf>Fn+oUz;czft~#
zQ!Eot19N%10JTt1`MY#xV9}P<CC1NG^tET0E5=6+2^%!|mlO3%3`35d5e0|slC90^
zkp0~A1(<~kQr$Kby7%kD4eLFG<m6gSZI3-){FWZD4zy&GMx|n>D=+*04ro?oJ5lN+
zFeH*h^gGh-;F?R8Q^q+>A2dBn`gOPrPHx1f_TcRU)VnJ4Q4-;Y{ak^$zW#pPH*W&E
z466h+8tn9Pa&y^}<#+TA)D)hj6w=T+ecOFa{?z}-m#oxz$b)WHH8*<v5Y)eGj@w(G
z(R%DRQE*27_gB@f;ZpCh-`I?LE{5gWQ&O3z*1RY^7bpLZlTcp4+jhGZB;dcUN>BRN
z&4djLtXMjC=Vp*zffx4%cI^7HkRtf#h|;t%-KRJarE*4>mA~nfzmoMlm3onFlA>f8
zT2CqK?tKNMO?<J=x*g-xy)9V$A%|**^OXkc{O<1VFaTTdZ`<nV=;&qZB-pew05)Nu
zsi|oO*coKf+kA8^Z`)WH=lvWt^-!5d4(5;XLKL2~M+hl9pLWil(brXto=xRgz5FMS
z@DZmfgQQ&_`%EMmniwQ21kzplqksLFT54~(gGk3nq^jS9^&L7zN>*|bIO6IMo0@kv
z;RRPtA5b~5BDhxD%d6x1c_{xn!pFbK)RoXY|0Ez~N-oQP@Z?#(FSnZ9PbMH`Li+Z3
zS+kOrC?|(rT`w~;F)?xH<>25TDZ+}obp7)o)*O@^#Ts!XMSM^6{HjVxx4t0BXb8V(
z8Y|^GP`g~rp8i1X?RR~Rrw89(6K7QCkZ<Y=RC(N7_6Q1xL8+y_#;rOuzV8gcs9o^c
z3L;=~q}jtIx(cU30E9kxGM5ge`rKvujg!5~8k-CRcR`Z|7Ent7!M~%%es0D{HGX&b
zd1#qY)a-!82ZBqsiXM6^)|-*jFb&g;dpM5PYs2Gs>hwW*uio*4Iq9#-zIkdzkcEJx
zn>#)}i@7cx#|9Y=szsO#pPXqo|FdG{wqD5RY;n4Y=jUhLHtlnT!I6<qxs&UMhuWu*
z`vZuVdWEi~XX}3hEsGuM>KHeOk%7zP(Uywgfxas&N*StNig=>v{tSZ8$_buDGJm#u
zf@CKAuVfU9`r-&F^i|g^F^eKnH13qZiX1=wXfhO$N>)Eo$-rLx1=lZnb|nYM%<$Qk
z5;wZ@Uki0ut!7KimOX}Z+|c>ltM(z5^G16yhj3pwzbNGC+$j4$AS}>~Oz|bV#P-qo
zZw|yo$S`EbIA`$3Om;+V%|^509a0(>q67~i>=|#ID=)aeF=oYYJp{ZoQVQ}CPrK95
zNKlc-0wwYzYghbrb=z}O6k9$y!hDHG*foUFO73!_&A~%Nx5>deJr&u*Lrlu63zjX*
zZ42eZXialiRC%OVmtR~}p6_*Zxy+G?7l#rc?HP44pBbOFrE@2z{@fdS2W!8N8|noJ
zbzuv289(43nPi>C6}y@zdWPZ;S@d;T3m(HlRS@m(7=qE#MR~y!0W}um#vI8LD||hI
z{?CH8%k`4ElfS)v)gUWe;M;Hz(utdMf9EgssvFk9R!77g8^KQo!IHe@ch<_J&2?>^
zUmDb|Pgh6{3=AF-4Ze-37`}XqZI}kndhy(UYbMjj(k8#-4K9OSz96rQP)rBC-KRYJ
z!u^yb9r`WoLLs|+BmUt0(QTobkfrzS@wOkgm|pko)kuS!CsL7C4J{QF)%Wk;-;j~X
z%gCT8q;N6Rmi`_kayA;0>!TZoEOiE(;PWwti=b=d8XYwpZUyY-)nb}x*%A3ZTxUx~
zk?>$Xx&fQC?1%o)rizFo{c|PU(b5C5<F@t;dF7;4ZEfjf^eAC{)qOVs64(BCDY<@u
z9;E&T?ePUOZs?P#>4#GFGLsV$K1Vbs`=e9Wnr03B^U?w=5_!fCw`2m{Hx~(#02RE&
z{cQKVdba9B>4&U3b`HI(@4B(m#*wyKXzsk>oH<wnkG4V0A2hf#oQR5zP3L`~T^G#H
zax?dtkavH9bjdZ)O^>RITUQWyJTjnFklPZe?y-L&Y~uX_JF+qmQ@^tOj?`^TMyG7a
zq81gk)pN|+GBwCSbwuJ{a)lANv2qLSsfC}-a5q8;*Zx$BrDt|7uDcCf-TaJ?msC~7
z$HBq*^y$+ZV&b7+)Y5~0|9-5}Bj*(fnK}F9P)B>$<WYUGNyVcp_{mG3fXYvSFU<qX
znJRNJ>bA|IQsPfCut*;7?%v?yc2}DX6Y<)S6&Dwer1Hijb6RMd`Kp;sbQw4sZq7SA
z#e)2X7iCXg$PzKmVnTSOIt%DE+bM7U?UxkKersS%YsaA5+Dz^+LI4P&>+6Qy-Q7Sg
zoNF<j;o&V{*WQ>4P2*0{S);BHX4_|8xcrmCbd=WlyU=b+jns+r9V77wZKD@1LFT{y
z&2g{I%}w#~3?Wo>6}%cOa?twr_I8zt$ryT-2lrDWTD5cMB1|rF*9`mN0|WIe4JSI*
zr>@=fbBo{2C$@mO?+?KwU8vfaE20OeYfAuM`CYNyN;CJ^7*FFT52&Dzu++(mZVjp+
zl=JD<sL@4NsD{0<n^TU;g<e+}lW{drNTk^QaI{%H7ZKawhluoUze?oF9VSe<PD6oG
zrttUR;K>$u`x1=?zZc6IW7?-FH-gNkusZgmLH4&#<ZCLtk@KK-@kf$uU73!~pWL_3
z>`+|oTCU2uG|Zm44Lb*mqhMQ~^F5J8qDN#>lNZA;J%O!L8#LdZmNbQSUswTRB0%}H
zS+0))b}9}9g`(AAI<wyC>G~0-#^!98Vaa`Wk&%PFy(@heok35##36`R!^7&Mi{trm
zftHrm@nX$@lMON!StppOz~D-OxwI#@%WBw3oK{{7UW-?MuPoFPKD&aKEGjfu`rRg)
zU<azWRWb9Ft&PZzv5$`+K&nqL5pX@6DUTmxv0`*`&;TzJr_=>6Hycc4F^sIiiM$|3
zlf!kSO!$AeCIeqBMC8(=(J5urlIeJO@Xc4O9MWxXDaYyC4&co2Mb=04pcD~!NW-t9
zn-EO4`f2#BuqVz$S~s>l{QT_}3{*!9xpuPW7jL{7sHg%52W8}P$Z&AFhT`bZT})=d
zTwM06y3pr|IjRp`iH05hcP$>y@9*KxJ~5dqZY$#BFR59N2xQ|qx;quDoz*x_60@Z!
zBq^gfl=f5Jwka`YKJ<N3<p=-X1R*X@F4kChZbk8)CK<Ev@+RJ2?bWy*sXskEd0bCQ
zqmOLpu0kZ67K69ql!P~$_A+W#0~2~6C*yHX^K|ZzrQb|Me1y2uMmzFgt-ZPC#t)jV
zZxSd2&`}Sa+cUhhYY)Ku=xzMgVd|y%zK=Hv2nEtH6dqn)3AYnTZFLM6#C_KSuoYxM
zu9e>`n2GNB_w~>!#)m^z3?y9m_S3u8E8Gqyz_WH_!mb-m_`R7^_)CPVJv~7bLSCAR
z1OPkybYH?iy|8);d|`pA$N9I%Cd9<T@RkZ?4LflQePv{J_HCQKlD2l&0rIt%P`P&9
z#|dq6GBVwgxIAp7?g*UVvL8>F?HLFMVH1`g1fR(p<2Un;rcgC6SdMi{Xr~le4pY&<
z|5^|5XW()tWdMuP)YKr=1JLf~E;AqiSyNX-ek@&JpgR~{>uv{*<wNoGDgRL#GpdW`
zig4@F1`TNLSo6^LBmEOjOPD`d+VIAaGz`(42MTM0f0+8Lw(<VPx8+X)n4boRsP+^d
z6{2>yb{uP&huy0kY9!ChH|u;K-Ei^n&{yjH^iXWkhl;EA0+yzZkKn08#XE!zN3XVU
z;zoWqP<~QW#Cp6rsUcf#_kBe3sb{`yAlk2MKtiZ5si5u5hBZUqIaky4Ad3E1#0o;_
z+;*EN0R=(7Y0aMoJCPF>$!d9cG}8PLbURKmORuTn1k^9oOiYpfUJV~=EOjQ9+b=dn
zm`gd^?C#qu@g=DMx=ttRzj(%mX2ZM3M?i}~393Wy+43TCE#d7bWn6uE@FWw+P2ebA
zBkRBREbl#e9D_;3#ZpklAIlUmc&8lw<HtHMVu<z5j*e9H^n+6qxVX6DU%vb_!$tDr
zM<q4ki+B3vPMvT*y?8=TUwdQQ39Mw|$t=?3ew%OVg0ePVf-NCFG7Txr!%qS*S+7Yi
z^i<#4<SlJr1-L{T4zp^9Nd84pAwxsZeXo(b1`c_=s}gzrwnw|#Y4&q7pc`>H{YNaR
zywK*&>kow_?u@7wPEY<Zk(Ps-Zu7o{t4aN>E{j^rj;t<>&fv)AfrWtZ_D?n(-Ah2*
z6@bJyXaDrl(yIl5y;ub=wz;0Byk64<zZa<c)by-%6Ny4s@6zyagnpZsmLdUQWGxQR
zRJD1Ovg~WROHudgj}1ptxR00WhA)LJ>z3fkQ2`d{-TnQq#KeJoxx`cOj@7@$Klzuf
z_cd-OONi|#r!rXzQku`SWzV5rmag5QZ0pI@YuoFa>~|3Ig_nqjg&I&wE7R{|447|0
zK?jq$$YWz;$IA^d(8pVy&5l$?T7dWc^l((kq*)aJ@F`(1K=45IvDcymRq7RSsp}bb
zDV}Yz8HkPf)_vx0zfOs~=fowZKr6-7lsQl7oj#FxGIOnkiVFtZvFo8~AOeklg#Pn*
zb(dk5{aQN)DNiIVX6*e#sm9$2_TUgArM-tC$~Cr%{cmGi_c+nL-CZX@J2m|0WThn$
zprZ~yJh}etGJMC%iY=);0j{HOAN%RBb?NCW!tHpf`FQH=f#|xv(cNP;j}@R;ckp}+
zb`!d@Om<HMSisTb0!#yL+k-g6pr9_%L>4_J|3pCj)!!#+3wyCdAeWMEj?$EhpskE=
zY;4Sw*#~$D$AhWz;))WrQmM1mHs*o3W>;&VE`YjcWIu^#^P8S#i$uw9DSRYNRoMH1
z^>Cj4xu3d$ZbB;e7r+)JkV&8h5dF`eKc0gO4A$TvK0qKgn##*se?k8cA>wg+)x;#+
zf?1O@F;z%mAF61o+Cu@hpSk6{c@AZbsWdZB#qs%f|EJ|e5~xhxtM(y~81BB;yd@kV
zLr@}co-Y?XGrZvldNsG+4WP3)fNP&C9b*pA-ie~5>h0IY04wt&GO||n0RYTg4rhkG
zCIfl~NHlC}Faty2Y6nb?y7+M51T-u*=ITLffrD_WLNsc3q6nGH0NGMQtZ<$kGFd#5
zA(TqOZB+%>k+}-J;S_FbKL6)dGfH5MY;XT-b~)SHQpw7(nJ!QW^9?^q7p25FQSy$l
z3Rr@>UbgS}1FS$eIy6u^R+N;=JIe}{LaZoK-u~Iy_}{<rqbWuFq^;{ZI{X7a->_C1
zKBv#93USJ|V;Bgd4D@tWQV6zSz_t87GaRE-#yb$vfU)`t^d|A?6GC_npamcTkn631
zn1Q~&zDrL*kMnhM4S74a=_b#<Ps;c=rz^wX5YaW8U5a-!0bl}<JCcc1j18?YXE9c~
z`UIcCuKaapx=(t`d1UOjl^Ocn7uJjSTH66IpH%yHSwOf`uCAE<(GX7WYeYtuVJDbN
zJJ#Ia-yaE^0ucT6?9W$H0vWIm;NX024hBq3&9`=nX~*DC)x5<PKU?DN18N2=3PG@J
ztJ~SCg1LEt8T7<ezRqeEh#{-5Ic=9}8SLx;r{tr#@~(<R@E;av{?F<=QAAhTnBdTj
zJ69Nh0p2Ze)+zx8gMYEg$;ui5$b7;r^UH03er2Ka-dI9^C=iYjELMHR!P{S0Sb$qm
zxb$cB@b)IWqd-AJ`*`i0)r!j1IaZYabhPKlRk|GJue2PbuHyl(=w^rXEEMDLbX~5#
zQ&dz0m&qLP068Ko>In)8dKJqR>kN(qF6r^CV;p<!C1A)VC)u5R6)U+jslRQ)(hiE{
zrqBPm62dy!%NRgS-QKTY5N-Dj;QtG&aR(~?4S3uY*Cr_QgH0h|*9vkTy_xa;LjL6X
zaBWGzqC0rI=C7oyO9lg^BSwFJGstHM61%(K!ZBkjD>opi0k~DO+liH|VKb1kf-{#l
zarf`0t8OG6DzT_LRP9&vNSczXwm9#+S^2!hGxy$c5-UOvC46)_02R~#g?FTefp0uO
z)X$<vla3}c0<0uJiVq1ww#W2qNp@M07l#F65;WC3z>7EMzIQT(OE&=A9>B?NAX^@l
za|5(4fMAS+g9A+$WCu)8c6N5cqTpB>|Lc-VWHn%EG_l}4w@0~P|Cx$7+FUwRQ~7r9
z%ZmV9qMCi2uOTK3kZ{%7E++sM#p(1Pzo-ywlAtFCAfC)tFW3H$z4mN{hRdddTn7gS
z-%*L3RI1jPD*!_;ZDYeE;C_mbC6~+6y7WNfHFpr$FE#xZSe>Qg<z}ZB{>~py@wLkn
z0AT5SH`N}O2%KB|?xW~`)@ErbBk=UUf9p8gG}tW08uf&*+bpnw+)pDF3#l0x!jy#X
zzblDAf7jM>`S=PEO^Kuj`gZkUar(ba5G9S0djI}CQ)Zi9i`&~G&eqmeKnRfVK^t&p
zjgBT9oSe)Z@W7A&QoUZt(fneqC0=;W`gAMS5FT8%j?nUGMmCIp|3=L|#?ZjbgVEjj
zMl1AYxz>6<8t4pI_qP1?fVTAQUCZG_wnQM~uY_SJO%V#}HQ=R~2I_2=d0wsTk)V(e
zx^5T-78c8Fyo~jSiSV5B7#iwDZy%Z%jHjPCLi+CjBfGcxp3j{>);W=ykCaeMxBCE2
zLkSQGG3(SbIlN`(<0BFg>1YYxI_r2b2PTY3xA6@SZN;TPJWKnC`gu>Ip6m03r;{XF
z;+1gHNu=-Ao31O=W_C8e#v(Ld7MB1U<t%g~`7GPkSMvsAAI<X9Xr}0zDM#|%R*Z;d
zl_6}YMkOsTFCkDb=<tJj+--!umdyaUlWvP!xv02r4L9~oyBhCfdoYl|0AT<IlEg36
zS_W)&;X*8|)-&5WIjLlN2Nz~C%`7)UEMoW?gVa$-8zhO-0*jDEw{vkZ*>bwz;xA+6
zn>P;De`^~X`R2-X1^`>T(CS&UV*>zhn1Ja4)KTa@L-R$-Jzs}z4yHqQM^l|a0&U_S
zChi%+yEl45nafY_mGFTzetv$w4k(}zwOO@m!D5=4M8Hz^@btvN!|PF6e%U;)C_m(K
zxwmxvQb@yv5)wjnIQ5R`-&wo8QvhRGvG>o0HNRBI(d?Q<gWU=dztaa`Gi@Xq`2G1n
zBjk0xz7b0K>SKTq+zG%4h3~iV5}0+=*CTLdSV@4zBA2-YAQcy@-e;5CC3pp#`9!T|
zh|PA2WzG*#uMGpyUheck5|<^eed`4h;0@~SjIcgujlTwVVnb^=d#9(-9UUTo0xh9~
zE?rR?$W@6!F7&`c$_EB1;Y8)@*Sq7I@jGB@fNQ$u=M%MR%r82juxWQkI4$JMKV;42
zX*pk&8EF{D3x9Gu1Nwn5duw;MKb}#8iLOXrUmpP(IW9e2!PGP_pW<Y#Lj)cP!|>%9
z%A`||A*l=o(%G}tE3DyEy^THV$73hmpwc>%hdzSV=ZB+&gfjr-`jwWZn<b|SnC3u6
zf%~)EF9gvtS(Um?U)K>%N0N1Tg`C{oS=k6^xwx(?`n!Mg2CCx+m&&q#N=M&Ity41M
z+S%MWRl2mjVkAw|wNx!-ZhHuw$c{^4LPkW4Km2Pi%c9%Zna1y&B;qd&)E)s5Sb5_T
z*Z<@K{NP>li3?*PpR^Y?*ta$(zoVp-EMFf=khm*JeQSr!gzB=R5|0O#3@q#j6^ZT~
zmN2qfNMFWJ5Zz1s`Lh#%>R;D04vv+yCNNnPKIdW-es@Mo4fghyQEgrjtH5ESLZ!84
zfnR{`8<_2P4XaeG+8@3MeHrbr12v+&PZX<ZpEycx5!t<qtJ*R6=}7(!#Jd@z=6w!<
z?6jP<n*iIdx36zwbF=%kb~4P`AwQaeSm+MuKMa8&;ng4m0!pn1(?f_C96Y?EqvHnf
zh5$q)UPJ_hH%m1FN2GV+GK@ufHOJ7v;UX^7>k}5eW>bFUg=(`18poT<%N*rgDImT9
zaP%{oqRImESbTSi)+eKF9V>Kon-5Rqu!K1?mdBT+tLy3<zE_0?ZC{~^aJp1>jSnwC
zKzlW(WB<o-OA5y8=6Hy)vb7sn6t4r);#Pdc-rU?gRb?dc+M%q~H{jI)0&Yy<IaNR@
zXe@)-Bf`bP!UAwi;_itRF7BsNk0d9K^%vV4yvR5tQBr;n2Xw2xx3BpA<Pk(ZnIjlT
zil>3rj_}P35kM?L=_h{@7dDQ^ZU%c27!k9u=4}5pvIB~_{e7?;<@$@snP5*2Xr9}r
z=Zj>%454!P4=&iZs^Nx_%%5}EX`5$%42?OS5pmNNkSg$o9rU>vM=}?LwIP1_!${5)
zpb7Jfr48fg=`M+z6JN5|x1O$jzRxT<nY$(_(AEF!wsm|OcohH;+b98^>@{ZU7Jqcs
zRaWjQ%=G(<Oz9(PX~|GfP#}0U!U>qppBX~lPW%=^1K1;t<$E-;4i_AGc@mC0P+(q_
z)eGvQYU=9xR#!8CH0W1+JdHadkT}YL!StAzn7|c70{nn<b0nEFbYkL5JM=~e2pi>8
zRPf-Du}tqTcjN&SQ%_INJkNLN;ag7w<X@4BxEB?4{LnB?zh^R6+7di*15d_g2!l?*
zWY`}`oRpmGIyC((4b$9;5F7o*i|SYD0JaklP2AI(n3w>QWNLW&k+&cfc3;X2KCz-;
zSkFu?w5d6`qgq@8o3-#)<HF;X&0a)J8)ch!Q+8?T{=^$2X+{90a9{NdweNr?JUn;-
zYRBQJe7R&H@A|t{E^M6oGHMMayjlqHLz3n%4zssH&D}@$ETU_kkQd@CSTx4(jMBDk
zKc)HV9N3$2M6K1oe#He~rKhK>@yq(g#6&_MJTu!JN!HY|q!eUjRYEV~rjlBd`GB(I
zDZz{W<B=#*!RRotTWw3d<c6fz7a>tG9DM}{15g1-uf9IPzWqMr_2w8gDhzFGUf-Xg
zRc$I;I**lu=YPDwu7#DoXyoqf+%q&3wzQ-TIE6Z7o-(FS-?B18@2ZcyM7G9YYR&5!
zzD&%|TTEn$X=-UXZge98Va<$A2ne*k4^OeNX>CYPlp?xn2|N9<XiOHLgk<z^75(Af
z%DvQfrT0yalqFnk2>=fM53b%is>;3V{-!}mL8TFu5~QT0O9TV~rIGILZlsk^kd{=W
zyFri+N$Cb@>F)Tgec$65&-1?LuQSeJD0^SuSZmJDTvsc|mg9A*G;V>x!LT;^ZJ~8S
zJkkOJT17?0>%B;y-9}XW-J>y`11R>9#qjpwhjevysrfNX8&ZRAi=$YfVwK?2wF;34
zTj=08*4@v3S8L13&AornFMI$krNi~5+u_ks0>29j>#xgapgh-9(AT|2(o#@<J>Ayb
zs?|fgy<qfAW@M)<H1w9Yw>Mlo{?pT%Z{NNxZ+uJJi-n78H56B0S{l_aeDf9d@>@+!
zS-3IW&i`e>ER4_hTW9cmV`-|t();pqp$^JfTu4q$OmqzvfNpQ@G>VFXg4pA?1K$gJ
z9Gmt=$tH1!gHX1)wdJ-^P<8q*D^-Cprn;u4#PiJl$LpjYv_#d+NHb{xVfyS+@oLnE
z{kZG9NH#V$H7%Z7lU%btH(rB-gSDDvKJSpuHJ;EsP^2ea$xh-93JNOCXA=<k>g($}
zrN2BuB{DXw{K<hf;VG^{gTkj*-`>{!)wP<%C*~Lax^-n+O8MgOue?;>`A2#>@8945
z%2)7Nt0FmidkZnC<j-#u{rUmXXiQ8@#E=0X#wvWcr;ye(j0CTJgd|>0&^kwYeDv^e
zdE+{)<=b$ue#0+<YAJN9zc8>#uGm^=&72d(1|v;bSyi<Iij2??AM`sSC^@;f(zCKU
zT0@CK!@>+NPxq_}H(9l6TcdSwmV3HT_M_7?0*R88eQfcNJsuR>)>hCJ(uivHFpPSU
zVK<W7{A|Kz{pnj+$il_N<#{xsny*&!9!e%C@0u0m1>4H{g*>#j)KpYrp9YWr^7$pu
zXl!B<?4pZQN6wU%efMpjNQ}EL7P-uwsD)MJ2C9X*Dz&QH1BsbRjJ)oL*`C&ky0>(v
z=8~9>#i?&_3@M3ND5z)CyuHoSobiztQx=kR9MegEa5dI0l^zpqkl&&nTT_{;b&I~S
zof4NSDJg*?O~GPahn9K&<YW~zT+4}4iM*thwY5&DkrMcv%$a7~MMaVM<1Fd74Jk0~
z@uSoWwdYAUNI9mRNgmewV^v}|kNqYVoOxHqXW8Fgn4a#J>U|!{GwbS)P0Uh0xapEx
zi<R~<*CM3u%M?y;9Is-&HUrtlKlc18$}3rqw<C4<^!+JA8Z?+_Xl!PKG*EKy9v@q~
z_~z8sCc?6ha+&jr%gA_C>$1~_&;ftC+U0F?I^v4R5H4xmDb7D<95Y2pOnC3%v-XEF
z%w2Y&(GS%gbF#BD#AByEE|iE~-^@(f(Xl+XoPmbsMeO~1_wJclywcJ{%)NkW%SMr9
zfEH<H94(8>m$!gmoolf2Jn?-I?sZek_}SWTm0tfJKj^%#j#Oh=b>v>Yd>LnUT!#i-
zvgmJi<nGYmzn5mXQ^JC_C#Z9O1t{5LtLVd{C`GJI?&$<<<rS8%Gp@!NRt_*R*_Q7&
zPRz=0GxATJ>Fa-=`OC`qt0y|uHN(<;>0@Bdcs#$^j5ibPY=uTB6S_fB$u1~BTqAMX
z?x7tIe12}VNsuF5?T3#1OfLW2G2One)!E!v|EY5QKPRsXlBx9jn=>l?L5bqun*>Kz
z$;hU?zq`8)pv!w9A<-3d#?8&$lgRZE_OF0|K>2o2j$dK*ZOI0IY{`IBhi~oFC3-H!
zM<^aPe=@muJPt%C*>!nm9K;iZy!g=3(Ze1Hr*K(}c7D=SR&Ik@w+p^ndYx)wW{uW`
zwvRSz?3x-Hzd=4S{3xJ~Z)S+)U*A|!bb69tDaD7&Wsc3XS-{0NoTBS?kodO{&9s-1
ziD?m0C!cK3tZi(3Q(_#cbX_@SGo?^F_D{5b?uFEsdbf*<nv++~zv;0o(vhz&+aOZ=
zheuqpdDlJs-zX>YzIBV7AZs|wz`?W>DSSN-H+59-yKvZ@(v1*(Mx2;QJT}Hx=SNae
z-<>=>q8|JH#_u*0VDi0qHF6JcrNe>xgc@@n&;D8?!c&;YE7~~iKBf5s8JoL&JSI-A
zj~}eT#g~zhVVym(qu@lCCSz_s;qxT!s+V%5JVXDCTRiZ6<7_2PSyhWd`n;_46OTTf
zv~u404)}U26O*3XE#T^unLoVPuG@L5pg_%PnCiBYuo1N4h>=DsjaFcaQugxTYl>jw
z5as#t9c0(lQVC*>d$`tCR?Mszh*1Tw?3>|Zjjz~RK;=C(_NDKun^v+_7Lvx3rU`vm
zcdFEx1kj{h;R!a4y$r~~DMfZbj^Ww&Q(=s;oUZ;nrtR^1bQ4bcBvoy<OvS`2Ib=sC
z+V2#^lq;=|KgImOFb+h)z<HZR;(qR*{)x~r5a03L$nfwBID=xfrn`u4X)+H-tZ_eM
zYwB|TL>9YsX<NS6nr4(O%3pEO>f(9aA-rct)t2ONVt&=Pjw3>R|Nj2<#m>RmYARfO
z{sjfhFdyfUyue=T5iuz<(z)RqI=??3i@CVzx0;IB>PNQpduy==T(qw!k;u~0G^KdS
zl60pC*M_D|>y>>~VrZl(c<lqSvl+N8$El|Rq4Z^Q+18|TTe>6r`gH|T?Gnmmja1|Q
zFl$=jE3~wwOcC#AN*39V!rb2{$N8FhJh&5U|Lw_92y%DAQf1iXr6+i2a-gcf^@-@y
z^v2mLjuOo`POnbpvS1+1Ou^!DZbL(glZVTlvzE9=YfB$qhGhgZ;T#YTXewac`j;VI
zkia<JAvcqJd~&zMoF{t<c_2$Jc_BTR>JdMG5_FbjP%<NP32z$C?ELvujcT4qd^>2S
z(F4bI4x7#xC1^;$HI<QCWzFwE0eL9p?|@H$>sR9uWo$|(hboPmw~s!^bZMm2<i2IN
z3ti~mfH*!o%&RFaEp>%?R7O@pzR}S{aJA{K59h$IJFQ^<{nh|8Q`NILR5#3|Ukf!#
z>39#<%m$+GIJexY#D4_UM3HvN+Pdc4dhs_=MkgmH%F#`oRccC<wbEKXJni4n{5?-p
zKQ8%7d1+~(Y+BuNaBQrPLwAc&-u&ICcay0}rmW&H6IRnd5_*vu-%;<d;@tUJlr~eN
zL?8X^?2Lq*{9q|szSOWQ3a_oo<!&1G?Fzpdwa3w!b)iVc{eCDZtXe@tp|+xrO`888
z2c5T+u~IcpO$9?C5Bq7n&ZQ`}eqp-CfmG0)^LF4ud*UFuQZQ1j^m_+dvPMtk<d;~w
zbK*#4Md3)s-IXqmd}QA<1~-4f>>!wB;-pbF_t3X6Cun%_lc$nne1E1n>%fIa=-n34
zmPQ9Gk}5B2E#v|C#NV^BEF?7PX=y7;ugW)#!oLL?t4`qaW5+octns~#qS-SaCLt_4
z-d{r1;jmsK+PC6C9RQ&;U$Z8vU+5w%H<vNLpup{KAaOj8jdn*6J{MALLk?tvh@&*^
zQwj`X9duHC`@U*iI1#7q{5-Ei;tYAdL(J36rmA+u`fz!T(RVZ-{tSd=0iQV}G!*%T
zi#h0dQr;zw4&CDB@7HJ*zT7(bi6qCo<wASsM^xO?G(KSoPC;cOOFG@jJ{hEn9Kix#
z4fFwO7L@K-=Ebe5xtH?tkx;>v>NcSi#9UQaj3L^5wHkX`BO@c|0!m!>^xa(lxzkRK
zOA~(mg|kOLKk!6q%$B00Dk!aqt%vv!mzq}`cld~mV63XDmVav%`A31zJ7-!EnW$Fi
znOop$&B@C%*js3W@<gm*zFU+r{K(c__-ojG9SP!AWJkvU;=_3?<nCqF0C`7u98IaW
z!k@(+XrglYx>tX)>ty{Nsy%U$V0(<oZ8iCQMAX{UG%Yvx)rOM=h5{V&-GkvFAy{yS
zMx>?=tc;9}l^J!%a??r<F;mfLvJ-mEkhk=<eULFv>l2jHR!;FuN=PRab$+~rS5bH7
zt0Ll6CLO?&H!+s$45g!&;9hUz&zrX_q3D5}IRaizPNbF=uj9X=rbpDXE3K9AtUj80
z)*t;g+gW$7xLh|QkAp}nnMB{jzc4~B*fUKT+aIB{K%MNSw>$07-^KlbGUphnHXH42
z5?9X?(zLy(f1|hpn7(5oBm)nA|3+s^l_K!L{4;P!f+T*`8dq`8FgbXW=|Jt5cl{>C
zGuK^RC^O-rtbNM<ulUCy#t0phCWbNGs1C?9#3C{6JJ{qzceV=FpS^Q>I=^lljb)=;
zHb*BiHX<_lq$21{HmY{3=o@NbN^he8#m^v-s)(!0>8Bz9X2E2Z@AqCtrKHT7R}Z$e
zWh#?|de<{cjyhiroQ@1qq+AYT10dL-_UVJN1C2BEfaT?1@P#iw!Oao>`}eD-%*<}v
zHLIs5ziwvqrbI=s({yCiy7v_=5|;Tsq`nlIniw+wwT*I^tm;D#T|mes*+PFaIeydQ
z78`tM+nk-bp-5eU+T#nRP3=2zdHzlrlq~A%)w777t_*eg80$X)3o}J?EHmAf^YBl`
zKP>mda^VumGlz<~=PPe?R1_gXKxw$SzEHX4yu7^ZdAZktumrxu#Tn@7H9_6m4ei@h
zIET#`W^FYN9z9l?(B}+{<?R{j0ac`mU&NX{_t=s@(LJgo=j40U=QYeGXi9y!IctQZ
zrLNxf)9pa5&Y2k+b0~1w-S!O&F3L%lnWb%cm2rpEJBF(!s<GmaBYkJLZI6q6{~eN4
zf5~k~ZAkZB^S)Nl!ZY&`O+Pi7v6s<|q5(;xb$g~xy}~?(qI9s7lW5k{%L@ukc5tbF
zc(~a8wUV(nw6bh#u^u|@@coCwwliM*C7|_epCTOZNR+piAT91Qy6ExarSum(4m>#i
z;r#cOXP)$WbBL&p^4LD-(b;B8t+l^rbZ{c6DjD2VIU45ZsQdjJsyDX^MK`Oig)uj*
zSH*$+Mb))6f`##<(89jy)j@pvvQd7TG1N7dm>+2vYf*2v@DA>UC)|Gg7`a6owaFtb
z<V+|At%HdA$?F$B{zw~5{%C9$XTH)r$io--O$}GJITMwWgM(3k(-4x9hDJt4UJ6*5
zz0&bMGotd?ERn;_@0GfUI#R_GmE)z+e1qx`*gl`2%)`ZYG3jfyy6~5ue0gSIpb_Lj
zw}W0Dn^$FX94jdIMz}uNz3OC-ifH_YHXtkMM3ueG=7A}4(x<*sfjJOG^Eowjj&VU_
zND!lyts*Nc>plZ8KII9{llO!Mwni!+>+Fuip5Zu?ca%;{{rU5!hD9VUCZ=nw5abd)
zSQH3YR@kTEl7II{E@@h{Odp5q0g5~6G4)kbQ3>0oyIT*mZKxYNXah(!nQVUr|K!hH
zQ#`Mk)Am(S1{noLXFl~bd6spenr68omE^}}OL0j4&~bp0S^E7TcCUTp-Rh{@PGpJ)
zbg={koFVEbtTX}wnihF*?0pot+--rv15H&`6^^igq9U8TRa4UWfjm{#6pwA*6wm!G
zmAq3t^qk?*rW{V>p}nj$rEf>e{0fbSSU$yU{)Y}I+47m9FX=eV9gQ?864<R#BEv}K
zrKF^?tbPm*4M8u(2wpLG(IsX>befu)8Z$+&eXZ4v&TP$@zD}Y%@kmQofB&Z!=aLR9
zo{Ce>v4mrR+6kY=(mzVGpk*y_f^Fm4PT=}<5e)<5x7O}4XC@JyEHxLGvP^V*N=o59
z6TE>U9br%nE1EH2MB<t8-q@x4J`W>#Yp_Ykk*+aLe4l?|^gpwKy^mD^FO1diV<zdH
z>uL)Ga;@Lbt|+;e?k`3vxT+2Gl&yoqhAm@!F8regdt$U$OQzS`$%!i%kMfc6j`neZ
zdT$q#s!>fK-uxI&Pr8>o(Qrq~WYEh>ljUMAjwCPMxNNf^QblppcD|EEAyxYJ=SOXq
z7@@RQ$d28ArM$enNXscGD6H3~MH|g!WpSn|Jv(%KwjRriJDZy#q0WFRD^lj5!%pLR
z`NemojEp6STSY99eOO+kQtV-33Vrl^(ba-{CZVHK(;TL@<~b7)z1-kBjf;lj<V)$O
z7l6&dGWc0{#?M08?Q4%48W(8C@svlr%qx=RmioV^B6+VW*xX@B!wfR}B<ydswS2sT
zmq)_RX|nw2xMUywysal2r^^ng_7f}2(`e<#lRBFe5{vUmiSe6N-W>6ZqIXkKQBltF
zRwpAdOTYeksP#OD_d@+}K5}~T#xx1Z3&f5I4y_80bGV`RhZW_WCacJOUuT(3f9@(~
zmUk#+h@RVaOL%f+O|&$NzQ$j4ZH<ImyiM@*;-1*_-&IwN#nZ?s{DE9|GaH|V3m3Y;
z)fdF)8V>Wx4JEeoBy1Wl?B_Po#mwJFQQ&q-l5v`n=pAm1$yajRhWuB1d(djCPYT=s
zB;w-YLKmAHR6dt`-0456tK$(@1H$e?R5$tgh~n~W?AJAI$LOphp<XmvR#|{C-!;wG
zF#;XFqdRProFvm)FY2Bi<*&)ne{cC`>6WCZs!A*@EPOT%T+rlJl*K4mFi;>-jf{*S
zu)Rp3^XwUHpI{`7vi01DPDN|7sgo#UiSPVx3;wZES!>BO%r=`9E4C*@{Qyf|dFIwL
zc6PnaPY;L3#vB$}Z;_Fa**ZEJ!c0v8Aj7xJUs;%&OUcO@0|s=t-_2TjU1KycHMMB9
z-iFkIJj^v}ziy7wQS>1dT|Dq%^jErPLtYpUkDGXgd8(e?3gyT&c^*$Zhcg_^kfZ$3
zc57>EnJoqAjIYVb{5ZqRTo-`g7_1GXYs}FrpQ*k3O}R2JbR17c7NxD!Q`vK1*(=b$
z_min~r6$!(p01XU{m)23n$Ypu9Cd?s87Hzv#mG%Nxy`5K<nWXfsz||O64v@d!ib0n
zN+IJnZ$8K38ozuQbbfZ2<mZ-lYH#JdCC$(5dv^Mydh)0DpFe)rEGY4)mrYIDf}cE}
zs5&MtE{w2H;2pCu-hCNCkZjRQ)BJAq%Ol@2sFF24uOzC3#gpdLbvIP%W9@Q}|DIzM
z$;&M)jM!<oh(OdaV5Q^W;E+5s!~&rXc0#A!GSc%}Zu~1Qp9JSW6p~)a#l_vS{`rny
z5BlEP@)g8vcK7uaDz{dYl{qBcnV~G0x;21w)|Sw7(}U?1_)v#M)?l*x_?d{Rl2Yer
zWbRpV+ehLu(?1#9b;~?jNpfWi^)7CEt&y@0xAhjYiEXapw?_V>r0rl0ye90?MwjIH
z>wojFW1&ccJaewF_C%WL*jxR@98&d((ZlZNWyH@6(tfxx&X!kUDgArE@|`p}hmO{j
z@h0?HoiJ=$n58DdWIoo`76?xZA^9LIaR~|ZGUHzTkzD01^Hx~q>QMHk`dstE!UJ!=
zQ!DEu=d1>y-nDXiLWJ<GP*QWsw3?i?SD3i>x2awRDJXVljnHWd*)P<-smonr>pcL@
zRPS$dpy$O_ReXH>ZZFRqn?d^#W~nt<8kdau`|rucTjI|>2*^13ECiqGtr=ngS{CdZ
z!7@5h<7s-g8^DqA2Cq65l<%KE-+@aDYDhuKh)FAR^96Vy1|Z`@|MBhHH~-*ZOk-nX
z@T!6`9&I^F5P1>x;PyVrV~VaD#1o{5|9gzZfgzh^6B&V;HoEaCv2!!7_ph}i42Lwk
z*YgLkuQNf58iNe=M9O;5IS}eH8CPXvK|w(R80Q9(;rioE^=c&Ke+GtMTxJSX8XK2!
z9SHFDe$Ph5si<0pa~D?bU%iREp|3xLS8vf&`Ry008ynNXGU0Vt*#$q<(R4ij`}Y7a
zR7n%TI9!^7g=8qNQ{?xWAx~RU2cI#I`k}hWL5`uRb5u(sl4t#kEVo^1J(;ZBT=VC)
zB-eaM3B4N?*6joWI4=XS#z!@~Wm868n3<_(ab-d!kepnA-owa4N|2KV-xAQPCEw(c
zr|MRx8%S)z*Khk;TUstHbw#(fx0l+?2?2S#MDr5K!NGy<bz<x3-r`wS15=D3KQViO
za?J9l7(8`Slz5JzdHq=Tae{}->vO++nAEm+3ASGkZ(*4F_U=8{n>lVn!waKc=}hH(
zl6$uBk=6C)>f{L{V+JshPc_OzKqWd4!h6f3NczC0WoM%6aDa!->bI&Ixy$QpxupL#
zbNncrjcxkXJA!Njs-Nt(9~~bbR~?-pf!j~Iwj5L_roSFHX*>P>ZWF!9Npt4n!W@_e
zY-){;u)x*HwR;~n2L)lsr3gIHX<b-cG@4T(_;Hc@Hx*31JA{OTvwAh|M~pl5CsaC~
zdx7~Lo|_YGKmQsqYZz2%#EjT-zCk+rjJ9Wkgz81IfGa9Kek3f<(cVT$khtZKiT```
zM-Kkz(9mOQY6NIDJw1Io?~e~hWIF(xIzCr1eyEtUyUUKZ(_(NxRa5GJqqa-hjTlvK
zZ2YV{BJY^kPod=TDrw^V+zVMRF*hZ?CTRP_L2?3RFpQJ~Lpll}KCAc!=HwvQ4twIU
zj*hGuxu~}}hTu%Xoa3($22RV%>jslDe#8MoV)cX66FJ%%hQOr^Ll&iSN-gYp{q-Kh
zILy<mu|+lVIo(f15@vXJ7-L?63$uKZZ^T({-a*63d6$EO1B`%>@NiloA?>_xU}?)K
zC=h@yyE$E>I4ane$`d(zOL8L4AyjLE_siI=Z1&Ni#T0L^__g&O*EsL-ct^|mP4~<%
zG}q<ZEgN%J)IK)K&L&pU@0FCbwfnBGE|zyh-@c8ytbsO@o{dex+a?&7oX*8{aZzX`
z*d}#l{E3t>YH4={DbC)#fz8E-)EdfBXkBI~iIHCkRGg4zsWPSnEJHr|mFMT0clPKw
z4<!!5v~HjWGG*2&>EO428A0(RH3>Sj%bKf5nmgR8CP*4NN>TCg45?%A5Xv>5z&*1}
z(+=Fo8zrRx@MgfSn=flLQ*y2gEzG77{d2D-Pu95hv@9(!^=FFEAvsbZmHyg5n}1|G
zdP(Qi%F$7B&JCIf57!g7?IhhkO}^&CV3Sf-<0=9C^nKZ1+CicwuX2<K1#H<Q`)Esq
zF5Rf?s3tNyY?NJ)&-?OzF^QmFRzH|MCqnEb;2>bAh#Lp!d#y0QH00yQ$Bur!3&*)X
zC`$DuOlr#~)|!ej5Alg2%WTbWN)j_cUCIMJ42;lj;`aa+z;-^`o=L)_az!jV`0Nuu
zSgpWrfeVO5$LlC2<X2^>uX?ne7|Gh4##)eO_$w*if>`&Z9QsTY6qF@L9#&S?vnwWM
z<`ER4-uXKOLNf>luqU1!*v8GE#NH&7wQe&4Vsy;W>`e0GZMJ8GV#a<6?GgpGvW{8u
zFR=w3!^x<@JxzFKbQ^&l0D&uSB^}k-)%CfbiH<I4TC*Z(vbk9dTNae<;rgwZ|6u_%
zkSk#EwgllKMo)j`cfkuReE-iH{~-T{!``LlXT#Uw1<t6CZ>65RqD^P}TxMiq7FF&v
zB8cNYZ^0Y#8802D^suQej+lF9a>rJyKk!>g$!D0!d%QK(0Z0dEya6io^2(K!m0?tZ
zi6Hex78kQIpL~EH(e~Vv0K|?d_!3Di$uT;UR4)S0aU-MOIGSefB0+ZcgYe6jcA*s2
zCBSi=pX|)7U&G=@+<e>3DSS?^W_kY%4I!!%dra@vkURD!svLJ5K?eGsnc3B`IVIWo
zEa==%hWxJ#x+*$n2L)P8W<^Zq8fP#%d{yt1wNR&=#Sih2|L_cc(Qfcm-*8BXj($8h
zT4o~emJUj7dzz^K<P7hb4eA;=8}OgV$-80CR5;5~P*N_ZR>3o74waKh^-8v;lfr+|
zJlv7yf&Iz)ZF8pOALTytZAKHLknf_s&;B8^m@BlmiFb-=xLVAyJXot-7e*d9!sW6I
zar1v|$AokX9B8*sr$G?{$@x$lhCm(R;4`Op$?pDyugc@y(tnK$3{xYb%f+PauLAG7
z0Oqx+j`3kST8`7%2a$0>pq%`hcFy+r{Y*3o^$ezDQp}@X`%3LTl<!)Sw~MWd3Nc}=
zf+-AO=YAJ`YB3BjR}nYOt$ue;!NNLvd1Yn)GO4KS_~Jo;M^HFzPYTZB%lF!=^&;<L
zXqo~wLxe-+_3uCFkng0d^S(OoVxuG^e1zW!^=KKWYz^1v7I$x>!|A9o)d9v&!GiTW
zy#8r|w7l;50e9p+^^PR`Y1Egp!NpRxw=9Lxvr0)Sy^#2Y18W%_aq9G}^<PMbYOqp}
zkT9esK_C26;S)`0PmgSY+*}OPde<`;-OQX^Qdd`ZIo{s>5&@6yo5V&de)Uf&q~gHr
znJ2x;vRj*VZ!C~g@BcSAwAoM{T(Da8`Hg*I(2vpUKcnp`uJ7CzXS&HB$m}f-mp!QE
ztF=No4CiydfB!ycw2EqK+33T7?ob10F64Etpr+P2TklcPI5;_(IGFK>EHSZo)3~S1
z&)=<itWIXWMmj1)D6W?rGcrm=DGYCS*NaH4avv4m3$T2ltgL+J!2{ORDKOIvi^oPr
zjCSXn_C1IwDDp@C6y)aW1N&<~d^wV*&;FyRsJ~dm1><=p+pv72ZFg>g|KpCpO5UyV
zQ(;t85}_B4N%a_&H+K0sfm<HB2hI=i{~6=-3;qg|_?R*}_+@2v)k6B&Gma7Gn3*~k
zdhqM6H$ijO;t`N2yDmc3h@yduW69y;$B#qj>YFolRrlFWHpX^>!ii&yZ!)XqByt&u
zR2(1g_%~c7rSodh+#api^{y%(2{KZ}FSc7#$QZpL_O&+6Hy8`GnYep`M?1E!u(qCV
zX%4RAdq9upQs^Z%?}K2Wo1G`=-paVimzVi7;lEx*GBVkG>0q>7ks%njV(#weR)0d4
zmzFxGzC?Ntebk8lnj{odN@X8$JHGEV8qLFhYvc^NlTG*<eQq;2Ag8&z^Wh7fQfJAh
zwvUd+!lM9nkH+G)lcOV`K_g_-b^2)^FDg^#S9^=MSpV==mIX=Dx5N{$ug?WKHT&S;
zDdQr7IiRA$y!gj5ZxMhF)JV5(-`0l)>8pU-+it5RkZPVewzs$c2LP#te?db>?^@+H
z>?9EC8nkI?FTC~k<d&(>b?~H9i^G>SZv5myK2C|U+*f)R!NrS2*#!jdYtsFC{Qu^b
z1gFCf4;=y535|$YS^M*CU#zLCOIj`f8y9zkjOGhXC=>&xe&H*v^-z!Ea6x@{xHFep
z8q=0PQyFn0U$CU#MtkUVl&~xmZE{JhXEWH4pOc>BS5JJh_j=uhqLV-5?b{50yz|>$
z>oJqb22dLRbaz|WSnmE59E|vWt*u!gNwj7wSp-)R%nT{<II-ronU#eZyee_(l9FKX
zV{X@<Xs`6AMy3X1*7WzSep}2^b7D+b4|<vZx^{26_@MxIC($AvO0uzZ<EX>LPg0+-
z?|BOsd?WS3teEKY2R+!sM-F%JsC(yb;4d({>M_Nu!@$I>BC|&rryHXMFC-<)pfGil
z?3|wd3UDa_<b%Lm%wd?B{f(&+H81<a_D_g#F5b1H!@Zf6c@aJJG3}Ajrigl!uSLqf
zC3j>gu%p|X^4(rT`ZT*jK~bIh{b|B2JiOR`E7(Akw_ls(T@NAEx0$fRsu)FCi52EZ
zob*&x9ZQ<bdCI~POUb4^eP704;SPfZ`*7YjRJ=Y7LL5y2YqQ2(IZOb01U=-g8(B+i
z-Y9vKdI^fRWA-fi&6}KBo){7m%I^aSu@Cj3^A^!#XUgW5mWf|J!ooslc#)%c>-(#w
zW&c%vU3|IANqcH#@b{XrS=k_e0MgfRW=9-NXo*2bsYyRo!}&TB%dyTA`a8~*jM#st
zb!@F4Xj)i5cP{h2{?@of#U6l_jA?4R75USAgb`-R)9Sp4d>~vn^7>$x^2It5=HC%Z
zgIh_?jS<h;aZGJc9ey;frLmy?ixnCG8sZYzn0;;+Z|?@rq?nj{0OE2&t!becotT<h
zgpSEvXI6jpo_CRe<xy>2LQ4}@@r@*d^{A7^ao}eM>ccmGv+e40S^Oy<v>ZnU7=Eim
zm(@fuSn+@3vKjVp2YD;_?%lhc*nx18bd%|VG5J4)S9`bxm%w9bLWn(Uh|H8G-M1uq
zp%8forwn^uclM#s^P@X8F*PJ>WNaM<-0~MJt)oll4<7__Nm%loFbSusOTG-X{jjDt
z;P1Dx9lw@jTewLP$K){OX$d?ipryLQ*>9k@0WvsM#WyN7mDPV4cDqsW=4?GbBn<%X
ziHeEI97#XkJEj{JiBpio#L&>#_0nh5p`}rr%5E*JyVZ|N-aIXuLgGAsz<|tOYVXQ%
z&5)JY8%tkTQ*1a+z_>vkQs<h0dh{xSCd>2}VA-UMp!Q68A{f}^{>{C`j^!!P7pp;e
z4-E;SW?+!Y^Uovw{`T$dl0B5K%9YN>KM3yA)b`@>kZfh=1#9^<J$M(09sK#>G|&BJ
ze+foCa}3SS18PA*E$=XBaUZ`7`_xogpq11&YK3WftFs;>6n)%e_|F$t35&6Ia2hPe
zj3M++iDE1rRqeW$p<=EhCiXt|m>daJ{9(<&A3yBX0FRe0it)#2)57Fob#!Fsa_9&k
z?LzLVvwfWB$<w-!lxAN$@l@UQPU5u@Ma^+b$xOCKj%;w3LDAz>bA0xZwLup~N+3W4
zD+eJF6BoHSs$VnzIhkkB*P6y(^EdDQZQ-?BKuhAb(g&+l%lQ&;=M=CnzCuNV2s{jq
zBUj-J*h|~eYB^?>;fooRHTn07kl3;c@UBYyIIEvIYmp(G^S)q)Vm%E67U&ie?=h(W
zuq7fP5jZ-Z@*{;vOG{69>{qJo*z?bi=#nWc_4@p6`?2g&-(?@aS0W=PSFd$^0$!~M
zC}sTo{LIP4!NI{#gR!u%Ocg)6o$VVjrwREq;Ktn#q;j3XK4HGEwMstmb3~KqS-8~$
ztdMPEj1xiB#rwV`AZ6Ycv7r3%BM3r$T?3i5r%ZQmi;0UbJ2*HyTOW-><%WQ~YONAk
z9^lPTlIZc^Xi;ii*HUzktmlz1;v{Ua2TL?y7jHXX#hcv<4QPI$LhI93`IT~*bF{ox
zK9WkXxV-#36trNEGh^ZsP&$LKx4F5=z{(1)Y2bGnr^Ih?%X&tLii+B2mBH`P5{%!y
z>U_BN#_5EZ#L%FPfS<fjNu})K5`8`9mHWwBA9hRde)fq7&g3bDRAb5kvFp|R8@VOy
zqzY<G4}k)AyC2nGzPX(ZbLZC0r=+B4Jj<$N**rQrLNJka*tzKnl$4Y&!$?E{!HlMR
zJs34rp1)Zpyf?BMS!(mkcK#Dh2@B;qecD9-Mf=9^x%)UJ>MxsnO;0c|7VLfqO3=UK
z7@TTs^aZ}ZYxO<}iR+MdKV<<6tk&wcPagdux=o#f^kKMV2o{t^-Lpi(oa0>h@PPkV
zG}-IxKg!IUbT<O!ff+N*4VPPKf;VUG>u>whEzHv^%<LQ#MF-clTOsl2FRQqx-LsI(
zOLIQQ2mGhbeLr}x_T;d}W!FF!+vApu;?|drKk`j=4;tF{m%9^q9q24Z^MTnLW9Xe(
z0P6poA}0#`KMz>4{6*U+<ydc}U%zgfn|lC%H&iy06p3{czk7SzprMDN5T3BDv-8UG
zuX7hl5ep@8+^><uY+ji)>V~HI{f~!?A$Khelyz1x4G|%s>sop+q-XRYW&i|8iL&EX
z=zGnemtGliL_I&L3pc4*NCfM|C%E=}yf$^W#R<4&mvUS(CQ;V@7mxq}6QQ(%y<iVW
zWToYVn4=>n5MrhB#c()-w1}vZfc#PIj*y9>0RZv^1U4;3ZPFTQ)qM+1sGAj(MtWtR
zTet><hFDA5Sj86>1Zbr;RmLc&Khkw;)+hOeH5uRe3&DhfRRhW0FM6^LYT%4029-26
zHgeDKetAV39M?+ZK@*EwgUS2FN3F6DwV3T$){m0d#C9^ABb>9W4uhw=UbmSHj{1WF
zUO1n*@rLi}YX4Mul-^L-kRLOux^~d`1ru^zhHwTm#JmEq6AYMnK=(H1ex*XuQcgw%
z1%4I;c>;oho|z9b!28br&HcJ@V!q6M^(y3QY<dN)SyOI>*p!w)!<pkpbuC6F&EJ7<
z{H)<vy0IrBhfYqPeSg`r>`nSTrsyWdzAIy3z_`AVGxRAOXU#~?G<6otz4????N)Av
z^4>9chRx71LKS5U5v9`hI!IfEQS!%v8BQk~?^|(nM9ct!A6QTL`RYUx>Itr?=f{2?
zVh<hB4mXilxj36sNKTnGmPr1si$!jWRi>XUPR^-6cXX|7ap=xv<8E-Oii%K*Lvj?2
zK{8f8JW<AFH6-#Pp}EWyZx##n+P&Ivr&d)~1|~m$L|IN+`i}Sz+49=jIO9SsB+4~@
zIZTD^W3V%vPMz#_L@K<)HLjok>&9k>LE_x=uFh!Y2~q(R$1G-)<k#-Wmr+me*1CsL
zQFN6T$CYqzOsyEa6h^NNI>Mpk`vR}MffdsXU$o6%oEOqjt?}#zKzp!)FvSHVP=Lyp
z6RV;+B(GIai&g0FH0-sL|2$vMyIg#5GryW@V<pWjO$Zxbf2z_7j5c~mQzFe*B9|M-
ztTs=0*_!#_zbgFj5j1O#&!0ad41$9BD)Ek|tR~yjHJ8919j=dj1VW9?`_lF7SVfd^
zjPc<=P8x1hCl&HhQ)5BBeUf-~=_!Xi)Dglo1AYDY|8i}z@oWfKF{c-A9*BzZ3JYSe
zB%#iJy-h|+M)u78&=k(xCqS!#uDb|qB+}uUk$BtggxHkj4$DI2kO6f8Z%Un#;1{VA
zZIo<Gw9qe2sKTNWpGg=geqqrG;L>efe3``&smedTe|vOvbYDyl2we^A!2hyn+hgOy
zQg7dsQh3`xzrh*_AJOPf&&rehG=ueiQ_{BZX|%uF-jZt0^0Sqo6|^bLJC+Y#gm@8b
z-``$RUMcyz$(mnRoU)m>p1mJT<&FjQFaUUT?*{w^hR>maF*1?>Z4Xi6_@`3O4|Q}%
z=C!PDZ53>XaZ9@8<z_Ubh{uh5p1lPKMMX6YSRc>E`x1KwpRvTM4wP#4yk1s_{p{s#
zFr{_nWts586%t!#_0L^*)1JDNR#nCEuR&U$XLIR3F)_qwHGp!Fl9jz%G&CG=M?9~Z
z{_fXvXVV{HA6|{`VzyE#Y<yGB%)>71n_p&$!kM|ieiEF^>91t%;&dCh&?LQ@nr=^t
z$59nYLZ@nZia1(5%Mu>l55b3QRv`=n6geXl53&b16wei~V=-~bh98oO)*%(-?h}q5
zGFx8XS~qWOe&u@@@K4v(f%4|Y+1J|qZK<#kDq_h_XOyk2Ggf!(yrIJLK_@@h=$==S
z4gh-V5j8auIXO8Tx_rq}dASHP#pn7U6(QxrbC!JSy^X?YOPn@U0Dlw48fdyfc$BYz
zWQX-3o5-0n)Nuu{TLPEGeYhauWYO8SzobXfP$tsSY1?hK^*;TJiRf0myaf69_z?O~
zjl=3bz~IxvD)}_h|5@Pbr2%lpYl9n)pke^>2#?0<W_ls=N{oy`=DlY45<4ZOt-bH=
zUor8mZF_NYfqS3_a_t>Y*_2r(47mR~8hs~lGVMV5YSi4)l14}<Mfm#E9|A+>W)GpM
zv~_VIHVgcF;L2V>*%(JtZ7=8R=>GKkRc2+8Vsg?95|$6x%pZQ~1iM<Jr*Y-|{HdwU
zU^&;IrOiM>O3DrkA0ZvWX8u>x^a5ioc4S~!L|^aGyM7yk)iy!HSfk3)(xI`Dj=VVf
z#fxwg&$$v;H5^Nx3I*ZLPp0(+2+HFtC|MR*13f)qPZWRbLMH-;ZMI5kH$WxkZFlJD
z=qlfn2a5TD0^)0KV<R9>+0@Za_>o&^aoOpJqWVj>jY;30YM;?<4Vp@1N1g%t>?%RU
zX)V&3{?D_5lP7|s$w$vSelso2AoVsW!Gsvn#H*_-ewsUo&^cm?#q7@ncxsN0jwBq$
zI0(UUH%%idQqX<IV0fhEUs?0T*?^s%NtU`7WmCTl(t!vMJLkCe)flp)bU0Z&5jM4|
zL5ab4%-PZPyDiuE(@gg!lZ+_+iw+Gv6TV$0nAb>gaB~|&en3OZh?TVnkDjG6pBz#F
zzd!tqD|+Z2J|J~c>L+=PPDF)#6~RSz{SLH?#OLv~p-vaE4mG=+5b+PV=v(@{n>l`D
zzcs+l`Ryly%b4}PumU!Ny#eGzxfH)oTW;>|8A^<ZBv)4s>B7$cS(By&_JQzj%Gtr(
zW8k!1U0oFv6~$qJLT+~7>PFgkl;RQekO%@6@EE)G;U^*@-G~2%5CA)Pu9UH^kgkn~
z$e0~erLfZgW)fWfLB|vKLQMm&4!<$g<6)7=1@~j|U3*kv`P~yFo>diRSNhA2fvU<-
zAWBLt##k(^tR{-}P#;S=Pj6~h&^Ew9;8xu?I(i?3KNyK+_?)B6;d_q<;i+WEc4>p&
zhCkPEa|3)k=wGz%yv1jt&t|wgh^w41k+V!irKh`?i$1J{4qUqi1ETSEJ1_<Wa%l?V
zWY6U5mRkD>Ww2zC-&&fuw{pYNuIHaT6qwZ7r24Ul1c`pH|5kYQs$kWQ$A$j83){nE
zQM8Hh#^z3x%<oG}BZ`}y6I~H`b<d--oNW|%EwddP>FDSnXcGf_5(Xm-KJ+Mt%Zs^K
z?)7V{`uf>2On=M#-IB`I>CTYBM2#rykI!w{<4W(Jj0}xrmIMZV^Ai(It7m>p8L0WV
zarQotz`tQ6^@^t9yA?O*CGGTVx6c`%f+3{#kPzL4)=<ib9U*wR8X^O8I3)k9r>3?R
zc$_XcX6$SUj^8JWgCk!=Jh?oKBPAW|uwL#V;@Nu67lPcRCq8eyZ*bj@T$`_A8mBzq
z`DQCFA{|Abps9%We5!zW&FtwsS<<{1uVJi+an5<a`If}D%uFMuTkYY7Vm$9CF-(<X
zGy}tRT1G1d6j&{{j0J8j{s`UuvN7yKE+X@mD0tA@WTs(R^JijEL<HVrFXP?LY}J2O
z|D3Nds~BGU76O1iH<t=;D}|VP8(ins=SKhw!l5wH;KiTnbxeA&+OIn2N}T3bhcsh7
zgKlc&i{ZK)qv{Ijka!@l=X`FiUQ^0P(I7A|;8x&eTRVcWPt`M=E^hAqe}X}GiLqs9
z!6ra}bl^?mR>UJoNJt>U_3&;tT?2!bi##9Q%DoD5p~BwZwOBeu#Q*%M2XV+bS!~;1
z&M&gm4tZ0}ue9R7!q!eckY~|p=z*_a4d%%OKgF7wB;?`YF=@w1zo+U1bAhmpX4Xe3
z>#~c?wiU>Zd_@;^ZP896Fk7|>4k#_X-%0TvhR@IXw0vJ}9DX@Iqa{Z)W2PKe`^sRv
zGsA^LHrU_fxpRb^7T!Q2tu!I%HnIGfkfw|!zD;*dxAu>Lp-A;t<ji3$`!vfZ-3wi@
zO@C6ue+W%C1~<kt;qqQ8t(T~gIPGmpbiyc=;K`nWu*r4}^B6IXw4wp}%M@Gnqym&~
zxbyZQTDw0aFWiNi3PD=Pi>>~kl>`$+zj5Q#PA>mEtfl7c`@Q>$L;7lW0`6<jF7B4E
zw_SV_BOMv@%OA9leIU?Nl*RFx<`!z;dv{#2k+G4?;B@5W3a8S-$>!lNp}mEq2JK<#
zl93Tj-3XWz6!(x19~K~qyx_&gt~Xh31{@6`1;w=aLpun&1?wBzR+7dma-a!Vq-3O|
zoK_VU4oJ$U4py2tO0Emz4Xo;O`PLxW4|4~vF*OdoW$v=wp}odOyK54cD|)7x<7z2m
zTCs&!xKe%Qy4!Smc6xq32(zuG$iAZ|Bbh4S)A9?~X{tOL7)KgCW-|~9OPFL7kx;&2
z2Ec|;^4YZ^EYl@T>6em{!2~<OZid5ohb6Ukw=Lg+S%n1oTbOc!D1hE7D)L{JW#u9Y
zL62igz*N9w9GR{Ei9OJ$<!fEFjpAtGTbP&E0?QiIP3P_DF3|J%J%7GV;%2v+<OI9w
zx#OBb9$vc-DnJ&XJi_Fq{T6)TVuKF6#Kgpbp`k@Mnhl`l1R>(O*1Pca>(_`h4@9;<
z_y`{p^i=;E^7s4N?U$`1u|O?iL$2rN^5j8bcD*|LfMk~FKr=kmK4KkDyO}Y!dgeVy
z3vB<ox!!oCu0k`izfeo^{2udLROLhyN=gM1tP3JRRFW&!5Laz}t70SR*S75H1*t;y
z4{}H3NN(pf^wi34@lXBPUhY-WOP?HaIr%YTc+aS!Hqx)z_iv@2z~(XA;(0{Vbg|XU
zamBfXXYx67Dy_GiggC7`pF3x|WhU#WC?=~UYr6pF%VRv71k)@fMhIoXsDU?^P}`I)
zt-*GsefDfWDC#lugH5lCkcpDy4D$-2ketsC@4pPXt30OgW=^A0|9e*B)NjHE7EGR2
z`OvH_^p_=Ex}Q@J!ivrmogniKJG_gAOZi(c`jO$z?gn}=q^U}LZoHvpJO%mB=w|=Z
zCxVr}<nG}d#dmFn7IeYe(<Oce9Yf`#@m}u!fxMbOaWaiizP9-5tzT_r;X0RF^N8d{
ziNhsxX4gUYyKm)o0(yw`zK9HRw|C{j4#w?=ivybsCb1&gf#-JA-^smt8%HQKnOjp?
zRtB^>sriVOy_iPKfEUyg%D*9Ww!WM9cuHtDGE4VxNu`qMW(=|BI{Gb*wi=8KS8owr
zQ5WUNKBx$5;MTUH$XtZ}0XfC%SY7CJ9`*Wa$LFApSjX{H4O|L{e8U11hnsH$?N1#N
z?lk(E^9GAZ6~GDLPd(Fb#d1H{er;}Up0roNoL5{-ATKWuN;9w1`dtu23bqgC&;}A?
zQd9c@tAY&ugQZVs78Vx9y@?WrhK5P28M*QCBvkH8cRmzp2a4#pE=&Kd{WHNAjVo;V
z{&6p*87mLFrP>?xjf}s!&IEoieS_j21r-&`6hTg9t7?%p^MzPlZS6CYJ__Uhl$pPM
zKU4_y3<NLt=r+%lusxC4b)P>+xuh4p`I|Q7w>3vct10!xPh+(Q`8!#~kD-YPbtbje
z*=vmU1}h~yJ9fw=m$oGt>!+rtJQOboCeVK%n`VVHf;+NVQyxi4>};QzJMl<fvWaa<
zr2T}vqq@Psj@{{{*rST)ucbIxHfc)Oq&IlO55wZu&#4X0!&a2^P%sY`q#0hlm>4dh
zY1ojCKrD#)CVym*GzRx;Mh|bcjVzybvslX3me6teTO};$KOgRk%|g9o<9bGullEyT
zFOcT`UAxNf<@YYDo?1j(3;Ywtk{n9ew4r*{!=mVOUrE|C*WuvcKs;T+c5gvKrAYg=
zlF}E;aOhEOAdT7rl#Z>FvV{mV1)w4Zcxq$fQ+*L7Z=6-3x$Ta7w`A&_vzT>*+E6&o
z``~TNI+ou7>*l^t@+4B2Q4tY|gW&lSJLB7cg#At?96pEd7c{sSQknen-y#nr8hxjq
zA0I(EPxAkWe#mjJct&o<4jP#Enrb)bo0G$vWjk0mYDnL<<3!JIZbL$7VQGRgf2u0X
zp=P0$ZyF{|Y^%!&ze{CtYp6tmPAE~-ALDaEf;6C$;Jfe$M9@eh64k;tm(h?AfRL$`
z=l%a}UuL*PP@h7x0srfn&8z?n+i+?z=#6LhhnW=zFfL;sB%dm`)EkxmfAZ#?H6nBl
zmCEBtfx#av0zv>@cAa34;cjT}PKxcW&8>U$DE~_!sJIouHC0BsChmpmTC+D3{NSNh
zXEbYb;{0&EhYJoau94gC@XA&jB_{cnD5CI3BFiT{0ocJ}Iye|XeSQb@5+;0tFRkpx
zS_m(<iA1&pF~ayN(~3Dvm+*t5_uVSc=;8Kep3`_YHTsYqvJQx5OX#J&eJKvZ`T4m<
zN`rO>Mn>FBEN1D8w|dX`SvCzN`{dq!P+Xqsp)V2={T~(p=e60FFG{1>Rj8noqF4q!
zYnXffe)DDVVzXSstFxfoAVzmm_2K~+<VcS`2_A*W=z5~^M(JhBLdbo>rMP@>`}XaO
zcMgu)@1h<90YW1^xL?A}!$be{sjR2Z`}ejpr;l?2+rVS`vgR;>gi~3uGBU3hX1ARe
z7|_PWwb|3qxKgIR*;OO;`T#FD<E$*ha&iC9(134=oL2&0vGB2CR`f{P7h*<UUf%ty
zu<&p|;nJ(CNu;CLI60U6Fawh3z=V6la2F&kLc-y>B*=>^sOLCG!yD?~1bb9}vHRN-
zO~H#y?j0;sgYs28sHGeeP1D6_hEYUSMdf$O9>{nIlaPv9nc*%D>9YV}6=0mz*>}#M
z`mW%S#;-FPMPUh_N$`d@pPp?0Vdp0duQfo{P*!DWn(Wu>-_d`#J&(pgEHeH=#clFZ
z68U_9EHt2J2C+yv;v!H8=7mB@Po8$P&Hj=}En@b?cn&Y)1c(UUoeXc~&0ppoR>>I$
z6z9|F>1pNBwhJ;@HTJ;PDBQ^p(g5=$;WQ-xb!@Td*YzU-fuA#ri0%z|6C@C&Bd7!?
z3%}3*qJ%CLmOkk7;gAdiixBDW0-MqAyHGNca+!b0C+1gF^t<?M8gh<1dijv=;mtXw
zETzj3F?T@}O?u>Pv9guY#zs@sPD96{!8U5^=uNHD2E0|6wStgP8J1P%NShN-&ulMv
zxVfKg_&Zsit~oJu5+XA`e*D)jP68#}ctk~H-x_)7)>K+)jbfn*lJ-eRU-=D9v90OJ
znFih+Zcfh5cPH@JT01(rr9F}00ntEJ!o<XcIOFY?LR2K-ypQztMIx1z3?q5y|HO1^
zyh53_f_Af}I&QA)nws0iCog_``+SCU$^{4ixg^0hK~QKzwduXF*j=~5IM3*FX3&V#
z$lsN?{?@<mS*PI4&cMVpwElvQme$PA8*=3^=oYO1=(spZP0g<Yw%LrJsl(J7*Ugd+
z1g9Gmq~FMKH2fsC*EN2wKO~h)hny)E1N~Xki1f>*GTKbbZh`*Mo&$=$-Z=jLdJjjF
zj65Pp&fId^3>T9OcE~h>6E;doLF#f@Agb{G+c~j5J!9jbq9Rt9uvQxBs;{46NuTZZ
z3r2zdo@xWq9YM5I?xDgks8p4Ja`UPeQ8W!VE2OiY8;@Bu357L8{ZPKzeBm3PK*-01
zg>*oJF;yttGyj&C_c1l~0oYEBO-;x^`oKIOcJ|NZ<>e|(eB&nszSDDJGkV)2yu@L#
zP93P_x%(`Qll{6osbY^tI|4hjJp*m7O`L<LL##7tq!;>eefIXTi=yn_n4SjX#5ab$
z>HfFFc33H3;(>&_3wZV(R678;J_3B%D~RFg(^gR55yKlkeew+smYj0=Kh@&}60Ys;
zyeNo*cJ}t>F3RsTG(G_23B3x1ur~z^F>{V|p$E}2S=dK=)>}n1?Ay1;fW9y|tn{Qe
z^}X6^w&T1rdnx8B8jR71edFL@aFd&x6BsksH0!!|`Y-xLu3m=J2XKgu=NtV1Z&0<6
zx%?yffcE+Gco^+gvyr*3i84MhVb&D-@uLh=+W!fAo{DZTzEC<@QnjWS!_9c?xDAPy
z=s%x|oKqXSbe_AP%D?ya=2>ojqpFISrUajkmW3rkCXU5Sz2@KAny!^qZ^)g;(Bfe%
zf;pO*HIiS6#8)OD6KhOSZ!b%QQ_fLh;5la(z#sNqPQ4>V=3?AS4?okXS9|g{s@SG@
z{J_(rVvdq;#Xmy&ql8wc%6D?C^4t3QBB0d$jF8m|4tr>6X(9ge0AVn|clA@bG)XKS
zG1>=)T5#w!qe11QUhiIENooS!UcfD)&i;Oru|h4WsDP>%S{Z-TgyA_-BSj*~gTi8f
zT;Fk&@v3G}$sJTxtN`b_?O><>DSQOi?hT5wiOziQ={E`+xhI@#Y^L{6zPSrSm<LWy
z$PXeWJrqa~gs@apRq?>=TCfPjt-fWt3rXvP-46q9!UE9J>B)rdQl;QTY!!9Xez!&&
zsHmtYcuTGP5d$_80A_125eM`GB_*X#iHWbAohwJ#VMN2;;o(wZH2cIcuDR~EvsYE<
zyPO*yxZ=B4|E|hEq$8ub1)CpqzP>L{6s@tPU@F_*dR|c}l)*@s7m-Hof<^3U^15-t
zamHHsX>exj9UreiB9B7A^%K~y<!-+usoVlU+}no%EqjA9tQ9wg|9glw2*C{xgGmr`
ztOBu#_u*@<s`Y|NGx^%}iO`zt!zc-z6EznEIG&Re3S}gs_J&NMApsT^?q9WH;xL@M
z+>WXr^(N>)dCTFTG&D%P@Ng;b(+N@)e*e*J!LCRc_jt5<{&-dRU&$FOV`9|~ge>9M
zy$fl{{GYV5`&?O>nPw7d7R?DTB~1nG04!*?t%@<&xdfDD(j2^`l}c)TPL=!pja|(8
z1bebS*Y5hQy}Xboe2X7xyXj6=iL!QMwDWT@gYRJHCM6xWUg}NG?fdQ}k8h9PduX2V
zr?pz_!;c-F5l&>6vLBaGucv#Bx$2@wWC-7(DdndzWJpEpH=&^1{(vPFoN#t!mw?@Z
zq~B#7wfs~2B-3_Bz=2h9+fAtd?5m=Vj{K`vA6&1KncShThTueGVj>w#6iR!k{j<Kl
ztqWmba4qpvq^b@%HheDMA9`fQl5Ote<NnT_eaH-hgVJQ)?4RFvqR-~a<+mfPqTMy~
zt!<<ZOZJU_d=vR$g_D1_z9M`^_fH3WetipzZbX<Ll%6&VnOCm}TP#GVV}PiX(9uaP
zEq#Wo9sr}9KtNV|_s#)^IF^ViRosN8_vMrDM3rwWx9a>9Yt9s7oSH4F)zCoWpJbmh
zVk`Ta?0)FFeT(@vx=wQ6%E8yNY15?d(lmqs@0Eb13lTv0llkVn;k0Dq;%Wm=6h2_Z
zIt-<$NR(ZLIW7n}@VtCikjvnl;fK6z%iiq4r)})yTaN$r<nHQ}fzy|*jyX9sWng3!
zD3+cEi1>#Ow=4FbERm>tKJp7B#<lfzp-YAxs@=NB1B{yWot@)O-#e@%bCedAml1=Z
z;lOiakbB(K)de!bhrB4F4te8B56;4f8`AwTB0mghvk74Io8EY_o}sa^`Kj~HbWJqS
zdw*9~C80x##9i}3(-3*!w~gV2cD{{P3nM*X0|ELk4T+G;(_@}9U@8#uJ0x)su)Rmz
zi~2b}zwHo2?Dld`66%BJZf^(b_0sU8+nal+N%*ivt^5i+&@F0JZ;;C9XS4=Ogh_=i
zcsRM69xKkbWzE~OZ@zx-|3rb_^nEZ6shEXDo|+flty?gRRl|~P1JAw&fj@@?fPeGN
z9PMSG-D3M*+^@^Pp)3E%B&-J*<~E`$RDQB1^jLOtNY5S{HVmD;YsG;GW>;tDK=vB)
zqbGQHc%#&9E{~_+wuBK|9L}3+5P2EDShgm#KT(`ua8=%T<&<E`r1Uw9U(O)yZ~Z@%
z+qo=J3gz<O+^sAOWje>zPX)Q(L+&8~Y70{-E5m1`z~~YeSpjfj&Qk@C{muP>1j{^K
zef{?d2?;^NsA;JUVd{PU@q&f~(Ug>IgY`G-#@cPn+)H1L9A<f~D|x-fCno_2eC6g=
z)4m7ZI1ZJ78ch!jT~ipUM2$~O1R+x%a$lf3kr}}7#4<p9JOwRyy*9fpgibu#HUnRO
z&rH5Qr}*T_6Z2IQYiqDEsH$w{MtA3_*mQk^e(eKmT3Zlud}l4#zetPuRyN}!KE6bl
z8;U%?u^|KQGALaQtdG1P{sY7KyPEtlkzqO`{0x~(R_F=kvX^i*xgQ2%QU{#Rh!ph7
zw@4vEA02h|P#EV4(Ck+sFESgOofORV#{-Dva{oWeg=`*J5E!1A00KF{Hh~QdLI4|8
z*VQ>dGz)GTQm*yw#G*MOST8fAW<A%A_cmqji@imnz-aCfVdHvV#4V<UDvRwqma@%v
zo}WWH>WRG<Dt7N8kUqUcow`r;Re*tPxSj^mJ8h5PK!g8B(|5;nxrhCmWpCLlLb7*6
zR*{|T9VdH}M2M&;o5&{09@#|p-g{@S%<TQU?&o=af1KAjug*Ef{r%k6=en-<VoQa^
z36QJ6k!T5a7>zGK1QOHH&3ub~jG3v?0ic}6YSJrqc6Jnu87Fxzi-$KY9q!+(_>dC!
zP_EcoiFAqfZN_1^ukO$U`tx+qq2NXE9<vJT28QAro*(b%3{mAW4t?PpwZKy&FlokX
zTc=!q@b;bJ&1svwrB$Vxlj1qHpAHs7>FTk;3Yi)JpetsQzzmBp_db4HtGYe>zAKhz
zAB-q|g6Z^t&r{ilpZYPfY*5^I`Bv@{>LmYjYw0-HE)Yw4Ix^g>c1kfG?)7EJ{wKb-
zz5NYpnx`Rhcp9kEk*V<q2#ST=FhBwgK3{gZ54IpSOuTT=s7LgPI40>#N>DG)`G!w#
zB^Ss<rXG5qiNCs8MaxZI?xk!}`Csbeqr-y(D+h<sZ~}ZPez_1%U|Ak?Ce$*YVI4F{
z%Tt8H>jfHs7O;CpMO^g&Xaf;(iH_VEJX0tlk>TMeSc6qsx!OSz(#NFBmyB!pe$X|7
z_(67Y_Tt0C3qO1sVT2N)sX088k{2e0-tWro#Bk<B0f1c)sLjD`6#*&{Ydl2I3TJjc
zzhZWS{PmSYQ_IWBe!-JX8KxXBHIj_~u<+k%V0%=_oT`lb;`N#kv(Ty1H7DG8L3D~~
z$cdKVV>-o9ccEIDpp-ecbF;+XG>>P^N}yS`Xe3I3ML9(>32>ES;LgEp(h?pHWsR7e
zJP<g*t*7ppcfFw-NB9X)`k-=s8tHBR);72CJD{HIEvoEPliIX@Kz+2A-!Dw=%reYe
zmc~FbD~G~0Jw5E!S`kN_*)e$T3M`;~e7O!LtI+XFoZ-Ft)e2TS(2sP2`k;A^l87T6
zV8?0p-_qu`%uh1qMcpq&eYETBt^H9y7gL2O*Bn3HK^f^8Nvurf_)fVp+bKI?(VKf`
zq;aw}9@LOffYsT_;4}d@UG4Sj`*46(5p(i)gp#Hplr=IofBKXdY3oh_jI_45Z)tAE
z;Ns$X8p#}^SX^HI9eNCK4gIvFC(n_&q{_wcJ#nEa`kAs@(K2DOLpRmN=jVh0Mg^LY
znwpL86W_QSSF+s5S+lks80PqqRu-dFk;^N>_Z-@n^aG$Rz5{M1F?^~3ZTb26A$txu
z0(=hBqI~Xefa?N%57FC<3>NJ?U8aH+P=lp{`^fvQET`JdTVq{w?auR-uFrN4h|;TO
zj2Z?Pl8yPva4KEMUG|wS@eC{3S!l{-Yc&cqIrjGVpXG9Ry?%X|RkqXQO9(!QV2}6?
zzz;Ywh-QL?C4kSom57e6r__3=nq?4>bvTtjy(;>r_~EC4wMq+UnSA&9zurU?G*GY1
zNAvU`9l;6Q*buWYhz0<5;F{j?^iBK&t<)N66U^#Y<W@u%N)itquyb*}0IQx~2ZKs8
z8+p}DJ7s9zb@QPX(Z>)G;iR0!Xu^HV=f}_Vp^?Cf<iKoHrzNtBWd{HMWXl-p2eh+H
z)lyJyZA{J#`3zA$&*&h7G5{+&<mwG@2G$(>!vgq5f*1mr+Q^^A4<gtX71hL+%WT{I
zp1kphQF<R2rbi#QwHrnuHxjGNHr4d`81X;thK+5zNZ!^{^M1{-y5HACd_n!KJRg5%
z`V;DbE8gDT9;mlm*7}*8Ur#qS`okud@37#T{A+V_W?;3zzaycbV2k9*)>M>|dK-t9
z6&%ZRTJ-T-hx~o_EzyoKjK*C50?k-rVIO5HB_`Sz0wfY#hm`x1Unbof-)vD@a=6G%
zyf4?JF}Y{LibvarF@=J*qFT@2NS&$i1iWfQ2cDijcohJjID^j#(<cg^$4W|{78cBq
z-K?;~^hZN(o!m}f<q5d1C<DeB2Yf8J<obu3PjzCQmu#?~=qQP5Xm*wnlsj>(SzGfB
zO$6!4pYo5ozs)+RHPq_SS)Z`$5a-$WF5W1o6N_^0haomjyk`3056w>M;y_uSJy6Gx
zst+In3d|DkfSvs~j2<5=q!a#>OehoV16ZMdhz<q@J*eThv~$&xA9M%ewy;;<wV<Z)
z-#4Y)x@rFS?Q6ENCz&MN6w;kf+IvaucJ-_I%m!i`;)<nhv|a!o*f^hLn(&OHpYW!{
zsT6eDP)=IFg%D-;_}CIekx<{TgXCo3f#`ljd5(Dpu2{%udG#dGH-y}ZT=Vh?82Jg)
zh&W{A<$WF=)<hIMHK_3A{_n!9KE|MCWMYz5UbD5g-zpxexObsOY(*ZeSn2dAOOfFL
zK^Yx~{`(v0-}O%P7+6o6@}!(zu50tm(D`Xgr>PeK5Daw=Iz8#z5kzEU%>bE*x<1~G
zA;8DK!@+^`@OQGe;)uVdRJf6$*o0~pu4?J|L-cVktc3F9aDCly4&CW&QwiugDsNyr
z&)To`1|lwh(BTk)E(J^(S~f}noJ3q1!3+yLogxzxm5q|5dP71&zVeyBSKzFP^l@{a
z&$7!~IyoeZETm(rI-p#2NFhC5w_fLO^u!=&C1@vi;T|!(eZE4q)LHIC7DQ%pj{)sQ
zsAXe4O6w5=^W-Gz+KqNz-8{Wa&bxPs*>GU>t2yLFLro2BWct8IOGn2(X8VB&s6X+w
zH&s*w^WQwXc#3UUXVeo>J{R>auw{B9LqK?VpkYudMZBj+Q`pU=qkq)Gzm`in)Lpr>
zqNS}BXNy40I4QG_>E)K>bCqmO<efoUo*+^*>4W2AdG6Rx<g4}dQcy=SSr6B6a@cnk
zNYY&1tm7osGGb)ArI|^q5h{O{mN(!n&(8MaMkD9r50DP!eS?MPj>dLVnx5aW!&kJk
zCQRxfuP14S%fF>3@Tp8xI5MChdlWF6bQ_`MGs8fx)p()Yw&4B3uTr^!eIgFiQt%;~
zwcx-;eCC|(FZ59<0pAD-u>~x;>%oZ}IyB%xt$E$<yDkY=Fda92?sfHK{6t_=jD37M
zyS*d=H9+<K!|uQO4BlbkikaO>;yS~!zDE<5W`TVmbU>gu$Jt-!G=}no@72i4<(UI*
z{_UwaxzHG@epoPk{!H3{Vskn`DrU~IzTjN5ZQJy}|C&Mgj5v6KvcPVpItH|p&hYC*
zMn*O-!nKEtol0Lg=fQq(`>^vP&MkginVUVVZvA#B!L~{#SJSLV6m)ZMMqX<aq%baf
ziZ}3$@-^0`h%>1Qyfz!Z&H3MP7K4<Nv|imC<YEIUYw+*OEk`qb6C<M}QKo)Sw}4X*
zLQMmTz+v*2h&yjfs_yo#e3$d6`G%nyZMZs>_+|l@``N^}f{_yL%aN@mBNKy!AOd&J
zJ*PWvYo=}t^*GA6Baq!Bt6DNVY+JbPS?)aA?I|lNnuGpzVZt(H>(E1N60L)76C-Ns
zq+H`}`1<fr_!sOKA_1K)&q${=ceZ42qk4Ngx%(G02P&358nfQ=K1nizBE!PMQUrEW
z#=e&?fjNmfU$atsJd3o(S0hS7+cyB;PWHWe4WE5P$A8{~A?I??6*tLBySVT}9?m1E
zj{tmvn04>)@X|45w2)z}^Z5?3f@ltXOMax?@MzN)QR#nt-sy}pMGU)AY7-Mbh9PSK
zrm6zt+D>q5dA*VS1CgBqF}1G3F9wIb!*Pr^MFOoXX6wA?rhk5}RO!7!**&WHlgKv;
zy38kV#)g~FJ+)M!Pgn-QO8VyfLd%gB9^~Ty-5C7U83hEQfEf!!EZtg9A=nv%PM8mD
z0P+7U7xc3@-p;=$W$#en=}XlcFg9Xvki_$17GF}luVaH#rSVL`X5xuPm+H^61GdjJ
zyuiEmGVn~}sy)|ngsBC5SDA%{znq+Sl(|X6@}~HCMO78Jm>v~DYxd^N8{h&!3I`ro
zV;j$`ISL6nm7Y+`7kqJA>(s(=(8=B!`=ZD<GL%`D8S`HI`P8ebaPL^$V!?jPoG5fZ
z0(WR<UW9*Lml*p`+*T170Nlc)l0e6TkH;-{Vh1mi#9`L|-Xnk7#c9m>>#~>eZ8p}>
zzO<v**t~CjJP(gQJc`He`i!3Qyn&a3a)VPdJ&AU^udbHxH$F%eL!rsp1K=4(3#tbm
zynu8pEido&itji<-FRAZZGe#b{*D#i#CSagC#+fpdt3H+jk(|31ERIZNQsG2>%l_M
z1C+)<LYadO`ga`bu1H8LtAf~FtJIQSoZru03)()x^s25J!mw^LF%LLg&7mGIRT_Wr
z0b5SYBhNH!Oz9xwZcNuQLQMri&i>4+KZrH(7jC0Y@V*a=ih35icsE9oU1Cf%Za->H
z^=1V|$<^HNtlI4U?@<RCdjyH%tX8S}q=6iV%R@xh*+l*P9vKQ*F|&)P(nmm6VkYSV
zWyYrv7f;|g-2gq)1Rm}9HjV~Z`@n1;6cX|Rnw31GnlC_4Ff%o+2l5g&Tq!Ip2r7ya
zzuc4f%O}!&46~y(``<Ic4snu|d3%+>MXHDOCf^$0<Zs7krKoaCw5BaI(Q-)Qy)7=_
zTRaORPawG@+dA?GCvqU36H`#ggHSmBl@Qjb6(VOvY#jK%BA#)xZ}wH5R(yq+D{bvt
zx!U=SP|Ki*!sd5`6{mtGL4li_`~^*wh~og2Mi7Q?4eGB4jj==&Y^5-RNu3?K)G2iR
zcieJMu(Vv`>7V*97yEV{6ll(ypF^<RikRD<_DdZLVs>?Q-sa#429+@CM!d#bkl`aY
z0t}4MPke~CU<eTG1mao{_1Z8g_J1#ADO6v*ougd0x;aZv#H)>^zkcFw=kQ+O)M<fi
zDdhtVxqZMdneoufat7NM*fhRA$|~sgwzlU9h00ev-%?ZG!ig#dQWZ!VVSfWlpuM$k
zVYg!xAN5?p*nJ<Eo$oEKD*{09psuOe4D5A;<O9FB88qSxmbrwrvMD#JX%4JYy$bz&
zAk7G|ANwB<k+z3d4KN*afX?`IDCMdd4#%HdKA&T1t4tCX9;}g-CN_+LJ6Os6_#x_V
z=<GO|Hng1n5-Dq_W8M2r#4&<yU*C+?TNdcje!yt;d5&vKf|ezcTd1E+HIb2@KN1=N
z6bV~J0|NuZUmt#+1^CgdY;EC)Hvkp=Lv5`D3`FU;ZQJjjOlPy(`W6Mnq#9`p5fD(i
zJ322PjyOqH8Oa(&kI+9W$m}tdzlj2cj6=T!@~ZfNB7$^<U~w876qhtGJ>3n<=$4$9
zFx@bQYkGNk!EJMwoAY~inYXv^#|fe|>we+4d*~O%gTF<>N*BaiSWLvZBR3)^Nt{z6
zx<C36pl9djFN276VF$*c?1F;tnVFbySP@nq?0ySw-7k=VL<P8lfoVxc^wHhi{M(um
zXH-W~Ny)d&s;|I|Og|fpRb)}PL!v(=wlmFjUOe0t^2nj7X!E(t=r{)(xc%03b*%Q2
zKmXNhj*9oQ<c`Or+dyCwtlB}L4wDL;oSP_M1e~W|mZPMSEt3Y^YF~@x&KM{kOc4sr
z!KTvEe(!&~$hRM0%=}Cd@u>9wc3E3XN0Tu31~swc-Dd%(vW&~EadOyP3H?ffO;|dy
zsoTS~A*qM7Zkc{{ZtEnzXG1B?kEcOtfwZA;B?uYU6bPQroNiw03-yE3!W~{-BJcu-
zb0Z6>USLHd3iO8+sQtAr{2N13xKO5MQp2Q<L#2Lv|IQDDZMX>9LG&;_9tUeiur<ia
z$$|R|oiJ=W;iKkBEy>!q*j~~TXEGLAVg-JzUaBEBEC*^^5!?b=bMLuY3QYfLs+Whm
z1Nx$#isQd8uq)kV!@V_2pRnA}Z~82?@W&&CNT=a&&OtI$A}upB5ZMxtlRtX$1Y0AU
z47`Y*zQQ#%wUNH;Pjzp!`|IwAm&wLXjus*Li{KwYu1+~I$zM(4K7^SL^WfV0`b&X0
zR(%5lSnDB6CII`QzI^#kEMlK$+dY0|uKQr~iN*b%wV2Ix5i2WuQ-cO~ANG@=&k0`z
z%X36l0BeDYj*QHpD1=vS7d$GCY+N8suLWKko=zs%=>esPntn`7<1UxI+?e6_Hkpd&
zoO^CfKPa4MlAgN!_!3fOIT7yqrAy1dyIEI<LJ6JfMM#gJo6}!je*X0Gp2Mb|9+l`J
zH<n&#?bP@EdOJH807;-8V~kOJ>c;r@{S8k%hDnwasTv%OPanPxPxrG6-dS0Z-M(LG
zIDU0O5`+O)Wyo>3pC2mBz&6d3O-+l*LMO0Sn@?r-<L2Zf%V&V?kL-KA=WwjRRb#Cn
ziXQe{5R3(V&gQ7)^_W8H-@kwN{{35kZF71^hDP>MaT`61@pNy%cls9)|6lq3ltjZn
zAYOAAkWbxFO-+c8NBr?&Prf--5eCbXp!0HvnNCQkp@e1+7B#PuMy`^vkPS89*v-w&
zMIg(ZXU)yWykf>!Z(-*dr^Wq)^F;T-z9=US>6^5gcN*AJq;m`j7=&vruBPh`+PfW7
ze>VJoE&xRaeVeMR(mj%(0C?etrl&d2ME`tJvj*#)xV$f5x))){7Nbb@(B}b+aV)NL
zEVG?HAMrA)q#3)`i^GX;YT1(#WZ$aZCOtK6VCnAZxex44n0A*|S3eQPbo}?3MnqNl
zFT?|s#-j%sa<hqx8~iE<hG|)J3lKMmX=y(LWD3eB7N1y!g;TlsC7Hk><-X6^1kd8;
zMU(CxMez>QR@6t(;EPO{xV^1?giXu+gzQNPTwnqh-xo<Yy5NVt|K_A9e(DC6mb(_F
z2*88zHfGm`0w@F6n&F!s=%xe*@whlSi=gEMdNB3*jfAfxV*l}$7~SCOb$=WGskK%L
zJ0sd~aZ2EHQ`9%2y8JJtupK3Qf#~RHs|OaKPpz!1M0#<EYiaN5dY0Sxj<JVNNGLNq
z`-#d;0yO8t{ryGgJe$UHL%XAYZ2z6!xf#{y5S92w`HdRj=|L<eN=mpOYl2loOiYZ;
zlF9PRnfrz<jp^r)ha}N1op#G|@05^LoI-b@2i#*sMNI!74&kqW@c*dbY!X#5Q{!Ml
zM~Eip&C$(x&#Mir1CX@#LOtsRJMDT1v$=CsmzReJ^bpvixA*qm+Imp)h{5m18J1fG
z)R<Rn{fN<u?`f9_(gXduzN%=wLSk^XrYiQqEC7j$ap#lyQ0#LPLkyZwfW<)jD)%B&
zor->CzTuSdVVoF!w{q$k!Nuv0(+1vwz`K~jw_&zBDJ;KHZvQ5uCh(qS{P!X)-RK$a
z!)9}`6b1V-DZH}MQUpeYgYzG(<&7d&2M0AGP*_-5gZKV0j@POribg;}C89(HK{yor
z@1><=UzJ<4FuR`6pO>e;tTWM!m0fZ06I`W#`(5~$*qe0sE4I(`Ie{qDFDul-h>aq6
z)X6<K2JY>rbIlkCTseCRtTVV*w6wN*oB?G3aN$$)@W8-R`_0L|-0Y4**6bxL>__Jc
z?2~OhH_~tlxJFn_>pd*QdQ9Ihy3u~#65>`0irla0E2-f1lYsw@M~&j$Gx+nb>!R<z
zstW9Gt~R`SNu%ew>!_)v1(1KMZDB2B4+ucaH|Z1SHWo?PH$&M5NBtkLVStNBD8Do;
zjIg1h0TBklat?@=cNOWNw&8vtSh>%j#^2>g<hzFx^r%I!94BKEirnW0)qqMovkgXh
z<wLQFxkHk&9<B?4MLb)VKId;+B~ZJlz~BlJDt;6dg~KBaiI#&aw!6EV7kDz@IR5je
zdr&w4E0ossl$7n@G=ZF}yX@>zFz}p)sW4h856Y@BZW6erXjByE=H)E`jFaB{<OMEv
zV~#(7^nmCNkBmG8(*&l6&Q4B5nG76e_KuEWa<|u+`w7v>Utq8{hyPp#ykFnR#bp6D
zp>x?kl0D~)>0<h>;05dG?Y%J<5q_%d8r0ly_VC-iIvPLVlOxhT;H3g3kXKmy;R)1u
z%Z|u97#J8}<`m&zzU(K7HuK(#wg38T?6=3tS4;YnI1_O=zkL{D%*^lLJqMDbtg$gY
zxJXgAfo&$_v^+L4qMMkQ0PmE%sXd=ujOz4s#JQlv=hAcU6V$Rqwqy@~P?zDoubK1I
zx7xjZj3(!eEpo4+k50RevpI9-o{i0XpGiXp=lotDR{-f(|7LWjmoHy{-3U@USd6!~
zw*x+WfDA4{7gyIugA)T~js5-MAb<auG1%9q_U-l%SZ*PXvEVjP<#Q>vw$U_9!Fa&M
zTU>WubuUIeo<}oo$b8<+rlx4P6Z>%hcE;-yjE0p0{Fzg!3c_3#U~<3}jJR0^fBH1S
zkq=cE75&EC&Q2e6kxxLZ7beovE7^~e>HnFAYV+=7+iW9-5QEjRk3+XpVz1`;UBOP8
z37yguD2!yIWI=78OX(Ywe^7P1CH6XfnjfwhF5y0=x%zTR=pZuF+T0ufUf3Z}EVLF&
zhQZAe+L3lZG33BU;Bj6VY?dIY4OuE9I$>zGiAhPBCmTTfMFN{Fz}w-7qnpAWCj!+R
ztEeKgATzKjumynG#%(2?tSkz*A-JbW>FltbspLNHGtzod?I(Ga_#cO6IL&oxktyB*
zwq9v@c}@qH)JY<6pBN()*zev2VwdhP;B*iItoRaE+(NgD#BGntN-@r>+Zr*BKbw#M
z8fys*gDjO@vBEe2-pnBopuAid>gs=%mbohbq;&7^?(NOTzYB0hr*3GW-BgkO8*8h5
zF+;p&u)MqYkjY1&wA>WdR&e6_;1(A09%LZm=<$K?2*gl;4_H08{~O4N8P~)2&FUMW
zk|16KEfVk~?!Vfm1=PtB1k}`Rz%_$;X95H@pkkc|hWPeX#~GGjzY^Na-ebTnkkeVA
z)Bw<$uCA`t+K7)W(%O-(t*lr;KynMHL|}1l4ujoHXbCu8-r_O-40tzmh+EnQM+Ue1
zq2?p^JrrIub8~BJYvOMuO-)UZ^Fc~S*ef#q{3P3By!z%<)uZ;C#{3jtB+pFIR02ur
zCb6cD8QlZ!i~pPW!0^u(H(Ba-L_aj|5J7PQqzd31^q(lZ=dED^$m+7xZmeSl_k-@q
zkFw^gMdq~NJWG@z<M$JeZV{s6w=?j*x8Y2MbZwm{%Hw9dmZKbh?UtLit(+R8cxSaQ
zMJFmDOGZ;FYcP~b;BlT5Ksey5!oDrS$48v|7l3>CXmV>s#7WnB+4Jt+PP1Zg)>(MM
zC4ToiUQTLmrc$^)W68}x-kIkdqc<9t((B5SeV6ksRYw0*j_3wOgc9@gdiqtEYnwW7
zQxd>6_Gx6bxmgYrqgI=#%<hn30?ODtsT5IX8QDn`e>Y@r^kU-Y3$w(7;$}HXEE5@x
z2*w{+?xmJ5>M2_%a7I76WJV5+bp`$Q_&P&T7;ZA)AiHM$qQfp$8ft4};cp?0pn!kX
zo<BPsBVR7EvMClsX}E;MtBL{r#~$i;^Y1-a8$&ab_dZ%qM<*7_kAPJ_BV6DVJjP2f
zew)SU-;My__FkO@IFHs3YsS4TFD)%vDdgGf`0=;$DO22}GQPjOY}%h%1E*u0SIqrM
zPmPULVZzo;+x}&SwED1i0pa-)fP7}1X3oyVXHwPryzO}Jb+G~_C(sgmq;}@54tzx*
zc8v9+^wxriv@5CZk>!A`Bk}$u?(r~>^Fu(U-qg5@152IkPrLl>2$H2s3|zMS|4e@~
z{cd+Xl1u#qF+;G+>=EdD&;(^_Xp0i4M}Ykf0nQD#Bp9h}pVcGnXQb5MjSZW%f1Zti
z=rogoZ7o#ZV+>E&s3XV@7-}Qd#q@1YIbvWZ1r9iUKxb;mEv%bdcK(ft<_+O2upb5(
z@143gCfF@t5fUQR9U*8vV9n197{FbT`De{5^#0hyf51qd=9pTiu%TiKQ-m;6!xY!$
z{^z~ttCNoEhSeOAg};6sVcEdP0260+A;o_r+S=L(XBPmY^PfCtzN>RR!=9oiLe1WI
zLL$-@7(*t0Px|UxqGU4fyAlppXpOr4H+Nr)vAd4RbW~jJFDm~X`3KxN#4Dj)=U={j
zu@Re}Z~9=X{129ckN$y8S=M7^Xj>o2KvM7LAyh9i%nb}R+8-egcDNK7m(Q^E$<Z33
ztUJQ*Etj$f>?nRM|7=>rXd%{Xn!Y8sP}scxVwBoJSkpAszMj8t)uZF@>oZIPXLsH6
z|K?!GTd>7IBi4>+O-*Gi#xa0Z0>pQ;fqeF!q9Zf_IDs<xUlI3{Eti<R#o~)qllz`!
z(OPY4E_u2i1ydH5>O0@^eWuUiZmR<PW&4xP$KM`-<}I1+Qw|iK`uZ>!Q&)dHiL$nZ
zE|xJ2|5d|b(4A>vekSlNfL90hnk_+wtF4+~_L1zdfDN^k3ZP_?AUTykNpW%QaJ(Na
za6vD@XBKc0kMC01`!$M~R`tDoO~94Dy?xv*ynTy@7gEt)clwi@fhw?f83)rqIW_R_
zSqg5Po}K+XQ{)Vyy6aR-!ae&wS7_r4BT0~ypka4IOZuhBkQZC<@7HKXFLBEY-i`01
zWfODeJ)hs_wPiB#nwk{~+I>DhswG+{-CI1JZzoCi$hzfqh6b0v_pN=8-J5ttmGUW<
zHL2ogNy*8*&_W_SGl(}^$S4uNIRA4E%|S8~H2T0EdTGSD1zUO^>;tHH&C3;cl)mi3
z{?|P>42B~nUfx}bl=;g+D8)vkYtW2=se)f%ASQe*qgKXOj{&uV8(vUEBq}Ou=!H~U
zVGP9W4DW)}s2Ey^f%0~^>@jg~AznA|Y1SBpJW*4#>@^S5(9p%8m&N=fmV1-^LriqT
ztEs?o$3T+AP1AY1`N!2h3E_+bsEF8=PF$OXlH*4(fL{=;+yqu>m9Fw12kBvw1I;iD
zj9E(eRdv)q{d33*O!D==^mQEL%Qdw@_epD9;d_*&eSAZt@Z%SEeM{1aKOE$*^mKKJ
zz(dP8AtAw_FTPe4B8LpY)lB6&Kb{O#mr|d1cG6_`HT{~Jk&-u!^=#WjXUT%I!>t1Q
z!Ov7D`lpOKo%fw3VxU>r<Wf#u3-u)Z9;*2PQ<mO5xy#Un!v(=1vqhd|+uOYoadvmV
zRd^5FQhNHTp{In&2l)knLjvCuCVxKAPCz(K9GGt^W<~-syjKmoy6-E=3uX?FQ{U0~
z5tLgg-O0P3q+*lpv^8{7{HNI_`+s7S|Hb}cId4@#%C2jP|Kv*YCw2>d31bEUG(siZ
z&=|wT_7M&KPRHP50%&OwYeC}#xF*4I3^5KxrV@wgEY|zMASMLxJzYcuE2RyF>7Wb&
zX6A2ACmM-2@9w%`j-~nUFZR`>{VShCmGF*~%IMja6&fL@kCuA%80zdKM&V*`P-w*G
zTVXzPQw(qH>U${*G@CM~)Yb12{q54~pSB*fe1X!tVn4DWY-Vphyl8!Gf$er^KYoON
zaRC8HZ}lGfiGFy1>e(1<>X`nE;g(pn6FZqpXyQcUm!<*3-!F-?H9vLuDCtoiNRF<k
z5-y)s-|p|}>G@219}qiQ@WRB##;*D+3QrE+LWJ#hb=v3q({6&l=N<GEP7Ks*AU+~x
zV7OLF>MI1+mwuE~MR_?tu%rOyfL-Ld#3RW<li$gRD+|;XK-4p&-F;Lr^g+)7t|X&{
zCVk+`Drd9;<WXeagOvkmY;%o2fZkh^WeA=lb7yyt$ipFGuG#rvr8~Ot=3DO$e~ji%
zG(*~X?U*^z{4*>imQ+pHiRuC22y6kY!$vJ(PDgkE@G%fD+>()m5(Zu3od*vRAt?S>
zLGfXa&hF?@sqolm-NQJC>|5+hGLOWT@Z(P9c`I>UvnzAxJ2eIhn7&ZPk%-+<&*KMz
z!3^p1IvTiCVQqQ)_5IRwBJwhF>b56$G#=}RnHF@>LZPi67*Ffki(l%o6Frtl$+&*V
zLCq1am_f}?NKVU17e_0m$I^Xtw4o9NPZO{c=fQ80w(y`e5)ZPL7>r=am1#ucvPRSv
z6&9@W4uyb~72G4#4P8HoCoNHMh$bW~*xLtIgpd=kw$gHQu6(%IFJ&jQdctD%LjTh7
zPD9^JNGg>8u#j0nBTpK&{c;#SM_6=`!pi`g`G~G%Wd90@;$YPRq@Ce#N(9~pSQgOm
z^W!bXYrCg4FDq~1QQ@8OS9)M>5BnEzcPFB`T1fNUnVDI5j&w_~Mq<O6I%<1r;$GwZ
zqnO)c0_(rYw#VUa>~ACEk2nrn6#WeUtRKWvuj6D?o0m;Ckd+4Z4LOEN{mJ&<x_+O;
z*RM|a{Z^4FRe&Tm|6#|O?<c8TMxD~g=&~RN5}l{ouc^4Qg`dgG6)~<!<aJkHaj$d1
zd4#3`K{9~zT?`8o<1j5S=LGsTOdt>jK}#!^WB}`fZRQ~8CKDfhx>XFMA(MqSXS?@W
zEmOQ7!@v$h_iwg;ctrH{y~rQ|wuyQb4s-w`UI+I6yBZjYtV~Lx^F3)|fMn&nz%PA8
zCH+ys2u>3K*R8|D(S5(Kzr>#XSyd_D{I0Gi8P2Jh`0{1yqq~Akvw38K!v($mG#?7Z
zJqTW?B%u!U)z*)h_^JQ_2Y{X(_l{oRaw1;L(3lh$R^2u}htdkUkGL@MFO=n$I>Avz
z_`(1q_)Jr=39h{j|FFrYXN5vbjPw3hkt6=*6<)c#G@um!#JW8)FAeH`YAP^kV*YR&
ze%$rB!Sc%-&lbJ%^}OU}{PCP0*$lw>5U}@IlOq#!kdv+A3zBr%p{aLcMZT$Xh-xK2
z9!Sbf<W{3-)Vgtlw|y8Jn<SpR$L>KzP?)N3XV3ctg*zGtr3Z524~R18xc=Q*QuJ%q
z9?==-?|%!9nfjJ*;MW3uumlJu%+TXwV#u$LeXr#~#$K47FRdFyDdIp2b~GL}ReS|i
zq0Bl;9kqH5Tdk}G?nV3PekUB0x@&QMm#6HaZ=K80*$Z=5zmfhp-95P*zwF{xRyy2X
zg3&)VkXEO%{X?B2JFnn-SsD3PWp|w&a|l-iprQ@JBPx9^t39@A+_%htXATTp<XB`s
z35A_Pb7@Ot<5L#a2hy|m+a>PVt>R3!jYZ<PRvUSl_MyfJk*~ae+u{3OZAKt5F%h~x
zHYf*QrVGa-U!wf%WXe@OIrb<nLXBH1i?Vj4i4#=tkdBsmR4bJsuAN)b6{vzq$3^)c
zPVX(U)8PmW5dyFepvT3$emOi^A9~oRZGzisv3sCIRE}}ro22(wU&EgD{L)WqhfSgT
zESZ7iT25wCjKHFq*M6y_oAAv|rDrO4=Pu_!Ap8b2gTQ~<1wG>QkU1=IPZ6TqUJC_M
z&Ag5_*OewV(T)xMl0gXqampFrzdvrev3wj*hg0f_c?OU>%%?Dt%7Gc%<WC#NwD=LU
zrInQ&y$_ca9Q_hpLEvO10MNKGTM+na^7~%BfhPU~$TYyvCiBJHm~gpq13_VI&vk~y
zd9#S>-lrUXOsZ&lM=vRE*W!;n>9GE7t^23Iyulw6zB3Y<0@WI{BaVg1#%b!|kUkF)
zk5J}o#(xQe1Q%pu$RKhe?rknW{lF5((_)9pv_J6RTOH2Ph1NLr4i>q!fk;*@=5p+0
z*pNWLb5t4jPqsvO)J)<3gr|D~7kyyQ4xAJwkU5Nx>+%ty)@P@Xme^rhx5iz48;G=W
z-adNG7k>_vk(~HRI?}v@eBS=iRZnlK9+$)5ln}p=I5W}brYNecp@L<?&KU>nUY|%5
zD$7&;+4yxs`CHuB2nP6oW(#Q6aX>>x#02p|e+zn1fhg?nj~2ok8LIMyF(Ao9&tSp&
zMNnD+fmTRJVty9OmS+{^&vbufg9O$p!*9eRO*sstTU%Ier7vL3aEqS`+%Gn(+yY^M
zbIMwx#t&&+YJkE4A-=?5-3m)L&pYHA95F_Up};g!@$e7@;vZSt-MiB=`u|96!V0U<
zU#~K<nJ^CD*|;^-Nb5$~jf<)#A}ugua4|@}7TRA_{Oq=7LP-}>Nr*bg;g|TL%1CWe
zu1qb{eR5vKW6I1F>l<luDGyJ<sscRaKR&KrxO{Vb9N<+TH8w+>(=&C=b@8X2Y~>-Y
z>lMnG#0UJ#ihKP@100SkrmYnhaVM4>IiChb2VbnLy?7w(jCqR{-OI;`%`H!=0{G}~
znS&n_f>FUGTD!Wh!%l{9GnwT>iQsO5J;l_kBdt`wb4FlcjF`DL!MwGn9{bc+1)A9{
z41MlcL@A=0DB1w8FyvPbc!!-M;hKnZA*`#w%b{6s%QsT$13^fJ?;$7B5<unwBX>|3
zB_FKBKGtvty+YuhBFKG!mB9a?myk*dM2W4ZvomQ`nE$x3{Mgw_<{Mv>B~XrDnl4yT
z&Z8PPHDs-J26J@Q0DAzzY#=m&TtY$u*nLe9VKbG>0X+r_M5ZFChae=+el4q%sbOej
zguHhzrMo=XDsJ-YxIW9`U1gKrmag<kYf{BNc;&aJ_I6d!i5^pbM*QsqPPCb4pO4|m
z!1e<nC7joyM^IP^xBJcQ89LRxWxy>jzTQCXosnb{Cdo-H%n!4C_zG~QrKhp4$EjYL
zXbsXADkUB{)Lr~jf_7=-SA@Ntod@V`;3*LU`Zjt}2Y&!Rq)rGsKy$f5OlWwNyi$x(
z#>SN?J*_}bVtYoKku*!#<8uG*ow12T-_YiCl3?5Ieku!oDpYrL-8r#ns$SK&1F1EG
zTii^=OJWER1A({3vyxA-@bJLaI~pW~5QRRxMagl%@;~kB+8j*=gYVBazMLFPnWPwW
z>#t}H+R{q6)9?K&euV$Gw#jl)Abz0(yDoOW+>x!3^*ooX^*MnVwcq(-GvCwbt@2)P
z^te=N1X_M2CgsR$@9y5*s~wKeB%~HxtE-YK#CDTSN_Ct)!yndum&W(tOt2fD;_46X
z#UxfK4p+Vkt%6CP+7kws-0x&Y_EZA$JgWmij^+;2;whJ*7~oVn@(Wy6z-Q_$14Lv<
z7NT$)|N02=v_MAyu(a&h{D9bMP7>?Dz_IobTvjx~cJ;7L1TPymJ>t`Rr3bJq5cj+K
zYjWKQ>+r{rA`bf7eRQs6b}}5`{cvjEhh3K6HJ+bCmj9<9C$5R$g-<d<y|BABytTa*
zSveVpkN@jm=e-N~3dn*LSWbJ*q+43N_-HeOXr~mIjfr0gUt>1zh{>4T7(&Cy75_9V
zbda5TyPpGW(xX7u14Q;Fh^;vVbx)1q7kdW>*8BJMSia=kU?nf6h1^!?$`C6f7HMr&
z*{b`(w5*({`+n*M*p^|Xf8)e$(qA5!wu?H^qsQC{FDH%v8+fCsh@AKJ*SWhh`1ApY
zZ1m@=9b<6i2ZeNDoRk$%FknxDkY>P^H~;78K&xs{<h_wV+8Uq4r9pGR!XLl(u9|d(
z?cy}o^lk}#L)oN;YILT?*x+oXNPBY-Haa>MrD!tzZOw5!CFeVbyWwQV_eh5J%1BhK
zBTpcjAuf`z;6$8pNo+s`1;fyj>#Iv^D=S7gx=jQH<|FElTwsSZc^Q=uZ>2v$Z!?^u
z9<Jx}E+btd)R2oQ^MLx0Ya<F(!k){pyQk9q$Cs&)jd=!r5^SH#{Hb^npp=TEFwYzX
zh4&7whF9a{szMk7bkM~MDd3S!zdAqve{t|es$^o{zAY&TVgIDF3cAk1VO2p%FuwJq
zFs5O}KKu)65;u+aSM{VLCrmXR%B8IB2dBBW`;*c%WuCwEdU>W?Refu+Qz}b2H)r^A
z>NlyFO)5a9utn2MNPPivmjVk)n%qndVD&7ltc>1ES*{^jd++&ihLLDBeo7|hbl}Hx
zV+_ifgz(ET#Z)aT2mAj_CPnPTf5vx;j}43ru>H6B>4Dfhmok37=bOA>sT2vXTaZ?+
ztgH;Gxwg)n)wUUYquJ|o7^_lL0+NQlRf7{oqj$=1Au7yOjQumcQVb}u+(wOET@WBf
z_wV1oqO!8f7OYYkf);D1c}7DCd__!!vW$Rk942UO!o$PQg@<_9*%1Ry_;?n#w|n98
zrj#_C;e1$eb}Kod&cFMCAMdL1+wTM-{G`lTxi_n16j%-U31$2-rrw@QS)$=L{_#To
zZwrfsjg3zDzae615q$AL<QNP0Grk$PAZSMDy5kjESz9v!8MXg=Yin9~1y0W=esk$`
zCU>Xx`tV|Z+9nBI#y+v1M^QOdsCy}mSA_vsjwPtO?c%)adWVaXPK6j}aE&bd$B4WN
zU%~#GC^s3f`wi{bglivyptdNvj@Gj3qcUG~d-Mjg{7iyuxz5HkT#un0{6yu+J5RWR
zZ~eU2>$Ew^7pWBHn$8*Xe)K4uh?$O=TAxOa?f@9x*s$dT{||_?gTxFOh|>U~${Z-f
z0sYOrZ~eB28d<x;=49{qxNU6gxS(3=wYLOYO4wZc2uY|h*Rn^4P+6<A>sR_RF0Rsc
z^0@W&z1DGJ&f~*I1)ZCL8XZUb6nZZ?WBfZgSl(7Cw{EpQ9tbux;CQiwLW8ztkL^}!
z-Z-`Md`bLgmur>FG7iwZp$uoMklESU#YuPxgCPLG8JOtrwsnle6hq7H&x5|RJ|A41
z$BAOCAAidDH;_rwTcUY~;$Q$j#6MbAP3Oca$J7*M+dg2jiTvus;hx4i&yTL1k2h6s
zYqHP(+dA5CbH6|QDz5TagrsyxQdk)$xjAj!cf*|>K!30h5K=f??us1=9U2&Dn3*vO
zL3lH^f|2e@sR`7K?%(E?4UO)u#jn1{U79$V#lhvcn8Yr{>EXiPR59op%^1-fo00l%
z+rC-M!}qF@CXUjLuI&Eui+;XvQX(QL8ygOTayuldh0qBTt=iCU8}?o%(XSAAm$}Hu
zBN(7m(1+(_`*}*eIbDvufXMCt0B-=^`9Gh9{tt?&juZ`nvg+YjA@)?H7Dv2xIKOtg
zX(oFO)c!vgfOGHqz_0aBxjhZEaL@#QK(=XNz8C0Vs0%a?SbkB=(1y}0eJ`$)t&YD$
zadP#G7gOjXW}&@hu{%r~C-KCJ><x<BBS;e8xCSr7cA)<ZjEq2-WFQQSe@Kfrt(d;V
z#VI_tVC_`Z*K<2;uoVla<04wCNz1`oc<6_2BCEu#JKo5=Bhx=5Gg6HYTXSe3VR;s;
z;TM|iI6FI=F4)x9SJ<^-0ehVIHaB@(K64{;P>r@ITc`s6G1w*;b_1N0{3#-r#+F6v
zGs`zAa&S&=yW-d29;*)-ek=9`H8Bi0%&}c}xwy!`&1vxp2sBMi(Ev3Gq#jT+boPs3
zJs0f(X-3Hl91*}8f_Y+Vr|p~kh>Q-c^5?w@8U=IPdmnH<RSivlu2K$r7<#mb`H<&x
zO`Kt8jAD5**IZL{6q%-W+MNjAfreXNq~@^ga#?v?;}uj8N?<OgcQ;4oy;wTv_aPkt
zz#drjgvGV9MRfrAiUaUlvln&Zi9M$xwpV9(AEM@ErgVbp@LI#j+YfkS9b7%-hDT%#
z{fmSI{nMB{s<1UeODN{FKg0rw#;`|mU}ol$B1%y5fZZ`r!-`!~=+4UY+xJI}x@y|N
zGimTGK{_pDmJ}QN<Eayv=V}`;z5X>DtUpAMQIt*W?s#hu-BIUla$TyU&!!%)H?P3X
z9~?jJwBoAJTDIR7UjNwOgvx!lm5aO1?Z)=NH}^i=Ilar%^sw$U5vKDbaH|4ypwJJa
zCu7AoIe2&o6TVeI;q}I-HUNGLakJJVgyX{xUP~Wl1MkIleHC@S4TKgXr1t{#DWP&|
zg3bKk)3iy$0paLSR8h5nv}g66Sd6cWSQ-IFLs*E=XTDQrU3Bz4r-8#@_ManoUyQF|
zm_BR|GOLiD`^&1^M<+@!MH*lMV2OtdL4eAwP9s9pB<xZ{?p0uZnhBLN7jJ@A2(-GK
z`zt->5Gg?|ay>bzUq*zGfy;<+4Z<SG#BaPaU|ge;-UxHDd*)>o{EZOfy=;AysT;-b
z-KjCgBWBUkTlxuUDH7e@46NLk_A}hrZU9AtNju~Jl5Cyzz?S_3ypaf}PQt5Pza$5T
zPL>l^s8w#+wP^$G?O47w&Ywm0*-YO)eDq53L~!xix%?j|3CjcFrKht8UP(+A5;Z@E
zD@IHHx^U9;*|RQJOZm8!x$XTtXqM^oZ;Y4W9y1*1S9`~}Oey(B6zBzr_y_Xk%lcA)
zV+1en!e38N5*9Z&!hr|+3S(?nM_=FS(U$wUfVhDHYdjU{nrT-%sxSiJQvBeroEZ<u
zPUbHB{BJpvSTgQt;c`4jeRpUA_X4uIxOKF?amD;>9*SA6Gd3*9S%LZ;xicJP(0^40
zAw1^OmCm>C(+F66ZpBdtrVFr}m|@lgE@!ruFhHDKs$`nN1{`j7w>hhBd<c^nbe2BG
zGZjbiXOrqbsrisIJSx3uoW?OT&Mx}08ynSx$1ia;Ine*!Wewi`Q#~%gjU`KWx<9C<
ziht8WrNDSako!w)<EPSaX&)c}TFk>q%p*0jz=s0gA)5@bd-p!U=Xr?U!}VN!YGWo)
zz`yBzGG1e{|7jm~axWq3Zy~gN8FN>zy=;@>#Wl8R>#Sy}zJQ`V-~W0~uA9_ZH5ZdC
z$a!rNDgKved?-5civdaNM7*s|_gD9aF0{Nc;!QqE`9ggCuakvOgkIy<>{nyh2P)ZY
z&;HcFmI&<I)S42Gp3eaz5J9Z1tlFRbRn8u&JtoZ1Kpp#uIcONik?|NE@8Eq}RD(6G
z)hqqRS`#!QK!||nB?=};fK&lY7X{|%+5@g`Zkja%g1TP?F8afEhv>)4b=YmcQa6!4
z=e4xi8I9dR{c-u7&oI3W1aF8s3+yjJ;DUlyj0b@TpwUH~_Q3spZgH_`G~WO&$uwAE
zMMslG=t96sP8a1eoKx`Mt*^Fa-IC_uC>3m6H*#XR;c`@Q=gvlbyF3Q})_s*8BUxyU
zi)@8xuO55w<tLhMpX0@q2iS&fsO#%10=E%nAX+iEkKb9Pqiml)XA*k=+CL-nue#*l
z@@|8Qwp2urnv#-;h={Sy0z??8&MI`pH_i0C&&>ZhEHg_t8m`KF*f!JP2{4Rqt>1g<
zlEWF|yIRP$-RVGWFh#3WdT4EDFZbkpEMN0>OG`Iq%KWTH-;p#Rvmo__h^Y=(LAM;}
z1j+eVR8*W)08X54r7|9&;0QSpu$RrsBvND{+1%Pfawq#9UO({J-qx(!_q`C<$4g1+
zO#PWg4BIecs`aL>Z*O>O!cg+G1O0nmub%QQlRX*{T2sj6Jak#ya{fkn@XuEBpyU=?
z9FW|Ao38q1b7o;l%Yh{)NXEt{NAxK$Vn}pUSXC3-Lu{vn3`Lh*+C`be*N5~P`v4vb
zU{ec-s7oDcnkrr)*qp@mI>PO|C-98v^9y}m&X3`)^D=&uoq-!AH~h&m&Ba#2nPj&U
z)UOvXs?lZ6PK%ljmup%(UMaVgmTu~6_mcyC89G;zOokMCQB_-?_ok1=h&|8dHgUMO
z8&@`OsLD#OUVXy<tp(>@c8G34^&%vx?TVoczk)CNE5GH1Z8!Zfzwg5fthQ&7ryz>e
zg&gZ9Z?gc}k09uF2Bltk?bV$y?MwcMCF+w};x&-MhbJWDKDWkb<lhM<zvr15btSau
z`&H^*%<e%2E6eI|5vWQb9<<72vr6)Gi4@`bzE=~xJr^=r^Dq@(=!LqILb&yCSw1C=
zK~s1~qs6o==|C-!RiKPC@Opf&&;P(f9r##U0UJukaHwfE4rnmA{elF&HxPrPKE}9~
zL=o6HZQVHeuSu@OUNk~ls#j|;GW28fAZ)2Ya0e6(Bz#!m*T%+%2k35qA%F`rEW)M(
zgOBWp!rqMhlC3z%$J43Q!t0=!Yp+nB)eaP7U>Bsj@xJM8nuKAW*2$@?=ui5ke1(v1
z=TiP>t{Cvj%ty%6Gr=yx*VngH+#?*8<yjHsp)s$%D!>2z`?rVvx5ptQxv2m_fWku3
z<I4M|QXk50`<(hL{T^$<tpe&F*>X?zl1M_M>7P%v&}IB-SGT3tlvYzB4D;FcINGo8
z8YERyQiEBse{57<MCyHm*~=Lh5aCez0#>U41ReduKeMy+K+)~&?3AlQ#|n85A4i7o
z^%Y{ik(k(BX6x@`0JWFx=FLt;#VlLv1_hlDv5!x7JJ}gHtHZUX@9C4_2(KhB^W3#2
z8=IEaQTk*_Z&F{A@+KIlCmz$#y^a_s`1`w;SU|Vu%nyM!pHos&CjWXaw&J_*%&OSg
zaRM2JA~yr*VHm1zZ4=uVKC`C7feCs&bfWQSA3|lGCY-g5W3iq!Wd!kNeW3lcL4ER?
za8m-Sr^nRc#z1b}5^y$1@7#rl1eyZC3u*5+<{NCSrvk0rE=-e<03lk|vJPKcqs`n`
z0B&Uv()f+J!!1^}g9DBSdqTK)c$+bwp1QgUAkhR}4?oJhUKF$XTc0K6M2j{SJf|TU
zeJkgof)XyR;UspM&EFRS^NvYUx$0xODu$i07jG_HH>h(5hWJvmGn2EQ{eYysI+N}a
zy%;$=2U8iJHJyVYuX@4s*|}r6RLHMDIJf<<qxgFTiyEa*P&N{Ej9)LOB`x`)Uw2`6
z{(a(cEFULmO#%|Ri%0mwz{`h0hmDIXq{Fc(m0pr9Kg8({W-r6rpAY6=O;*nd8utqy
zA6M425__`t=L8$K<wSBAQwGqyR8aWnQ34{O(UbFGfS-}TCwNrYBi81!OfFutZoF&v
zmD9)6IpznoM6a$V#CM9=3y$#r)VL9?8)2Yk4|2*q3cmAY-0X8}v(>QR!)HG0)GBsC
zd?Pzbd?Q(W#m$uFMRMlpS>xnHG`EVBNYy=OE-#%;-D6pMboq~8Zeecra#~KjSowPV
zYxOn3r>VO=f)$YGCfGv<mU-CW;5@6()?kx{pPp$Bw+Uhs0#c=f*3v(kQnq1-nu%G8
zI|lffT4+NxKZpir<lhC2`c?kYtbE<vR^Lz-o*OhS4B4h2PyPii%_ryQO1|gNtI|FF
zzO`#=Wu=>IU!V0jQ!8H^ak;hVP3%aMZS=M%I!9Ojdv+tqPR>fcDko83tS{1yE9l!q
zn?kpkjM*r0Fo&Cc9bWW}6W!OO>m<@2?RovmlXidZ%9!ZA$}ZtX&6g+ubtmGpilj+_
zsKZvp7g8gD*#izt46q61t@S*NBw^R$Aon`nUQT*7I}ZsNqj&AxO3TeWK%VFEGBu%u
z*)#vZ*H~MxgHT@W-_`6p!w#PM_V$l(*C3QYS~^<;7bC#m<LMA<P3s9>@Id3|*C0uP
zp*@nv3%mirz&+}0Huu|jjN<OGpF>)IzJTaxWt?^{)1f~+w#j>wLQ+_3k`XfM31cM$
z`)ByNHK^rwq$~FkpHmRgK8a!-IaAKo<Uqi6=BzRvci5%Aw8(`hs=t>Bvks_6Q6TX}
z;?-&`Lz4JV2h#TuuT;=ReJs0icE)Rk_dpEO&i0LDFfP+_2GNjZ*r&o(HR&F-q;I)I
zgT`0X&i0Gye*>HfvDR#GlnP2N3ia~-oZnA{Ad_HtLH>aS_{>kDL=Z^x+m3S5A8#~;
zJuRGx{FByH%aRKA<IU|6%&)<r{)v(jGi*J92QX|S1ny<#&~zdW+CLVYME*IF_pgsV
z@Bg5lt>N<E)b87bo<S!B>R=JzH})&{G~qVro~LW$ud_Y+)LJDvz=UGDb5CFJ!R`8@
zptNU`Xw?=0u6A~L{|((p-%9)SixvhoB!&k_by_=^k05#v2|tCcB_ELafO^M!6}Wox
z>jzt0nGC=7YfP5pw-S3Yf|i=tUV&(0uVp<bMju?>CJcO^rp^xe9;>jd7_^<L^{OCK
zB2bVhrxSCV>*D#n1bv&^&TJBV=_+?OTRK5~DcffnDCfS!W@4{nEBo5X#0W9H2o)aU
z%U_X&(lerI``_pv{Vh%y-CE*#XrswN4$}#+Qa^)+113Dg@l24}0<u$db>rO7c7RM)
z>p){u77y}Zevo4y?ffCEho9ti!`A)bSS{LG_vj2;`6r!{ogIl$(W6*TvRh||L;bdH
z?rQ#ETmO&rZ94Xmg%IWe_#0&$+kv!C>=P!?C4z#3k#NtruU{j;jS1>C+@^zn#~}9+
zV$XmDi-g6NSm1+LAA(+a9X53*QzWG~Te>zYJ@jyti~0EGQPp0(ciWUQsW=kTcI(M3
z{Tca!^-H(OuV^4=1HG#ed=409Jr+Z5cv<m)P0)~oE4W7{jWJMB-L0(vMKF*g;df6E
z>9t%xc68=N?X`7p7<p7`<A#grsAhDp`k#v^5l_ybd?>baej!i!77cbo(#Da0Z$}_A
zSXCuB)};Z59U{kPJ^yk+N2*?!3-tLr;AtBXDgZEDbfTkFX@*swMVal7J%Zmu7;Y*K
z!sZ~~w*<A1dhQ4!kb_0TH*lQJcoL;MYl#*(8nQ2TBQ>3tJ2{_Zm`$j$+5-1upq5r(
zbv0&lSm0$fUrYXL0(zC@6rz6kJSt+R`y!%$Q;E)@F1YVxq^6=ka1H#z_}`Vk4yJxm
zafKuE!=Q~Z)KJ0V^0I_#P%=*|<@c@QNCd}a<>FGQo-8jZ8QKeY;J!%-VAK*2<ZNal
zu9_}rFiZ9E`#Al_qj)WU-4pPw^w>|*56TJ>-ZYGVXGwP8Lyl7tXt}@sd+u?}ySN~7
zw@t2qk{7N9KSzix&Te<&N)I;Fc?Q!ouZn{o3<U)RNT?t`2_f)VK#!wqFy>x?QatW}
z1o?v9JS-;79UOY$BzMDK%I_nLha$0TBXv;DNdiqJoiK9fBtC~Qa#DJ_zo$<#ow&tu
z?^JqSYEp6@4J-<Nn0t>Fo69HFJ=%K23c8s&&<8BRc;5nHc?hfnx(y-GQVVP=`%s6#
zLk1}qT1~4{+=S9V?xsiW!M=i&l#2?IpSTuvYuvehUmhj&4p(Yq4-<ET*fg4>=rBz?
zcLFv{NQUw8_6!ODt^IZ0Uf?=a4;)!H{ow;L&++(ib$vZn$Q+^6<E>s~Z#{dU7;bMG
zC08P>cQvtDBF2b)Z%a4^Hm4h!rr@#lSGg4^d@rExK+Q=4N1b)eIcZJBw$D^e{eI}!
zr;8$=4-eZRxuVdOV#4W&HsM)s)U%OAn~Wdwjv_ix>gP~-UL_i+cksvU;$?0pBU#A5
zfbIyiUUM6p&i%fJeW14R8JeBgXjDG@V*ziRVU^1x&uN%5T^e_f@2)&P(<H-buV(ev
zAh=UM5S4yR)ofO8!$&v$<P~;f8STfAcUv`#`5V+gyW|wxB*pF(5}f_$%mL=yjZkWk
z<{M5{+ssT75@HSA_xgd$7#Qhc0#9hr*$OauvFk!mAPrvs<>{r+{UJ6`Us6WjL)<R}
zGd<h?!n*J@T7lX4@+t)+WQb`DOj)M?W!}R)<FaGY$Yr`DNt9q}cI$NB^DZwisYwgM
zPI~w5rHrZA%MG`oxYLzgHLNwRo4*%7z3d~}oQSR6Gq}sywF0-v3jF>*va%3QY2etk
zA~0W0A>UO~S%3;n;JYC5HhG_r%;!yQQ@5OXu-D+naJf8tY2D%j&R4+n_%cfyqZkN;
zHPw+NhowL4sbh5hV>(h}1Vt-J+?62`ZpeEea(%Ec)p2(z5#2zoU6s5|E|b2L5IyC*
z;g06Tv?dP!{KiuSaU7i}L5ZQ%>@}^FEHI~zd*~|(Fv;UIb<os5015b^*j9FMa4<r9
zMnu_QmYWU;<i(%sONPQ<|AJY{9pmM6DPo5LO=n#xBUPo#v=Z~l16~nV^rQ5Ty%V4a
zXmeY@7LCj!9tY5r&h7ihvU~b3;k2G2;!f~ygQmyy>|h<Xf$~3YY9+AZT4~)$oWT8H
z-psWXy}~EYQ#>iKRjT_SOC5`XkVdC6`Nrx(%N14wEi7;0qXYmxYs3Z(9UWBH!i171
zrnTB;BU?lekYzfM{zxoF#a>xO1-l40HR|ib8;a<Wu2B;S<tZ*&_Z0&pXSdQ{Pru@P
zH+{n_W9$G^h3XMiv(XIKgYSi{5uUQlaXsDL0O^L$R`mDxdtP4&WpCe20%ik%EOI6$
z3^21ODk<H6skmPuJUX^0X~~v8S;1W086umlUbLe~P4ld`<Vu;t(p4z`MiSi?)#ivF
zlQA9E&kEbY7yj<yiXZ_+f@>Jw?4p9h5nL~Ou*Lfak9oG#D2uvGzXvv$FF6h-1s>b?
z2ub$%UCA8Wiq*VB*xEGD9RCF$-5U$<ZBcK(klbQ-nl^4%wSA3gpU?9+^tq|0SnIpz
z%_KLr7-PV&1aagAUc53qEO0Ud!=otZ5`@n#@A2@A71w`;oXf9#jb-@`CqXan6@JS~
z7s3$y7<|t2`=SxZfe`%+M*)-*JY0A&86WB-PcUSi2(h*au95=L<!zlL4010QSqNey
zJCr4Iy`1xYd?+j|49x;pv+N4fL0yM3@Go8@c_xW3m9M?m@=2abyyRW99Q?8!@(kY1
zAQ}V;Q)@>@5U{ADfOc1{O4Gr*uf#8~(z8;_{7|5y=HVatX_*i=cm=@o6R^xeB6@D)
zencGvZju=(5FiGYbi?cUzi3V?u-Ee505>o?y6J4Jp(2VkXjeNbJiK1O)hp^?6~8lQ
z@=5;WI~X8?a|Q+nuy6{sf|fZ3AV)EPm)CFZ{1=tL(pP=*+Ppl@)({#be;1YxNFY2g
zwMRX#pdqX~U7?xvz4iuz0r=i62T~Z%m5mwaYCW}$eyg1=cl-Ivhx?CI9wmlu%kUm=
z-nQHAHOf)sdq5zsD&wbpo&SzY_Uv$R{`@8h9i7n+Zv*pju+Bhqs{kh<aTI1P!D3ty
z)OS*q{xPkqF~yMz%SdZ2>;54^JClliP9%oWZ68>vI?bYiQ_^cKqlfasz@Qaxy$^`I
zZk`+vh}aA3Niet+LV-RR$u)qz72?rho#_j#Ij|2XAb)K3{5hbHajN6*6f?~RaQ_sW
z{ZYK@+&|R#u2?dRzxA9ru{+*^Fh=W*3Q4!K&Z4NRk@;VXJV#}<KgkDPc?=*cE;D4O
zmhDN_Ku1SU2KpkhJzh`0D<JSp)9L@U_m)9ft$iQp4M<CaNeV~`Dj*$7C?X(8NTVQ~
z(hbreph!0o(jeVspn!CDcStv!|Ju*H-}gP=&YU^t!<jkv?3rga&%R@=bzT3uesOUe
zY~D~XZ5#S?q@zzPnVLgRF;3U;u}g<mLz7?0;t-^WlY;cr9B0mv*oth@UTTSQho-F5
zJX2dtTOMG`BgZSQf19v--plU!8tPq1HcqQEm<|hR8>%vGkQ1uN@Qq5@#}L%uvr|#}
zLiE#=^$#kED(AZ`b7f7CX=OO+kd0WtGMVxqt(0FE{bhr?l8|4&m=p2K0SIyr_5i2~
zJcY6c+#d8nj2rQHvg@GamQYi>j!hU48|Zt}{T4kPW^361Rh7omX{Jkpo|3D4$E8YQ
zxk{|=XM3ukY5oU<$pEM;v_GU*d$vF7=o0Mv4b+s^M#_Q_#^2HLoh(J34)&#+o;u7v
zO1LHsU03)mjdmiF@66^!FfcMM`!rGlXEVycc(UFt|JPg4-H|H9t&T)j&sX6x9M&aF
zKKC~CcklSH(YY$&r1}Y9+SoL%|3LJ>pT%4FM0O@x|0!?cZQJo3kR}McgE5)WAhQ@%
zmvvkjeQ6iEnE<d-P4%aqfx=?G(F+5a&_IJt$;8Ao0h|TEl7Z0OaQ=-Q(dGtevE2T@
zr;3L=%+ILs<o5<bw(BW0YwSbdsI5f*Hz~{!&~ezN%jG`UNgY0T{NOE>i9!K!Yd^(;
z(4RQ2-H@hkngs7g7Gg7@sfQOM!8iwN!{31Mh8!5Wo$S9k+MX{v<_1N|L&(BNQUUS}
z)W?6J_1OY6!RmJ?{R=%gIJ~^2AKl%Q^}MNU=!W4oeVuB8qS`-Tom5WDHBz^ucV(xf
z%&+5|c!ra;N6wyd<$g*BZDeM6V&We-|3v8efK1HOtO^Iv=&BifU`N&EtyNzmk$Txf
z0q?)>HiRU01y3F~Wz62%J~_1KtC26E5<8)_BJc?p;wyVhI)R3U_6+!vK+V#0_JC7r
zM6&d7)8m58Es7jZGLhpUt+OoHkf=dK@CRcsILfcB6>>j6G6av2=83w`gMICd?@5xr
zU0>a~cEIn)Q!Pnf!{9%zdZ2{6FwU@O^4;rW65T7pR}z)Gx0<W2Ca#MclMR<oNcs=}
zP_zj#@B<N=?Cfks`DFlJt!%z34BD3CWY%wCB$=$6oq419mPh5i`DKEo<rz)JEJ?9e
zokh;#>1V7SNi|Pi^}qi5zz=J8bLuB?A>rj+tCng)!`SLewmdMgcLh`)GG3pXo9ik!
z*MI>Fg|_oDW1^&C$o%$EB<0Z2wdewz<b^b*DaYc>naiJBDVMO(ls{L0eng)spV+Ti
zGOQkq889v&?pb(P@l3|wT=@=l{~!&d5X8|87=Vu!EtA5-n-TTV;x}#2vz5Sb18HZW
zh3wa-BTyFZik_L8<B>i59cbFzP<4ZdP1a7}7g~EflR!LM3QPzP{W3U7g&jfe`=|-R
zSPZ|7LB5ZQ`G`e>n7jQ0hIM~>m+4<3l_}N+O<g#r0^&~`6f*V+uH34*-d7P&R9l<y
z;)~pJzzQ7rB}yjKs`~*3!B(OyEKEMJ<Z^|gn5E3ipB{rk4?SHj?IZ_Dram8FNy)8p
z-eZNCg!?Dw{hR5mfmR*^EY1ZUzV-&?lrBrA?F8x#6bnhDU7qBL)qos_zFS`8hylca
z(7u;$0T@>!p~1c8h!=NBHt?@U8*P?wo;%5IM1_=?s*+OCo3<;kE*^nsK5%HEYXfn+
zFzE^jcJJPYXbp9j205D=hr0?wgGXj=R=+#)%VZ>V6Sw+%_4P16v6<k2#Yt9GfEAz@
zpT$220n1aQT?1=|44ZH;wtoY_{PuX>Fkt-xWVRhp0x)woNi0AgS;AS$qwJ3QDo8yO
z7AWaOySDCZdqvrRN#lu}*CE;GJG0d)CW(=lwuHp!g4`%glfwIjoPkCBSxT&kh8glD
zVtMuMpjMa>6wi>-Z^D;MtI0*!60n(Ty-lLL@Zn)fm-igiqr1t&xoRYMQ!|{m8qQq2
zbt1B!sJ>9}>)#^qq0woENx?%1t{@Eg`0-;Fm6tcow*W1`el)lbrg<>uczP9`H8Kcd
zZxMtRIBukujeaa8wH>o_%`Zm8Gq&lE%Sl1BwxRrpb~fXzV|k`(|8rxQv~o!B9o%Zp
zxT>{YqnrLk<H-1zZyE<H8=G3yTSf@$^H4P+tO<bZ8MMb(??}Q>zWTQ6HvW+H*8Q)_
zB$JDlMj{U81O&Ja6z{72gUPBuf&3|d%g>U<%5YAOr3gGsVa<eVipq%_Mt825@Gzw0
z{%&iVg<5<T5hDb+=u35Vq$Jyh(rjRR#{vgRDw95wS85Jx=}W#MN$js{a=(_5ysQ4y
zkR_1Z!0_F6MnQJfFYunvrJxntMcuE|e9L_6DVfs$VgVR*uKfX3co0<wlZdWD9Z}>+
z1E{{tW!y<3ZgCD1PX2)ZOHs^5bk|^W0OY<{I@{Nd;}(0q&#c<5D8YML0jsS!y;bOc
zORu~Z=p#X2Z_2a&T#}9C5QL$Nn0Y|HM?qm>0C^6Gz+<)-0%^5{hWK=D`ZNq1zHRFF
zrloEZcS*iT-MGd)B6j293r#*)=up?@QK)H`2={|b2Izz(n%F&mPW)r=nPs47$8UMP
z){9e?Lgwk6tDmgON!;5BaF<M=(;8b#a-%B!T&pAcSD1$8Eb`)zV4*Fkg7#uoo2<(Y
zB<W-UFY+mqZ$tn54!Xb>u}X`%tK#jZQ_9M==CWvR;2MU6Qivc7UuWm%f53#{8fbUN
zsEiCKJiXN}Kya_nJDWi0)Yr&OR`B-sw<8L_`c1U@xrjq4`zZ`uEVEkfJo~IJ240(q
zz#sk#WK=znL<aipcTY5o^mUW{2!LU3L$7h)e56!Lax|ykWMF;rkNBiJE9;*xW$#{(
z2nJ(%PEbh%u01Kv;$AdT7GfV&QI5{FSWmQE=^cziuZ{!s5VhN`0-zqR^I6e?2Oj;x
z!h%qPnP9L=OMr<@$D6NTAK{AB(b^W_JpBCx_DpY0s*+3<-_a7DLGX=&%O;US2d?M|
z@{LQ1x1v<ZY>%~a@7=viOH1ns!rOPjIvx55<OB_X;^E~m>t1ZQIO9B+^bkczk^VI9
zy30KoJ@4wkA`4mRh^Ob5FHD&D7SFA#YX>&bf3dacS&oQE-=(Rue}?y@TXm)PM?F_v
z!nZ)H4apdB-e<cv3tG8V#0J+`lvyJ|IwBO>;hXaEo$(t)eQYVFZ>OcG1v$0o_TOba
z*XbB+cYS%|I!X70@2~xJtNUE@`u2qJ9ZB?b;w3R-qa#7k!(2O`xgZfSFiXoR7%bFT
z1YXx0h+2SXd<7!!NNEn@30=%G1(1qlm00;4NJ0}m^?qC3QhT<oG^<OL=-Sk&F?W_A
zK(8r~!`6Z4>&TwR^(@&2@8(SnM4c#{QHIa;@GkAhJLIl|NuUID>Y%R9z{JGmu%QX^
zb^S|8Zss>YYhDqprPwD%uHEy+@U7w+=c$Q-IO-KSdtLPkU6VXdmE3-_k<vhz&Pf-Z
zK~?k`0uR{LkK10DtdK*K-{DDrLYMvYxlUb*Q~P&v0XwqS0+s$&V_Azk1E7Nd-<nII
zmXwh2{%M920fe?r!U?`u-<fIhcVBFnsXJ0jdCXm>o@qh<xp0T3>HJdzmvL9i`=t!0
z<Qw!UxyII=^--Q21?=>W>*O7CT5nD;Hp$le4iD-T#$?c)p*t1Y>w9r$VnJ?k@>Yt@
zImrWtMmC~Gk7MHv%WTgA)2E1K6rzX=MJFjWb%9N4#+?)yeBD9ATWVduD4&)&--tZc
z-F(3qBN(_DUma>1(`lBl<9`?YW&t+;xHC!k2wVp;+`E77MhrX#P96xRM%NmHNDyh%
zQ%u`!FEN_C_ekA3wwS*6T$qfMd+CbGMh@3$&A!hstE51`A;CG07|aC+f66iKL1&tM
zK+H|7DTrpmSEF+%5^Yo#&*zBynBZb!CO2{b^NACSrU>Cx(ka)^#DA`aEp3E7vCl2o
z_VMvSl#8qF*Z{d&qK|C=KSQ?0esv%Ov3`e>32?Av7j}oUFo-`uM|%$_YPj5<3QHZT
zk6-Sj;J4xrCA*8^EU;NKXbZl3dPgqG8Bh4p`lF@!v0$8Nt7^ds|7`bOvowp_DK~px
zj2PWuNhwpnSj_oc<)B^r2&S;<#(R=4<FAo{w{4L()qEr?;TOZAv*tIgIeDCrp{{}Q
z**8CeG&u2#iRX2V>V$z8tYt(+2GDj8O4!ui^2`<VT!SNl)~D0BZfibS&zX`3l#<XA
zD1B4xi@Z<Kp9r8WmX^kcg@w&kO=Qa6M4UO@kC&N#3)UO7b%RVjOyCF`PPg%Co%a{P
z1_(ydH`#@F+I|*fX{tU??&#jlVy2ifO=HUn@v*9o{b$(@7yH7pwFV9w%8!Vq13<Pg
zvoapJW}NO>Ag!XV?CZ?Qu;Kl_Ew1I+sF?0cW#40wzkfev(pg3l*}`A}=GFabHr8`b
zNEsP_2#gMw+BJRSCzt>7TtEre)J5h%>ME7P?fjXvWSSo>dm<vP2E0%6m-{o}xFT-6
zq>wQMZRUO`S=BUwK#bsr-LB}EvOZ<W_!=!SQ|wIe?o<H&_~uP*E&fVO-02JE3oNv{
zJ?3X>=;s`NS7n;}QsWkImi#im18VUX=*wvuGA={;EL74J_6fTe22O*D581Y^<?c@n
zXqD;@oeYF7Sdg0BeR8e+VADJU)IC5+ArcTFdb+w@wMmK+lCOI+Wf4IhsIFL+kl~Xp
zE|H=CtEwlrgxsZO$1lI;;QKCLs^@d;{n*9&wJBOY3sf;YV89yAfdwfv?Zq4GeS+y6
zo0u$-<RI_Y`SxUj2drN;8Dq?IT(N^K8hu$mYM$-2;%4QMyq}uF+(@0GMz=`iJM{Y?
z@^#~f3mikvRV^mXf>l(H8x|H8LN^3u)VFIx_dpRil}v1Aef<xjBP8T#mA}Zw$~sf1
zdd{%lNNBGWLQlu#_0%pBBjAyAkvTC*N8Qpq`=yeqS8Z~?_Hd10#sjqga642!IX}Ib
zejUAfpwhF1e=y^GnJT*AgIZRsRuLBuU2P&)`R&{K!y{zTjuORG+}zyh4x9g+g8*-I
zW<|Y$=8e7mgJjojT<BD;LnA#jI?4rDSAd9XeyK3Gfhf$%20a61p~=w`5&CfKTx)A|
zPNHtI{GL^DaB+$FWe6z~F7G;c<baS42zv#=T~+`EbWk-%X5rv4DfG7y1KfSW5ilCS
z5m42(%gB7lamovOMW~`Yh{*Qgb;>_|yU)MOnS_=cKmA+gON?Fp*so-L3$x6u6Sg-v
zHSz*xBqHb*@{u6jWeaI-_?t--tmvO_#-qa6^UDh!Tp=dMY8rPk{Z6~ETURK(&h>{{
z(FD^}R8ZDL{T7QcZn`!_#pw)X;(Gi0J3cW3r2Xq_r(J`+iFB0;F+U<Mj$SCmwHS6X
zO_^*K?l{7w^T=)t7t*DmX^)-oBqs+<XLw>CSsti7<E$UWEKN;)m5(#)hwz{Q1Zv#f
zXu*3>?OCsI$y>xT;nB+jyu*Q|Cq_Y@^;Xl(vR}TJf!Yf#2Zvm-sQ_prF|?CEE!si<
zLVNeCr(3Wj`+<0}_d|S6lQtPz3%c|*;@&*b)4v$tX>X|_Z(LUN9xxarQ6L2ZXSHcx
zzO(>E)lSIm9*>u<&Y#Z>p%&HIg!y7Jn6@yc118l+?Rfk4w_;)_m@6!T_zN^iN(Z-w
ze<}yedrnMUB`CM};G$fS<0|s7($px;kd}4Of=>AzzY;qwclWxR-F>fIqLmE<sRkt=
zJ--#`_3S|Bji66JaKPoQO<BV2XR_D9wien@cCG3N_!^8Zh`O2b$%M2=msitsQd2(3
zdX_E~`oC~szZ_rRm``Vdw^Z%?^fAb8<f)b1=zlRnn@^i!`WZ{^&0^8)>%Gt4+4N0^
z3t=8YLQL!lD*YhO)-8PlXjYM|l>orB3u=MUn#lQ~Uj5+=K8%#DY;7ZI&CBjosf&vZ
zrq`l(qz6#1=rKvkJgyojO9Fl}kjTj?C_Y9+NXUf}KlbzU1Az#C5J2AE*-?X25O8<D
z)Ch{bFxbCmJtF}K4g{SHY(EsD?+Hw5tjpI@b}A|vAF>|3`4B;O$?lOrLT_mp*amfj
zZP;(Qui!d?RNO#kL}3w9{(|+4gKz2-AeD<_WZq};q(-TX_G`|g#7`{3W}r5$77ZuD
z8cf}#JlUossp|Q+1vl?!8waLq%Mv%k;I4gaM__b$_4>?0mvoMb3!m@b(GAQs;rghp
zvq?RrzW6eC9T`}UA+R3B#jdJE5LepcE$(m)@yyXTQ=J-9Y}ljK0c%K_7JMI*fO|3E
zg1I;~B?ds0Hg?4g)CL7wq893;<6R+{qe)q;oEEMrQS;8(^aa+y#HImyb!BA?s8|Bl
zh!oBW2Bsn$Aq{~2WmGdUA+->*B%bOQrDdmtA3X;Gq360O%9<l5RFpbIXyU0Yy>TZt
z9pW2|Mps$K1TCaCFb_{KD&4Qwq9+WKmf-X9R!AL0nBcbol*I<t&+=E`MT1|exPd`t
zXhl(M?DbUPYO+OR&Pyc~E_HQI`I4~f0Cxl%ithGpKL{kt0iH0KdJSkji0XEivO!PL
zcb*l`rVw{aXv5>yWyzZAgy@Wu;AbN$xbv)E*RI^ULbNhji~U#wN>--zzHBlG8!&NC
zhMQ6h@}S14sW6IxnYm)*UC_3K$Nw=tz6}x1fRkI3vjS*coYnluEJnPO<b;dvo5Yo=
zzsSDQm5;303Nf)ki3}B+8TC-3BeCSfEnI<QDri|jTLbJ95KMAU6#}e=$a>#w=q|V^
zYPA(U0HGwnHhh4YI4A^@CL|^T8cIN3c#j-PmI?`E@oHAuQ3~Q350ik}j675GNN+D0
zC}BXyV27>)=&9wgDlRxkKLsk8Vw(bhU}6BH3tCc$OSjn=^A*a2rEYI%laa!Q9Pfwd
zCke;8OX2seWeUEp8F4Vl^$}Vi_CN4;Lx{zHY%H_9RjzjNDCADJI`I`p7i8mVjx_|g
zKMP4I2j}#6LGZgnDk>auaaJ-bt;oq2!{efURh3HYC%$|<5E86EUg`5xIlCm$tDfXl
zz4+GKja$mOcSZxqTuh+?uox-P@%O|!JTbX-b{-R2;fo(gf917ezfP>0YogZmmaTFt
zx~|qbWA7@6pbrp{J}96nOoSXR9|U<$xZSwzO7LRCKRWl_ynTYtmwF6KCIz<#S>nC@
zj7VArFt09D_(+bq$?LPH3`I7AS4JGHK#+tSyxu`~FS|MKMfVG4974h<LKOhbLl?fX
z3j;$1^oDxWQS4p)`l57<p&f#RA)IfbSb~;MUNZLJWB3I{5t(<Gd5`PhMRQ+Z>vp(s
zJx?aAH<mXBz9|DcoVhaN%D~(jrl8P&(+dl0jZ6UD>hE7GmdbRXn5L^ll(GjF1wyN&
zQ|tH<);mH+2L?&T!53J@AiDAq7(8_iv~u_BTn;k9@RE<89~#_0;pSUwqonlj<V;@A
zYchT~o7%otw2C6T8pPrkrzw9m80RLRF@J1hs3%wLPKRI*W*x*fV2mI#qwG5MnolM6
z4i2J^<$x!Xfwus`e)Y2BeUQP+YK=>UzP3GClx8xKYE?_(rL2L%?fp!<(W@$-6e7I5
zjnBUJqp+r>z7KD{f;bN|#G+T<kQfNjmSAI4{~Qf1KcvaWu(2f>6w;O*i)<F<5<8<n
z%N}}RI=KF~&+c(?=>a1O0abt*o>)si!1rah%K`R~0XSm>D}(0t=gAmvebbg^1KJhy
z_}bDGIjQ$A6qJ<MC119HFh7*2i8Eq*uV-O01&-x_usJ+B`WP9>ILv{_;v<%%`)j|!
znIQf87@#E()6EKdCJT-fo1muS+Yerc3>g&3;`HLXApA;Wx`O?SCwrX;iYU|$eu25I
z?lW17D$ju#Hc^cvJL7Kv63RGWfABzMMGgC&-sd}6K$DDT%!)VF<9%DsA~%r49`p7C
z{`D3<r&qjf=LvFZPWP~mr6QCgmdq^8HzcsX5FZobXN<bN4haC~wcOI|zWEYS$Q6E3
zwh~*5M*;^<_6Liu@u%3GM?}-X)GtsWMM>1BI6KuM-K7tqZ>#2jQU+w3TDrQ3uU?HF
zxHf3Ms7=xpAJ&L3YIK^oBO|dYwLtK4pJr*|>&96rrY%tVMbCE0V1oEDht%HQUI@@r
zFmfgVP&oipI1_0P)<?hSJ*5Ew5-Ov#&!1PGIp)ZVd0R~m?Ou6V*%%-%*|gmy|Do~*
z)*p(`&Oaow*vS?MQp3dle*2tnzsYN5lKa<tV=rOopuLb#lP^)hP(Hi72yB!hS)&0`
zc<tuRRi&Qy@882{-rqKb03Sb>XQwBRq51eZfTc(s2PVU`lHu@53B(s<@~`(ZSrD(Y
zLk%b(-2C9dgWa{^Pe`<ffn$HJDhQ_`oF_xsOL&+s1;n*8pAcVO!jR`gKgk)0-d1#M
zqG)|O)y7sP!!XZy_l|^8;bbU9J`K<BXFnJgs9Nn$pTK|+PTRq+0=U>pZAXZ4xsRV8
zDGg2Gr=?7{ZP4P#Oi!1n?i?MBhfQgWM%<X<>Gf>K&6{?#QT-cB4)g2AYV1lSt5}S*
zgFd#;VVW}VhlkHr1tOy#nZ(ECDwNE&%1y|he>gMrwW5a34i2(2!HCXI#4xyu0KGah
zL<?LP$zM@{>_4=~b|NkSYLEn<hXWRJJXobHWzK+`NO{g&`BsuFO-n^ZO7N<NA3>$m
zW<WKQbNSEaV&%U}PaLPtqZ%HpGc=okUuXJoIh0N|pqQQA3>F`VJyuGJm_Pv-8v*Z5
z3j7inW&m4;_^2R8-me$dYwQ?yz7i5EwHW3XQEFevl<K;2T^kwHJBwDpkuUIXqM8=n
z+htF#^&1be(O)OYzSoZ~0&WWc6?$qe2Q54kBctR<KE$IkQtq3p4&y1F1$k!uYBkkY
zNOAzsMkC5mR{(R?CCbIoGE=6+21o^NHo3o8p+ja<rdaAPdY_ZUZutF~_~}^l(ckOc
z93{Gc9Lt$bf2Fw@+{-)|O`lI-oHs#VEOpMcBwy*+kByICJyiDh@d1!|*P5C)M3V6&
zw=mR#W5^8Jphel~)=q1%l?IR3>Kjbg(N)Q${p-I{O%0d!PcC3IF9lhC33M*Dz{Uks
zZ2i>pGojT+@}TFEM`&Q+M-4s;G(cQ5tobjCeiKa7eMaPXa`mOM@~<Ku|35F9J`?t$
zghx9l?|H<7E`<#KtKHpQq-O^Y!EXZtDgp`!{gmnL0H_8Ee2|iq<Y^_Vr9&eiK44_9
z^h@!lHA82}_r;J&-Be1gvp<zjzpN60!Yi`65=zCt*wjjJiq(`YDk;Hd<oxvP^fVwW
zELy_Gu9IcS8OJ+I$n*dc5V~iXU$gRAQ$)4!8qB>!gkLS-+gW_RlL80N2;K&&O|zjQ
zuUu8h=2S2n1o`g8TP@Aa=%S*cd3kvt#7D=&Q>NnL{qZA7qJUjE`X#xSzC}e0Pb#hS
z(;L5}Nu2lv)6|wql^UY*-0=tgkWnn1m{JO<*5TgU4BnI-Y4&g7#wRT4_-47;+>w!z
z^Bvp_AtAx3wirxO;4}(EYqr0W$-_tiJDao;d@`sWE-WvzB%;#N(hy3*f0P?#lIFn3
zN>Nb3!lHnI&<_~<r5u*Tx@pUcC(P$$6N1b?g-X8>NA2k-uDUU+M)n%6pHa4nbhLVj
z-s>{=?3lGyb=a9mJpkkc5wUOv?5YK@D9_z-Gw0br!C(Yxd3BP3k*h*H7esZnwG8(F
z;VP%Degn})4Zh;`SW=Ra(EhFopM<Yz>EPq!Zxi0SMzm2>r<Ni^73~Una_|vwss`gt
zM%+#SdS{d7E3tMD4n}`3g<A@`(|~I=mSP5X96$|cd&G5w-le;So)en=Ci3pv5UVlN
z-xA+WtUm4cd03fe1pYl!^C!AbLRx=kn;p{O?gy8Df^Pjj3=RQh^1^a&4^G<!KCX<k
zKz)V}8qkd0*C{DAEhS5}(cJdX+hcGz>+w^CSH~){xqteKXDC{OisZ6zF25e^=i7-G
zNJY5$Xvpu<Zb!Ou-tdu#4U^eN^}DF;pn`k$NmN?3*!NBXWhsTqK328)O>GeW@<<l+
zai@xU0XVp^Q||Vf+g39Xw=YS#<f9p_iIj=!f<<opTv-xGhrv=lIy;y|!ztN$CAZje
zy{mu=k36f-MC8e_Z%^yn1qSw(=bfJKTU0)%xl$$b$nR9D(zw5O&hKs@s;s!Q3!?-_
za85w-KSIR--yCSLeZiL{O2q9LNv=?bX87_uutAxoU5l}BL3nYKNX6jqch$I-k!#=N
zU(kxR#jY#eICVz7Cf)trLH6P!sC)x1nFu6yAqyz2URqpVzeDv5e7{T?4|)9JahJAl
zz1$HZfAg&?jFw3`U?t${)EtdfzKN>uBXX161_5mRo3$p>YZDmSj(gW#W&LFOhfFM2
z;+E}z0|Y&qG?N!Rnt`h020C?FF^?L&s0(f3yNBy+uUu5UMYRNN`0Xp^a1`BMr75x6
zQE~&22P{z^Wh;fp#%9~RL%ll4)T+ZFAowvjsM=!I0A}M(xa=;t=jT}OpS|iCB)&&j
z%-IyYF}22a>Fplw(by|Twr9_7rVuZ`F0>@3?rvzn{yG_Ke2GgY0*^5Vz57uwY$9kG
zJwaawAfA6jnv{Vk#F+6Kyl>IEer6H;#G@~W#5o9<rWTJKnYTO5J7dmg*IPQCrfetX
z;+EOeX}1bX73e&&K1E-kO6Yu2=>M|s;O6pqo$-m|X4lDgEsDq=M$ya@?%U?a?aQN6
z*EB)|pF{HiBQGFaNjIATa8CeT{p6yrrzf0F-0v<lAZF|xJLjjoL*+8$9Ep^=yWYg^
zH0?*veX^9u$WbK|Mrv%2^<chJpwrL+87*TlQ<zL?=P=D}I>plLiEkkFoC*bNHUBrV
zhq|VC^5(>ZH~XJ0<I$T|mYT_Vh3yCZ+LS*w);=7jd>MDiCD&}~)C|{Ui*8@|#O~AR
zm5yZ5Li#L5eECK=j5M6q274ujtW}CGm-1%87d~_cd4;JON}qndxccYo*P7w?fgoKe
z*Y?<MR8AcII5&pHY~+H1qg<rjW!uthBq;WzG=V(n=zH+5lzY#vpYwFO-4L^QD3JVY
zA}m_wn;a95ON!-KCEby-YG(lzh{T%Q>!J4e^Xu*7$1PZs7vE@3Tl0KrinqdJ_*Y+T
zwv2IveNu^NJEWrN{hH53VM4LE-MhRz)iP3VIa@h%&#$~8Eat?p|D@LLEg*-$;-|!R
zS<$%l<PYzp{*NGvo_<5<P@zf|=oS3uARcHpLolR<JqqImP)_EQ<$=EYAH_PDI)YIB
zbr3s%qO@nYfiY~brS4XW==f!wQ(PT6pNc^-`^YcFRrg%)6P8n9h+d_8;3159oU1M&
z8xnB|5YbAk{LZ_~M@Kif4AwEE>?8{_`WRE>%}80in~1!#DKta<|CBWKP&}gc8@dQe
zTAW?YwSP`Ke4k~nsU2%O_bu;;-Ib*7K}Fg-3d!5WH_DB1J`R_esxFINqo$t7Lizg{
zQJ|g83cTDcIn>r6ns6c5$lcaZ&Yf<iAn7aC6AkKJTrK?oKDCV?@dC6-b?hz(w9?oe
zE8bc0@uRZxDSJ{*o7#lINxnUgmUg(`UyxrBA;cbR))@8Ccr1Cnwi~x6S5ss+WNP@w
zEvq*tx3((1zDB5I-M#zS^a(NazlhUr2g|@pd`*o@+Ib%Ir2UA03@AuNacXz$y`Fxm
zn`ih;`VB4~`TM_=9eo^yddB+Fg*@H*6o%acMR#z+sK1ZAR63WRPrCJ1mS*HeT*>A~
z;h}39Hng*}J4EhcYmrnaI0l~iQ}AMWy07lHw9lOx4&p8z=v4ZRNk&0ALaD4@N?txN
zxgrRqQ)!F2Pet0q(h>UkR-K(cj$HOa)t3}Y{hgM6O2{)n)TMgp26Afx6}=>be$Jn_
zFIy$!UtY)A-0Ti8a${tt!je-?&R%;RQ%c>+_NtrM>Vbcsrh;{7uE5mE5Y9vDg=XE+
z1xY8<G?iR{<b0F>>gT8(sW|9rGCmaUZ})#>7*<_quy5w;kc>focSmMvq8R_rz+bQW
z;rG&Z53@~mE#DVPtSZIiD&#7=mP|`N*w6_0RK~QEH0GE1#znE<=T1*E$K}f|S0Z!5
zMTzdrOlKJjb$f`sng4NW{r4uvo202D?1uckgb+0e=!#q6u*6ov4lJZPFwp{dIC&(?
zfMQ~BCDnvv+AHrxd(cJTpJ++8#7Kpk@A8EMWM!ILvj})wgPvzAvxm6);tF0RULFdh
z@)k84p*hti?E(>KIQEe7cmm!z)GllH5YUkEO*gkQb*ANG#MRJvwEPaVuAhsG!{LmT
z946R6|3!OQq>BdBN+D1eiHM52?{;iU?Qv?I{kSOOKbvJz3tM!}2S;Z}*>2lN?{Z!3
ztJQm1N`b@mD~+sdT+AHar1^U<Mkwl*mX@%DKgP@4LCe2_Pg$4|kV4*oM!b8M@|Z;K
z0+X8LL$yq);oQJo#V#)Lvo6}>%zlaws)CJ#z;rV*%F^ehAR*~aU(Wm4^%||t1jDsV
z)7A3zC7pH>G`1GGrb!pwLAqac3Z@k%Te)YOP-8%f?B>7g=i{SmI$B$`=8VbR82RaM
zgTHS}i3ZMPcg%W9qF3~iE-kIRx@V7HYq*Jj)Ql-YV{PT>#YUN%NrZ)}WCcN;6{>5o
zaVy!ca{m?9YK4XFQ8tBHO(_*<Q8K3&($`7dZP#P|bnYC7r;Ld)7IbmI`7ajWxWr`Y
zEo*<|Wli1~@mT+(`bmdUUWfTJahksejIZJczhw(r+9y8sDOtFseWgoQ+~wmojP`~a
zJUkK?Z{9S$%j4Uo`n~1%54B(8ZJ$5k($l{@q<hKRtDnIG83A-khr!}ycSJ5xpudtz
zoQm$~2D$1Pl~CT)C~Me`nxO6hWB0+>58NomrViE$T29j#;!ESB(wrID3KJm@^Hq>p
zpS11IpFdYXr0Na+!*ROwnkVG8%PAy%*sou8=#xkbp4kLseS47eaV&^768NIXoKcIp
zFX5hb*$#bI8t+5uI^5lWs2&^``vhM&hS=3y)wP`OaFMs5r#}$ls^1ism@=}b(vlSv
z)DdyEC-PuAZ7dWNrR5c+qQ`hq@4|!0X7f1FntKo_!+?9C&Q`4H;w6N$ewC7vz2hcD
zs86N_v2H)Lp0N17L6ZJhXqfbGva8aW$5P~~5vfYdSxUKg2e@F`H*Gka^qoY9+uRc<
z1%x;_x*+({lVc2^IHcxOw%gg+fm8QFn=%(!IEhEhzo6TSa+c=e;zGQrUIL~PWU-OP
z9R}Z0IZv2w-!2Ah5X(jxV;JvuNCvs<D>>b&cwh5+oCnJ=xjoYuch~g4O$E|CFz%L+
z$;w67Qp5V~(fm;-g;^CxZ?oRYohV|?)ZjdOmMpK2tYtLrWu=%>11P+7#RJ&%wXw02
zjJ7<_9n!8`-sa{U{>E~1-8Cx(V-agC))hx<b!h6z$jQ^`qhOd7KlZMNm;ld+m}2S2
zg_o9S5PeOyuhp8P9(}DS&U4ixT{hcEcHiQaKlz{Yg@1Z=pB`)tl*O9hK>wQStEy(_
z%+n^HPrK*$vs&8g9c>M5sD7aFp`=>_V;Zyfb(hK8W7$lZSea&Z%maQt{I+_|6~+?P
z5`F5n^!Q6rpk>d@etYnae(%poeh+D-q_XZ8L9fwprZvaUH%Cbp>iTO{LN(}8Ov96s
z<Nyt1d3KD+B1lS`n>9^UO0*z<>7fh0qv2IYb+25^H9pSJXlj*5d<VAqd!!b=o&|yG
z`3vBbjR<Y@tb{1Rai`FqJkke*w<PaY`d4P?_^5T_&@^S^ryO)0C)c7`D3hi>=;yk9
zbI{zdw|bpHXG}&i!{F!lak>wUtoNw8=)&!4M5d#ucD|MwwPa;2yHapra_PwgJ+-`B
z+H&!Ab&2)uPVYVLyVG7p7kBS(hX~}3<QwZ@&y=X%$;|aFv*$p0)+-W;;;TL~7s4&y
zqGjX^Nme|+UHGi@hVoS^)aA%yzW5QVjgKRyQa9!ZVZZJt#)6D%n9w#(r_j!27-4^h
z?oj<O9A?!BMgmSo0cg<L(E*~3N``bQdo$vIf`bV*ppJXg_(uw~1Ur(1Wnke2(^#&t
zPS?baJsMA5HQ^~S#yGI^+0dRDn812ZHE!)@)+5UG%ejlo$CGZVmqyOq>pRcqNww_4
zjKY<w)7$|)yo@E6{rBYElTr(J+YM>O>-^8?3sZ;lG>LQyi6_}0gr;Om41Vvs<SNuS
zn&_}W(<&}}>+k-)MR~qb$?rNmuIeE2M<;~3tArvXx)#_T?sTO1Xd*X)c7OIO37`yA
zl)p}Pe;)t#gNE$|<|mrGx@A$Xl31=npZf1@sTVEvrDvzMCj6!y1sQq8&qzB_pDN@Z
zW<UR2*k|Z4LXD~(ad@RZn7wNqQoxw-MjadW+y<cYP6f2n<PRoUSiC-cp`*hky!}Wj
zQ%cAi{&{labl#zx&5dbhY3Ub$1_N;0=@`>6G}%a-m|Qr5|9+}f9H%N}qZ4KDiuK}i
zPhSX{d*~oTmq)Lu(b85J3({G@a=^jCnUcMrn~PR$tm;*sW;&rY7Bved;YA5Dk#qt|
zs`jGTXsg;|&itl_I2?E*=CGbF3?$_x%xs*uk9tHxGcA}{>;%7XZZ5k1R2bMiy9_A^
zyG7R(WaE!!wA~M1TuLX8(rg*(dLJVU>IaiC)6<+>_h&-Xl@~Pv!3+Y(x%yLLvOi!>
zfXrq|uUOuu+du^lXH&KbcXp5MSnXRH3$mp>n2WypB31bs%L`?z{L+21<?&~qI*Y*7
zoK8eU`+lExgL_?R(XXZ^FQld$P~`mv^8qm7S{^$9_)sro3Yh$XSbbJMhpvGERzPlT
zlt_VA2x3kF?4t;GIF0-{QU5Azu*S>HF0YM?$46UaW?&qDY~5n@>u%=K%^+>VWV8>p
zn4x)IGh^|a*t$k|A_u|LOErE&Za20rS2IbH>HU36X6)yl)YVP3mY2b6^?dRCutG7y
z;S<ZRE3<qhs$(l@#eNONiZpBEsai}{_*XWl_Nk|=gY(AXnqpTC`C@H?PPbhJhHaaF
zw+CV@V+VFVO)1jU%uvbVHov4I6=hqt`L3{xGf*usee6loSP{Vj)`HK7?^MsiMA1nt
z*}VsTU9^99yu>`F&l28hUrI?gT){K)R5qo{Vd5Jj@ORO8u{4M#Fx(VN`GnKw)1btO
zl*-4AujoEMs64aL4o5trDKsnTd@8fODxK85+Vb;TC&+zCg<OA_f0s}|la?Kju%Qt~
z`P-Y6Uo6?(HOH97Sn=o(yd;i6_AW=A?<YFr8(#=D;);<eA~y3W<S~j$pGjgObDsJg
zxCVaw_~JrgdW;y*VxR#_=H8D*-h4gdnQ~LXuHJXevB%qcZRvAgtgHUUg*p~wrqo`d
zV2v0%X5RC72P_zvYa_ro07Li!&ntU_8dM-4m-DnW)<DAuCN_%CpEpLZ+{n$1Q!b68
zV!Gp|d2$|y&!^62ylns4guywkjD?j9jV;#6>a_R?^IG>;`-Fo;zj`{`8+i2?J^^ws
z?RMOr;!6m&X#JSq9XVi$?3%dl?BK7h)%2~bU$0T)y(ukw?XBA6LM`G4wrY#qI$VlV
znza@#Se`lPJ)J?Lsdh8<o+5W=H{%+n%eUV246+MX<#mrPMrj+aW;6^tXYZRH<{Vso
zzvZ6J<InU08x!;Hg-6Kf1Hl6hlCFL;VQiVOG1nTCpjJ8&tud2rC1#<K=77TiscdX*
zG~Lb15B7B}a8D{8*59^B7&G15^uX3LGm<JY6kW?7`+DEud*81Fy+!-y9A!?-OEba&
zS83FJ%Bp(JsZ90mDy@9d(QY?v(X*(|S4=ftuDoirPid)jUhyX+e!RqxDEP{q8IUlA
zHbtU(66Q<qFOT=@051$;0|=a{pH@DA`2?jvy7WZuhFAUQFz@4lxtMQ7i<cAr5U-c`
zW_)mbIEK1?P`3lTg=4?2`Fl821cnX)8vAN=QFhNgKgJqiR^>XLJ`AWPs@KBx4}19P
ztC1a_$MvW@5?(hqoGowp4Y8Uh81z=#S1PJ29m(zQ_@;R@_3PiG()bj$C|Jh&^SHVN
z|EAu}vG&o~H_L%xkwl4k7&ztlKYM<KzuQLNZr%RH*~Zi$q~n@@EK<mx{JiFavwQF}
zQ^0*HV()C5dL+CLX&R?+w)J-Je*Jo}U`8FIt}FAns|gRw<xweiL=--n*2^8!*uxur
zwxit|tv^bN^wP5Um}t5+tt!5HX3=%sA+45rw>NJ4gFZmC#24fDD!=P7QI8F`$RD`}
z+P2GlA2=qwm-^`N^u$X>`C2o{jOIH|3&@zWnCVy^SK8<+Rb7j{8FM%N%<cWo+P5=~
z)FtZ*<<gZ!8jJ9}KXAe&{q*T;AXS!0<j2R~1g#>0vsS=5GB7ja_&(~MX}P#_skGNo
zI9jD~T8U(l%7s)h_sXM+2B(4FRQ`%O?5w*5Zy&H$(IxOJI(#eRdK4Xb<z2|)I+bCG
zmEAiBZIVfyHs3Oj<^`(HN?6Yo(3ox-QaH7AixB^Gzg1tz?$32H^~>PN7|zw49256D
z2d?I_Nq8|5Mdq{<Die+x_UR!H&w|h2@8*5IS1)VMeV^1I^SLJb_mX;%$};Z_ExP-Z
zYi1S)OX^P?t`0pfOKIy9yOg`Tj+$Qo-Z^sZ`oqy0=ApG;B8OWqx(OGkaJ15VuuJ8w
zax+OM2w(5JwodnY=o;xh;!SR{-KTyM9g1OHyj49eZc4JTK$yo*(=weSh6M$!lwH8o
zy-p*n^3``+TB<)xC@Yl37U$*=4O4*1+M?gcfujaQ5fjd3z(w~OLS29nyAx1oH0HNF
z=%tRkdNsx-C<#?<UK!rM+z{{Vv)VL-e(E0(uTC(`Vs%!W^>VU*NWF<GEMT<%@=!7U
z`4Ae0hVE;ZFDZtodXky_UqY2hgpV(LK44oM{IdO9`nR#SdDfmtJTtKNm@+2TD{N4!
zu+Db5TaEz_iytTYol7=#Um8tJtavm|QDp^(#AC_2-`>-lvzydjspsPl))hpZ@fa=4
zDT{c^=itixYW5n(8s`=SC+5dg(A<Eh1U%~N!p^q&Er6=>3Jk>U{UUD&pm(9|U-XFP
zFi0m)&(6}D5B9WNtn+CE25h<z_@bNDeP2&%9!9$@a+yqOU{&m`s@^gV$*ko8v;GE2
z^*DyFmvi>@7Pms(RU9|J$qj`uQuHx;ivwbd>kYfHq1!s#B}5f!V*U`F$$G73t5Q?x
zKpn$id+v9|ob~nC8dBexi5!97kV1crm7R`$=jyu8`4O041zI(Ns?Dv)-~V}W`y1dQ
zLM~k}nSn9o*OJ~mk)6FgMiG(vek>;Xe_u9Vy7kfA=36P=y?eJb^pzgr@rEw$0F!O0
zCj%d7AOu2S2mfEM_Hmg{F8^4K0DG_faO-tym_6k_@LF*%U)BeiUZA}c?@6)z``Pq1
z{#|o;yWrI72vp@&H&+mq0Q76Nq<8qoe_vHEjEg-71_ujy32%yJD=8`{bbo&M&(-~N
z6=T`lYGPs6x%|MVv|yq|6|lG<m7~th42YTkykZ3-Zy$(tCh(Jimkj83Is$$nJ1>s`
zw2{Dl4jx<+ApOALXDt6(?!VXZzTN!GZ}bt6K7bndFC11_U7Q_YbY;j~Mb*~UK6(1I
zd3?N18T)@-$d|4>hDX4Q1#oCP7$jAKBP5i;aM(UQH`g~Z`|01y`B;@^q!3X7&J$p{
zp3igT-(Pt>Rxk;k2V%o(@aG03O-Xor;;7j~ZIvX(|9Ur>aWQ@ijLkt`dlBj-a28ur
zc`pq!t^d3hLNnAjC=u;xIEAlxEd;c`|9Uxa%NB@zNl)*BGd+!a|EKT0jlTqfImIO<
zmOoyCr^kQ2HXP&FQw-Z<nt-<fbE`Uv+5f%<xCmn<qFu0(p|3IrQU(ZX1ZQWn04}St
zy83gW&Iy9pIvJqK{-66vC6;z+7jUK7Sy><43GwkUQoykch&~!z_y5-|_#lP51n#$!
zE7t$DNa26RD7dSz5$JJ#amfFC-uRQWOSW(_Y8wv?3mg;)92if6F_#m7U@iA5-~4+K
z`WE;hCew%;T5YZ7@o-~fO4B-M*Gv{G{I4aH%KcP~Q?2+?<WtN>plGRqNd40{<3CK{
zae|tFx~0#*_wfFG2jURYE)cehtiJ%DveQf;lf|&-e}9?&hJIXafE9u*c}@Ah-h?T8
zrR#tG^};|O?aqI1f!<61*Z)nO0Xf5~0xY9xh~&SITxucoiyHbdw4Ju=BNSlq0;KQs
z6OacA3=6X?Yx-Z;X$)IMr)v9Reca<iq}|!U*5I(#z?(X|s~>FxJ>wbva5mFM$Ssc#
zDwX~={x1E2Ve9dObhQaNhwBjr#v@Afin5Xcgx|#C2x6sQ%BD&7Dm}_Hj-!2d&ue(@
zGg@%X+L^zO!w~xHw{wBk*?T2iE(M)si?u}b?Y*WVE(NmS8|nHplqw#mdJZ-8(=dYL
zGVV!7;N*1z3@DURuv<0R^hQC$`7wB*FkjWQdKjJ#QVOQ)KW|#LTena}1Jno+IH==g
zg#S8lhONtC+)foJG+zhzz#8)njL3y=;Gs}C%GP3o{rwpwB{rM#FJFAx-`}Suz(k??
zbNEs^1vhJ57X>GWSC1hhy&X_9+%sZEp|VtdxTF+TEyTUQ!NUq(x~+`Sn&BX`50jJI
z;Ah|nEX+7C)#XL}27=YlP`Zs2*t-yA5KBe_z~}Ry?e>0RfDz-<a7w#A1t(CP=K^nd
zZ7NduIg-|5OutDUQ+|Z<7xeadps{`i#(%Ledjyg6-F;US>bdg%suZiB{n{^>>G#0Q
z8{lE{Fu3dlE4iJ6hKst2MGTZ@jv{53$I0mAeb>XkwFh;Z>-MJ!HXXmG{b`bN3#wO?
zMKdr^DDAc&jN{CBi&lUg?sR&bmyB)!USR%e!=dB`0D{@IYbyt_E}=Y46l<}M3u-o6
z^}>XsU|_Z(ucR6c{2~5^(|JxC7$70nRrlbZ2Lrbn4(D8fld@m!KKJqQLLgJpQulNC
z-TClk?OdJhV0c+GxCg8cK9FM^b>Sw5CzLyLhmv=~QE%dOfp<}l=Hw+js*k{j&IEie
zhyirSyKjGLPZZ<@*89?b-&E+0i-8Q3;I=;;E9&81Q3e1?WI23Ec>A=F00*9;euG8&
zee0R;LU7LkP-zY_cF}-lFa-6+*t@SDt*oxPxJbbxy>UX(C9RmP1WqYK;Js7MyNCw2
zuv`c2ELi9C*5t+cK?(GqwzjrwTM6)F_Kohso1}iAosDn0^FnokwCQi#@3jmSqoEqn
zGWb#oYc{IKz_cm${+p*d?QpA=^ABxO0H77qD<_!GX@GpZDrFa>V#J5D>uL!KCDzYI
zxCF~~;d3xYD;#y5yrzaPF@eSAaki?n=r4Nu%J1^^aCrY6e=7TzCMM@~-z@9)2DGme
z!@`la7Q2{Z68#PAqHqS}j)A(ni`##eE81DE&RGuU0t)3h=rbeSg+f`;j(%b50zjel
z$ml3|Vayb4UVH-h!GR+meCLA`@-BIyEeTcb&2wQCiqdGGCB<R#z>J6+g~Be@9>y4l
zCLF+Hh5%cy+p>j6Tei=Za+O0f29_xUXUk3m%ZvN)0D<dp%P_%OYGh=@+3pE^@D;dY
z-ZZDQ(DK;<Oe&t?=xoy?Ir(hOP_0P!A>@U4aUx-EowF5H7dKcjSy(@kG2D7(dEWp>
zoqr)lFPvfWWa8jH_|a)PFK0+=*JDgWAcLfvFp&tWC`W9IV(Brp6N9Bwn?^PDB*cNE
zDQwZ9y~8eXqCWu8ayArzPrd&GO&B(u84UV=w^eMzh{3HxXs5kk@2a+As;LKDWbFhw
zHXwsaxZI9+$J`Fa`DfpADFQr`OpcFEdK-jo=|Ruy%>*oqgX0vqe1f0@F&+EiH$Z5R
z?GQOx12L{SkhA-p^3Xe+;W4GP1kgS{LLK|>e~nQMK7<wH_8Rp^e=IXIGjD_Ih?Xg$
z?Eu9^bX=TULs1>rrntiX?Vt}63J3h*yzr)ze22hlo)T*$gEYtkL+cz4((9aHE5i?(
z={7ymOm=6k@`Gro6bWmwhR>;~)03egG@|uO{6LG%(6s(R0Zd#Ak(1$015lT$+NfSd
zE{Usyk6s##C{Dr53Nc<Y2B18kgnK|t9e`*Dt#HF4mLdg7HVQ%WUmwJ{kbT1%`IHc-
zm9?90n{DvKA{c=GVhTg8bI@MkDl0291~?UH2gCrJb^o7FxY6Y)vXYesLJhSAElB7R
zJ{bGtP&-Q{txdQFi^gK~pIZd*-b^@dXjz4OYSK_)wOek}Da;EePuE3U9Vhn7D;S-e
zJa$1kswU9^X!K0d$jYjnfa2M%7Di5BQfWRo3%_%&=o{f`h66ULaaFSa(k_T}#Ud#4
z@==R6Ly?_W*LgM_>F~aOJDRU0B_*S?Nd=qkuHeHw9=c9FghG8#QV>&RjpTmtU;&H`
zIx(+megL?UK6unYLXoN1fg_42(4qrvumeXo2e{lScn&qxlMwW)PPf}!j<nEGf7J)m
zJSRY(1~F(_02|fG!sLr4BKHsAf%Aao{I7paC>ScSfkC1-7C|epkcutFt0BHX5_=#G
z5a6Exkrs)3n$faa1LS<C*_)F?dzDZ$t*@`2?A4=Cj%Oo<I=5WczP8+V+Ma`-Y5+e^
zt0fF}3FIMz<m5#tL)+!SW*jPHmfvj+5|x1PM;@3A+|Ic8crJ*FV@`{S%WG@TKpE=_
z`^yhN;<+syhAbc@YcWfB+iXi%C%7q;xE^g=@X8`=T%f8&fMcska?<KA`mlE7W{$mW
zyTiYjV6zRf(L$DPs~R@ll2`2#kxdDvody6&sRG-mom3(>W)Q@%EzHM2i3LKm2mZd>
zl{@MY_#BnA($(cs48KEoCt_wp1v!BB`Uev3{_+tJ0TIvCHM=@U0|3fFSVAZIR4CL_
zaviVw)q)yC61c8FKK{WSBr*J(WvUAbEDC8bqZz<d)BT7LqF(-tr^sgQ&UWIat6|}4
zThyI4HN(XIv{V!-MV(^Pt9~KQ^u9@77GQQ+qcz`3g996qSdjA&DDR9nAY<j}G}Ln%
zV4^(HAZ!C9_&21#Sh&;kpphyc4h9@7(r^0|9?wI9?EoaA%YWkCi5-0;eJI~8K#n4J
zImu}u7JLV|0DKB!0n6)0@clWvOL9J7S8i-#vXZX`e^^;p6}#B|BKk+mVLTF=gN;2h
zejp7Fr-Ai$$_)w+ycQ4@K}=-wwCe=mVGcvipl(2+ruMQ9@pd71|I<)j=e5xMH}T-@
zA{mU*W$mlmJ$y*Eb#OPJlG+*43)}m_4ts5H-xR+E=A{XRZtJG}aMd<l53$#`Y8BF?
z_WqL*AaH|Y2T!N)fhGj@tsKC~^zy@ZmgVuKh#vixU%;R_6cah_4qNyjPAHwBxEx#s
zzm^p$ADWB97Pzk_g&MrbZ_GQ4#sS(?3>0hiF;HmbXj_Zv*!<zCnh9dt?vdf#`ym8B
zs6VK}3S@I{0jBgfEQRd?hdMTpC3+(Jr@`SczhYv94Feq|#=}Cm1X0xxnpx>^kVHFv
zW`sf!GQ)5Ee=^@e!URRGwMZM}a*dMFt06+IngZ?0*wS*--t3^z<8+p$YPUz`#MuIc
zx<-6b>Z=EWl;ayhd`4X<s=^+v5iG*-NUhl?g1Kv6F&+h|Uq)$;%Bgo_?kDRmwI{ma
zl-A!#oCa$q`d!H2x+sRfc#1$wuc>c<{Q!ysc_KHVn#oby;v_j=dMvsUjNE#jrbHar
ztDc@%C{J|s8z>YW4qQGnZOZ@G|0t%6xr<bWns)ul<Dq306^k(U=>+kJ?T=xu3{Y{e
z^xMIOD)4q;)V_vPcC;Pckic8N6*3;$nJg*_2~n+BA4v`OAqD}LOdr-qT`?gF#h9~z
zw>wzBUk3UPsiR+E#VnkkY<2=A;)RArJk+(Cj?=!#hA+&6AeEhivm|;tO#~u}u@B7E
zD`vsG1e}azplGZK4S~`IaJRBRtoSEl_IX-PL#TXgQ8gds3eozP&iQt<g(*-(ek19M
zN)VK#J~6;<71p*iGg}j=hc|<3O3Lb1T9`;2<j_+8O(amoqY*73*A7}w_E<#JwJm$q
z)f@Bxs@`8YW`hzm5{}R81#iPMK2*iH(S`IIAh5r)U(w(ewhEmjgf@+8D!3^ckfY%P
z#%A}%`=J^Ee-!7zSpdM(CZCKgI{)2(6e0{+*?lS1zZh&>p26Jxcblpg$kqX|t`89T
zqM%IM3ZsLb?(UCpFByJAX~N}xdI<Fgb>U_K6fK~T@%o>1s}H)}XIm{ORHNxH&vVjg
zNGs7GnjzS>VAtjiCF88L#6zWM&Q4$h*nkRyA<S*m>Rp{Cb~AZ1s;c;5s??68XgI4u
zG6s-SV?ib(eAf{P^aO|0OUM;~-<3;mXDk;JjJA<>h0AuS%XLYdrnUPq&A2I2FnrEI
zhpMhSme*3dLJtZm<Q9UWO<ha!@-kAzWCeA4!B!c>LnA<;FfT!wM@|f-@N+RSc-UzF
z{~!O~zad`LUev&*r2YlQan)4u_w=Md7cPnc;%Fx@;@^VjFtQvLlRez!1eVhaY+S05
zO^YRz0v!D0n--OHEr8ge9t3VeG8{O8wJ)-vplXqWW)mrcwAHXto}?J{zIx-;c939)
zXGfsnvw>lpA&^J1^YdlFYn!9?Z!_=#OH~ITPWE}~c%APgUnD@Kx&R5z4nFBEw7v!q
zvti#47rhb(vwEnLC+*H*hFScdCUDoW4${}^A<Rg(!L6e5Kbsihhlu+DG!Fxpjl?z~
z%LA0XQ<^C;qK%`59cFBQ2zbXnfXTce^vXyf3BEkT>xM`j%^(d5z+r<9M01cC=0{8m
z*X=i6`BM!DT9!t0hYAQsRyUn(tJ__eN;K4Vu${Z?zHzc%w1GeA5uJ#hZ3{@>FD($U
zk=k%_QPeTY-b}s_37`9(`a9ubC3#SwH2ESUM|3)+ug>XtL#@hsmb<GB_#?X=M7X<D
zwN8IKUN0jwJ}#sbwvcP6(^g!rP!(_)N<ZCfI@>XHdN%vl-6>J^v_V03_-k&7j`r%I
z_{znObe+D{i?@A($LsTogC3T5oL0Ni3z8q)<)-NqEi9qI;!aszUETRcb6&J+4I`$4
z%9$Wd_oMlLwsm?z!AU@3;`qhqJ=-%1sL+t|u!g!2iW!}Y<H3Wi#NDu;-J8<0O9z*C
zuer_a-n(!}dAvmCPb9$1`<_~~OK&&ulV9EGp0m<Lho2X9s%76m=EnOHEaxDRZf<?o
zIKNtvivn4fZ@97heq>nA??twT@BP6Lew1uv_B8Zp(<|w7*!!FN+k<EpjTKg>xxj<F
z?R12<1;iq#G@>k4A$WrI*l%D9yVTa##KpzcUYu=iI*&mdgogPm?<Az!LJW*!86Tu!
z6FFO!cMvRh+F9V{;dujjAHX$A$d~~pLIrzJz8RbzZoyI#)`qpI2q7DA{cT9?=uhp;
zoS2x1cnj|2Te`79!Tbz}8h}Iw!IT&nJ#oSRYYcJ|K)kLlci6Q>qM@?DK{-ohve|Gk
zBPd+BxbMVIB7F(}!I%v)hOyV8n*dQ$q^wT9I9<FhU~9xtH5YD?JU!OCS62na#M=p{
z#rcJJ%Z^A@?cozkN{EN)kaA3b5RV%Agx2s8>Xh23|6NGPAywHt6o+G9Y><5jt<dh4
z1TwP05J}O2&hlB35dW7%dv$1w_H!2PZEf|z_nsdF#CjoZ$U{BzT$b?AlbYM;xg7s|
z1bBAb(>ZEY)!rM^u&m?pIQ&NQwi%Zx9gYT)Z(l;C#8Li5OJqPgIX^<GxOj_d78rrG
zQ+GO@A1#75#u(^*@#!4*L?1gnf{Mu3Q|`Y0uMe)%_&T8FF}jm2?`9IV=n7l96FSFe
zWNHN7)%v3HI4H3)EBeTbZeY`bG^GfDR-S;&MkF%8(&AO*-)ZNv8MAJgN4gp^d7+Q6
znZ2m5LbX4Gw|kM5o7)B<E1UtTWlK&jRFTRAiCWPOr*F0{qpUp1x@w#bY9<qHI{29Y
zS=E6-B=mvds5kjM0(pxQ3=w024Ndgd2Zg%Kt-*Kx<AuC%Bn;CY5D*X`&-muVU%%6J
zOOLiq_k$W}CX7>BFi^UlRp@7&a1QbbCU!ZiHc3E!AJ(+(mYlhc@?<Cw+w6wX(FYik
zEx>TO^C31>1n3J2!2DxQEDRH>NGnWdB0~PT51JrxDIG5FIzxuMk)uGY0U^J?fnFZg
sC5A9Oapgz<<UkD={`}voCr>@TMEF;h$dN#ij6z9E%1IPF(tY=T0IiSw-2eap

literal 76087
zcmd43WmH^E)HR3%w<Hi;LIg<A;I08e@Fci9H16*1!QC~uyK8U=?hcK++Z4|`-~604
z_G7XZX%>B(Tes@eIs5Fri$EDEQ4CZ<R2Uc-46$!tf55=Njl#gdt{@|VpM-QP7l9W<
zZE?}B;1&G45o&Y?FDMq@l&oQ3UfVo>!7k~OIl#cYg%SHI@XH~2f6-C(*TrM+$uz@0
zRbfDSCQVR4(IFZ|?I6uX9Qcpcw6R9H#vs`As%EBT(Ktt5U!jsF&SdPRy(EHwCA{j(
z7cXY7w1$tu;X6aT7-+nm_(n+27}%MTkF+?>c$d#^+*%2RWpR973L=yHyd2V=9K-mZ
zZ`Z8e!74xhOc0srn+NWHZ+D=-PWS!qZ3gcbhX1`sKKS<k{4%IWky<0JySqCst@6Z}
z<g5RFQQS+E(NHQ{+Rf{eb(}YEY;Au2H-N5>U*W3$=l1%)VQBwnPy`?N(f+%Y79z^%
zf1ehN7x+KDY%k2r<E3CgY++H6?qq?SL=tQ0@u*CXnyalSTOSP}HAXWWT)ksI4mS@E
zR|AJ2@)*tYW34g9#K!hy8<*+@G@?!4T;;ZUW3+$jBUz-WqzdYYSmeUL@27UDF?_XN
z+S-zp^ugz)WaRT0<H(!UM`)ucP9VDOC09p_V?##Oje=hI_E%~qH^x{zPUleQ7BYF4
zuown-HdQ~nqa{habA_)Ch|tZo!RCj%)1k<ZB#SuYK8BIclUj$Klby}bHk6Z<G?pvK
zv+#|_CO<0)%)tM;ov`dEgqdV9gf(YUF=Kk##)iq~Wv1Bwu`-=+_yT0|#>;JXez@Ho
zu&Ju4)jwR$6R<#jiL#0Rx1g%ICD$+ytNvtbySx4Wa>V}>6>ZdSr*<war3!c>+S-i&
z&tm97gXmVNK7ZC8&y|!&<0UDU56;SZ&tx#HJ6oYgvdH}3qF5$0WICryr*a3@*0R;q
z)cB2DK*zkJqj4Y*$S9=RZ-hEX(8%b|>Uw|w_oAW_=pGt**PDW>hD0+m8eI`X25k5T
z4hz<BvM}Ox&wLAgAH?y64F7t)P3{&;roEriD~86MUHp9Nge)O+VVKZ!r;b+Dez_K&
zfWxsAyxRTrf?U75cS4T(*P2y?7~OJm++?*WG5$Zx{<Vd`wsW&-Bd3^{&RtylzRYY^
z8-wNK;v&+}(1;r0;NXCT@Y~tgFuSzm=K6in>51o_?(NN6oTu`6d61!8y1kuc<icQw
zz7FqXZQaW6h36(IX8Q)~t1sJ)Q*$Q6{oZCa-kb5_DC``$NtaX$51)P2n!$KDeux`6
zC2CL|+q>wjN`uqklIG_Dxse_&%&>WAX*|SD+zWEf&t*;;I@n+Fu1Tfn&y)O|zdHuw
zQ&T%pE1H_hokkAIWWMGO73n-)?Zm+1<Kz4M`7_PfR}da})8~hQk_Rcwd<u9b)ynG2
zmkO2YNwTVNZPxD*t=Bvo|2kZg!bDo1Nu+M?tx06ZQ4(q<yVJv9rVf4NrS=E+HbXvt
zZ#?X(%WFm7+2CBsM(sZs8Oh2iNbUp0SFpjV%FCmiot>pfqo$@7${Jf)S<yUoES4V-
zC3DricDOm+EK;ri_wE2kFyJzMd=KNPQ~}4viN|qslWJ6#Kva5r*oDcGn2%IUt0O4w
z#G=gn*ikKnw{$XJOHFp=tJW)+s5?abt-{BXTofGnU;j3n&Pu_CFe`zQLe-m{A`%n7
zOG-X;-RDWAyrZE(A9_;*<M~oB*<7;U@|5~B?2N-ZlIG0OX6rotNq%$QnD%&cg|$^^
z+wP|*MMleHrMt_W?AU>aM7!4rG9H1{ZWIoy)6sjn<v8SBn+HLMo_4mj9~l`V-%!h2
z>=oDk`1Px+!xznQzr2^T!3gAlLzAK+=*igji@JKNb5&lLGZA#7DdX|rU|8tpeuzlc
zYFzX>3>C8~qJ4|Nl6u|jeEI3|7j5a(p1JQrc18psG}g5Z=Fjm!;t`*Dvj~rIykI>d
ztEl*Q;?GaVtptO*?Nn!ERMcIlJ_`BQ#>U)Q`d|uGaW`QEo)oULlj<*ns94~B=7-b%
z^Fp^@?xr!my57(dhWfQHEm9Ru$*RgtZtR&>D_<_!a<L&G{Ky#37^fZ#C++bn%u;by
z1P`zZ=Ft#FMMXt_Bz0(iwvwyilOVE|EFp#j#^J!kN{fE$lvLs708!8BCWa11S|e|I
zZJ+h$RHvDZ^+&Y@N#5<nzCiOQ^QPH5T<{2sZ^)i=;`ivNsLrtkYg)JF2Ng1whJ&6K
z*!!!VJ(c_F=XWWeb3dJs^X2|(o{*^KS~Z?Y<s{@rR~5ZkO_k8~;ikDa&Z{7iu;@TH
z;thFMj`VZhR`uH&N)nwrRKt)GGtklNUR;Ie%v36_pxvX<s#P5y8v;{Ep?2bp{Co#y
z%{G{ZTcqR1r7r#9U`wNUJI<&ysZ{-*6nneps*x%75B@m`|H+|NV0_q0y0P3;7ICoe
zo+=3BeI*R5R;@>flwR#eF?W6$OS15iup_VnRlZmwd_cHg%ID?U5Ay$>;3pj&UD3@t
z7A7h8w;w-VbEX`w(U0zo7T*suhQe26>5N{V*Fmk#t}YBN_h-?`c!`VUjWc4wLhe16
z#()9UZ{Lu~$%o>X>nkGrh3@8B1*%Obq}o1TuT{^C=SkPf7=w`>e0wec&l&jtO>z6b
zSOfq6cUciv18n9)46&MLOG6c?ZCm~4ImA?<M^OE$VCEdA-t%El)9fOYNxwRV>WH#s
z4b4s{emSi&5M^7>G`Zj;En9i4DdqWg>1U0*kabfvg{i1C3u@n-umX_*Goj@3CpeUi
zpl0@ZJ$RyQWk(ZZ#fFsdjh+uNXV#5YVPBa)+n+^2gQRW2XH>&zX|4{S@>St48i{6g
z_^QEpcu{)=V?;%3Gl%N;a9Wn)rpJcrs6um5X&ls8*BlV!<>GF;?{bz}rr5^;M#@yz
zutyoq)QKj3NUiwk9{<pg7$n|^gsfS~6>T)+ELop0a)i7|h<C}kpnbV_TI5vgsxzmo
z$40eh^tfjuWx5>8dMV??;kn78-`pd$EEW1{D02CRin7VQdgk$6IIniOt1H3pemblH
z)nwl{b*H0aEx5CIZOK&h{lzrn;=B*Kud+sBm2@?7Sf#QXuE94j?b~6RS*0~IkIK4}
zvS;vBjqno4hG~~okuRVF-?|e>a6Y;&jSU&%`Ev<9zHz{-?(YxmWOd&I^B)dTplW26
zHYv7Utp6;6tGZ?`(((0=P_7>QdAYwpxy14O3~kbpQ%c$27PG%lvoV<FnGs)Y;{o*E
zpwIbLiXJ6pHS5vVXq@=U`Q`%sIb`YfwJTxYqzzC;*G{$f#?s8DpQg3%D#OxX-o$=l
zB3Ly2Jf^;Ng<AVH5wnF)O)tU~)@9>QPI#!^(u9J((ur_}u$^zP?MHJ&)0)9ykCsoa
z>1~X@0XvTf#Ka~EDhE@XB-m_<x(YD+BnJ{^adfz|3X;MGe{27iz|1&^zSK-;t=>=k
zP&P^u_F6*$bCGS3kqiD@*f*UDT&O1jlQTW9qO0-7QF%ryLha8MCZgzo3)#&ByU9q|
zU<;kr>vUmn<|)h>JCy2$$Eq6K`M5%HhGy3%vj|tYSiH_1+vjySPE}wki)s|UjLC-o
zM_WNnY`T9fI-@wSk4`l!6M#8zzV+hT<~=jT=)1d2kP<KoX)5V8H66%`ypR-iHS(mh
zkEd~i_fb6xvmVNSc|4wKQvR*^TLRsH2wBu6>DeJFcH?>uGq#VEH1bICa6S(=ugt{b
z^6*702Q$>9{%TJ=K|+V}|0zKJ|6g+Wzm4(#zjy??1Q<1W`JNC0Cc5h;sXU#&$dAvx
z!rf78TCqxP&5qrF2$y_NYjF%i2FG`8O&jzuUnM1n5a#*rkA(oTR%)>M6cZD}eYIdc
zGe6H{IGWBn>z4DG#=o@{DesJ5wrv&`cL`qBjbyQ6it8NBL<2cU{b8mDHka;&b6l9I
zLee#q`q~H{YNJ`~YniqB=Z>>Mt~jsIqxY77$>>6-m}TMSi|Sc8xRU7GU&)PJ+n;|+
z|01P}Zj+iM-xH-`rjjf$j>RVuI&7M<JK&!?l~wz4nfsoAo?u|Et=#(fxKu{X*=s_%
z-K&s%uux}y5|>fBL*Yo!io4FaqwW)f+h;@2NmR(I)L8Nw+qryG()%(TMmmUpM>vhy
zke8J-3E3|Aj$d>8<wr21ERIFBt@g!C9R4J2c2?39gk%Z6*<_Y6S2}Gt&Hd4_y(TmB
z9jobF*ZI!)7^Iq=latYUOCpw5Wr||H^w*{w!-gY!44M-H&ICWt#rqkzsw|ahPUlu%
z)29-w5tzx%G8;shMXezSk+r&SOl!d^0lxzLp7tu)8E1`Jan5+VMBeQ&x=fb0aHZDf
zJ*Kg|O29}U<(7XThk_+$&SmTQnvGWrN=AMGXqG4mEwA4HjxO>u@F@6veQ&I2Q{T#x
zOye`#I^l2NLCC{>iS$9=UbS&fuphS0p4j9Op+dSd!{Pe+qvElyXl>D?`i;}O<%b)_
zkDE8p3r^JU6*q>H%`}e}r@S?%iJu$jnqzc)37~H-?&CG&1kW2NmD0*<vy!GEB&6x)
ztq*qcrRqamM@`o4HBpq(YX=9R#l^*|)}Rx7`|+c1Vq&7adFzkG8*>(RHr>m212s9t
z-Sp-z^QTBO(_yS}3GgU=80K8bh@wAe=TmFWx9V=Qty<(t6>DfsO->s=VYsQw;*@Cc
zjCB;h8Y+j7;JjgLGCe4bH5E^4ba!2!PhBvqx2y6;kS@eq>N#3$fXrMoeK<I*v#b7B
zKJI=p9Z`*8rc%90(*K!m8STWfbvR4J=z78Z%x{`zA+ct9^&_!uQ4!tXDQE<)Z*LP_
zhWq+ZJo@|l<<0G%`z7_f@h@fz?47-Ga&p(#*Hvo)oWmI*J+(0kN+l$wC4r<{k(IGV
zB8}Ip@626bg5^7m!@Y+-o>9)bd}<_I&ZUxxL`!K<Ea5TDFDob;GZj4xS1S9MS-)1J
z=5pt$i$A)|Xli5~)6i49Ls%csOdK5}IG5p9YgA0Gh2(=-Zjv<jH}^rDXfZdvZ0{JN
z{VrFfOxW~&YlTniFss`@qlAJ{`&3o9_~YWt++0;)P-fMRozC#^u-_B?r%wXQilwt3
zOa54uMbHaLxZ{r@`5}6VQVvuVHh*$bWW{dq7iz<Y;=)bUNH*`j)+%nkzDz-3xWu%u
z_Iq@F%lN|U{%so`F$v9hPb_2?@AO1703Qp!NAMeZ-Bp=HqjH9_#t>N^E5IaQnI<OM
z%Gb+mx%MS}k}2q_OzGmR5&4Z+_POBQiT!<wa??s-;>7DM>=5^tVejk(?=O4O@?fhZ
zllkb$om?@D3|@cx9{lmdfd=Ih9$)y3;Q`G=MXan0yIP{SpgQ4+<DahWoXFpCJ%Nz}
zO}hC6bUQEyJ3C@_H9J@tMMdbLBHrHKLWQ+Nl{Gc`fB^9ekE7QZF0Ngg>k-u@SZyg|
z$~igU@GW!kzlA|Q_tJ9JRD6q$0>cJoV%&r^pK_q5)RT$}saJc_zXw&dL8vri^gU|B
zCBi|@?Uuv1!}U=^FrTRGiYmFuVixbNgCoDi+w+~FOI7NIaD2&Hv&FgE=B?u$BqL|S
zW%Xn<l@?ht7<Es`Z%opx$Gu(C2z}Px)%#8t=WDJOj;81i3P^I{#iYc|rhF>NJ+o|T
z1TG{a|Bq)?RM`{E{sKJZxa85TMu-T!fak${i~OfTzii@}7?eiFw}o9z(|or?gXsPN
zW^wAjhlhSRM_wIyJT$(gZgxx3p@eRr=X}hC*W{;N4<OW=tw|J+JtrRd{$b~uBJ!U$
zhlLAM&Km^31GALI2#tl~?r;&E?v->Rn#l=@$%zSzbExcEZOMDLD=8tv;X-B}Ji7|Z
z^x$Tudc043JG$0^_V>)4A4$@kFLs3?kk2eEEQZFV4ALo_!itLcgrT{UAm?d3!gr*}
z_NJ#CU8ubMLNc}QT`|W+^@rnkB!;-L!S&uY{QHi73D9e71vT3@W@cuw#xB@?!#--k
zNp7h<-+4xGvkSu?sN27zrMjlOs1lp8nd|-iO-^xfLQ+tO6hf#=7_#@(w~eV%%a!=P
z9{wMbWbi`e*+g%y8u#gq6)9{}Q&T`l$Oq8#1_U6#d;cB<8QHMxSHU8*Ky&sd9@YpC
z6927$L`jg<4{|Pq$9vU>8LohiGJN@8-3<DJ_ziXkb0tH(8d`GXe7Qwy^Y53-*$mS9
zY8XajUXM_y9TQTAzFAf4oc-Mrfw4nGFukXoN^;J;s@6d>x&K4}=HtVM569!-#!8ih
zm?WMBXZ_8)QddE6q=&od?Q4h?+Y@~U%j(wAv9T4V^Q@Pp*96*6Uh4?E0?fJo)X8*)
z4k}M{7ArK|<**&iE5i&I%a~SMg9M<weC@6^o*>;CjJhB?AkjLOd&!I5;V-0#oGj%n
zN+D!hO_(S^yByC>aw}{zD8S4wivxPPzjI_dO3mD{@V)C)bbJOghpP)I<PQ`tciH@8
zp75I%?j5e~2u$Xl4iGf<ZZocQ4E7w_w`nt)nv%1#Kdh~<2Zo1Jl9PLOb$#t07?4bz
zUL5mAX%Pyrz-sGkVQNiT8Ys=X`td_eDq9rhZX_qee8v8<c2B)}BBLrvoRhz?q2V<F
zL0`}t>bi?5l}~_BTkY`WzMRo9t$6sZSLo|n+IcQb&LKq7s%MZj#b<!qX8d!-x9?c=
zb<qOCys<b-aZy`0Y*?yiX95S!tt}%U1;B!8YikDt29k4f65_>$CDE6Z3@c^*aVN91
z@ZHn!3Oy>&QZj~TncTHbc1<t9=||(&fS>hWo$oAnxf5W%yuAD>DvFJrg;)s)KP?@d
zajo8-z3|RjzD_ni0pX7`vmMcUtEnmz2TbNa$n-YLxKLP4JX<3two|tOvv7aAS9*VI
z{E(k7mBRUol+^5XfsmkJS*5YRicfvsN~g-itwAY_1GM!xP3d<;k^1ZmXWRP$a>%=6
z4phDt^!e7G?)kRc%As#;th|$%``x%YCYB#D_(2b?(-T5YLE%?e(hx+aHrL{&Tne3y
ze$`v+->NfoQvGWx-YxuT>b>7KB41})iEMUUHj~!HlV!sPc<Lf#E3CC?F=NXL+qIE%
z#mAdzqI<hjdpE8duIVe%5hm}{cN04NNwqibzUrlvpFRc7%qVK>=*(6bVG9f6UD|MP
zcqXCgZsudR*EO~eMXl@>PGr0|az*93DwroGS!s9HR74svQ&v|WzC4)6=&f<SGP7w2
zo?w|qIfSzp&F2|#)MKz_iJop59wx*M^CzXVc#9PKLC^(v@B#i^;~$=X@N@3{-O0(s
zAcS?<Yu<~7O_*0zDmX{H_Y<0tff6BZ$7uX3#G;u)etJO1JVd~vue+oOAYDE*ZN(Rs
zWp}wQTb+3ZSF@EDwt^)oP52~s)ESRod~SMI4oDpYke1eKX+Hk7XXi6<i^}|0$<9$1
zsgc~B@pFD`@|RAr>GD-hQxo^=*RRn9ICyzeJN>beqRS!7gkD}=6IWP;hhMkqU#+f&
z=QX4GqyBT`cvMlBp7HV$tZU?>7nSVo>j^90+tw<r7S1ab2=}VZ;%e2nB79JDrT>>X
z;4M<GC|ib*lbd_mfkv;&(5)&b_a+#Zc5Q_3sjIWoyYZ-HUKewF%FJ9s4W-xlsJS;Q
zoOjL+X%4mC()!cPm3pwciX!IvvXAx0vfVGphw81gOk{3z@YMD~PN}yXwZl{AL<}<)
zZ<N`L17l*|z$0NYzlfL35*e$v-ku^QJbT<1lG+j8mGUU^F1u8ME?}75zAQSBtW{8G
ztC1$x!&aR4ERItCmr?#>fL+zOE#bpoj!dI6+?JntQXzya{TAgf2?!hoFq-SbqVdL$
zN-4fGDGQbQd-rP2tIt%S4R0`CGb14(6%`kMqGL^QYnz)(e0;b|I+qvy`jvU_ol(1%
z17nLP21;vN9#i@TF?v19;oWQEB+IcFOJp|}1Uotf@2b$|TOkDIFgkRU1AnQ!_aOQT
z3U0^Bw6wJR(q^G!St_=@A82VqKZ?heZLD`^Zx@aq&Kez{!*b^K;)AS;<K2wNc4z#V
z>fY@X>#=%U@^YI3j-k&TP8!Q4=<%?9Caj{@vcVVe-7oncH<_EcJrxO#Wb*CBjf(ku
zV})qymX?;uT=pgH+qah<`3zK4UI&ensrqaRbw=+(NZ?H!UAZ0oQm3W_e#y&MSZ_;p
zmLDwCk8jw|x7GEw%@w%uYd+CIV8TvWOF!LJYW2B&E8@{WKlS74%y7puI}Xq^y<RWo
zq>~Yv&6P|-Z$DjW_x3Zjv$G4q|Kb%NkH2Jrl3OH=uUPB8-EGjW7QdKfd!!JQ`8zUF
z+X^-=Moz<Pb!Vq9{)-O&Yf_=S1^fnk6v!CKQSGBMwMt|xlg|F$0^?RaMVk9ir{;23
z#;2f;KR@#%ysTVYxOaATC`Kx4YblwTCq~CWjukUB%-4F2*usb2O0xZJ)kr;1qsRV_
zB^8xV-No8w{*J^Uix;CBLIxjiTVEt`jg(>=a;auzD-z7PjbEHd&3^U4ek96hTp~3s
zKfDPya%JnU&a?1vb#^n!(zaCQYb~~-WpiQ27^g^5UKe&De@q~p)tJurw7A}(E95E{
z$F_ST5y1`qjd(Aq&t_DT3s+V0T7Jm0L!xfg_88N2BM_DNk;IPXmve^mSAj8t3#}Yq
zUWLvFtb6$IHQ6Hd=ERG=>7PF`mj)&cCA_OEE19xH!&y~neEj{1zI!GXF2bKXy4hg9
zHC@+#ZZM{bRWSG{0aG8f^n#6zjhmZ0zodlX{rhz<bvhuob@o*rYLJnUMX`CI!P$gj
z%C{i%J&8A0JKe$IUj^I*6DRW`^EIJ2wkYr&E+9zSa(oy51oP_EE3hX}xwyFO4j1}L
z)Edzb%=<{9{{8#mcD;&(ORqi%))=O2a&8T>{aq@7kU_*gd}~%fnyF}B<}9o`z^mgH
zyY;k`$K2|zGzNAQ{A2j`k!xyx9__@1Bf$$pkjShy0>t>g9F8$Ps?}6iZzdX-X#L9;
zdvX4GWn4OSHEubh0##LQ<p_Ia4&OE-x@Ia8Momtx`*5)l<XoLzOrV9J&xei(0Hx`d
zf<hJ*Dj_vB4jb3s!e&>}jSR%?c3#?%W7OTY&X0GK|1dqsc|($!e2Z(F%`f8yMn(+I
zwubz4#3dw{O)9Ia^=8Vn{8|Sxg#5M5OD{dxdsDM#1etuQKJZfqb)2mCVzdUAmzV46
zPDtBs9v%+MDX8hO*Zltdl9>?0F;7!;9ovNYPuV#QR?~7?MgKcq2Sc(JHTst0zc;YT
zdE=JX$68Tk$-jT|yScUW?JX=UNDQAvL`1wmKv1hXquR_`cHHb_yJ2O)ZO(kTKNoc)
zJ6>OZMO6eDxxzZS4hfH%KA!RDYvMXzigkbgfnUR|qzeY)c;<&<BNk~RDQVr`0+mU+
zSNkhkfRgB{G2LyFmXeZ^NMQ8+^fNbod?z3{_&qCYtdGx&PO;^l9zSv+KYS?)koH?O
z+w_GJQAZnbh5mg;CdGPxWPZzT{Znf_NhTRV&AZJU=R>LQn#riLasBq8b3B_lsam7m
zM>@JN7UQWhrwa(E_<GlbV0-_PpIUCPx3y*Akj2DjNbgJ37Ds~e$WAksMRBEH!aSS~
z&=)CH_G}0W6i#N28A=FRUmY%<TwV3Hw|jPU2u4Ol(L;+b&(E1T`VOtbRu1lYcVoob
z39heMzb(BW<<S~OMJAqV@aBlgN8D+3dodw%w7dHqZ2F)~O=IJMP`0JNRYv1wKy<w4
zJ)EuF_Tr+j)#J3am%tI$>J1~IQE!Uh9Ec+d12FW3u<SP}`RxtTRi5N$<nZn2RWTRt
zdne(Mx%JlLo~2rLo5Pm?R2hxuo~#AazsA9t(7!Vp{fM8Qv^xgkL^yApK(G=>V{8i+
z@KF?!+WRx*zvSdjhPjT8+wYEkD9d|V8?dm4H!2r<LFk82MkgmH0ZQt9*#we<h>}u*
z*4@oEzP-IY0FM~9xvsb8?PFsR4|i9E(#SUcN~;W2&hr>>zhn$f6)VUhtzqLCwBYDl
z?-^xfWid^7(f^khAhe~Ztn4?fN)5i7S<eyt)YKF@3FpM<4TH<|u?UwuT{n#D>4sR8
z754fS0v;Y-)RpJx=qTbck}^>`Y|RJS?c%4U@5`h@4tHep<rR=R_`J*loDt~P$ndGF
z&y*iOCfu9BKKJ{{o$uehJHEIuwZn2$RwgWMz9iIezhGgxCA@90F>u>BSXkpDEY5v&
zF8jC#lQqpF>`}=U7{%8;*6>Ep>vcLkp8~{VnBI49$!E>#`o<DE{#O^nP91*7F^Thf
z57asur3#<YmfN7gK?xW*IJkHYYYO`%M@kUA6g*gBVi(>8CZ?tYye@U*ZRkX-MeT|6
z!3fagtg<aJi(NQ2*CM*bMPfO<vKx5za`=7D&Un#`BABrImG4seuplWc0DQrHw_j;<
zv_t}yT|z?lVt4X6wemRcBDGHXmNdLXU0Ge#vU+#0-1>w-nHh#Y+Vh#3IxskxiCATk
zmJ9UGg@x2PIXM)yVRz>{Lce}RYq5sw^oG46BO6IzGLUF3OL^~b>6g4NBZuiKTCh4>
zjIk}l_bU8B`s~k4Goi=dzu!7KI%X>L(2R_XJe*)Hwo~k%iLeISy`JZk-5S2V{AFjU
znG4)wWAo3&cXf635E8Cfz!X*J4dBA~-C(TG{>}?H<VBU_l2k}nIKNM>@nefx-ukxn
zs;H=_4v6`VC*61`=q7o8-qLx(EdB~l_4bAb508U8HBwGOOgxw^78TEI^!m9%6;4K`
zrHwqJE<kHaTUvfGGfRoLFhyA}g>^<7yq}cunA)-F>g^?RyQ(o;gjfw**x89Z2Yi)$
zVfQk~Y!!^1S8bD%F?*%0sS69LkJsy1IKu5L`CJIKNF4#G2W-xMaQ88Rd;fHBh<}Y>
zu`M~gKieLm<mOHWd@0xf`s_8$&R2=|Y8Zsfe>bGESMA)f>3m^QA|kZdO8^KDBjYoh
znX0!oh~kO><!7h?dlnZ~Sr+H9dY{7XL+M({30ge&t|hAYNMtO5^jyM=f5Tl-i3HO6
zOR`aodGhk|FOZOu7KB3r0@^w|vxE=WE!SR@wmx!-$I=MojlWvi91wpn(@K1jQ%u9#
z9k7Y<DRufCohiY(cub*SYrJn~$t32{a<BRamQeG-&0&KY`ONoCxJQkr)US0IwLD-B
zSayF(=?=y_I~(EqDJSPElBEUq|6n3ZFp+uFD`#it;Zz>u*ZF%+XZ*#*1%<A`fq`1b
zEA4=OVkJyBF3_6H+Ffbe+baA>pf9`9MmW=~NQ)QNgs{cK7QR5j<M?nfEha79J3Jib
ze!ot}eKyGC;5;pz5Sx(DH&<<})a=AU#A?EzI>g1vxdtd60RhjxB$w$DH4MsG;97Wz
zgVQw@6e<3WpXeu!FGwjwPK%-RH~<SkC;nh7DJfAj<2~G7yazp1em+%lN(zVDJ^OO^
zfzHzu-=q;VM9C<_nc+OhV<}=cOW_53brb9CJgofHOy~X0nNqFPgIgE?aY<$kfuQwm
zd3tmfq5?h=oJ##6?f&RbHirud`}-zK&PP1*MT%OdJ;=L6p@oHo8M$_Vrf9m|Eku_l
z6dWqoM_+=#&%I(Ts=o2PIoq}d%u1CqEtCG>iZ7me+wyX&sr{4y>@CpADPX(nL`MMM
zC@n30dUl54E)75x6lzEzmHf9W5PN4lFMRdJ*VotOxE-nK?ywOD7k6!Is|Q<}yM6g-
zBuVJ*uwtfMGBqPFS7Zc$bPPQBBsOzag3*~AalqS=^!4{Ms-^=I2==R20>Z)wC@3g?
zXVB_t^?^Fe^_gm8JZkwOKo|%L3Od|u#wpFR1wW-O8bz!(wDTm-=G~p#pN$YXwJa_!
zexjmc_=FN16vS9xsnZnZ#liWG-#||<LA8y9OTHBv5fv5Hrx>4~pN|{rYh>ynkK?<4
zzy2Y|bZyy~^sB;8-^wUTgzceH6XSIcjami~4Whnua)=@(28QWW%5rXYUfvn#uqzG6
z-W3!Uo}8VncjIa3Haut+u_Z%}S)}(Lz3e01Hau5QTo^MkVv%BDTyN?l;j1#+gf!ZZ
z0*}wX)=_SCY%}Aehi-0emII0&05=JjCS~M3;P)}DUF-pc8K2!EJ%f#vi)((vZe}d&
zq7a({5cZG>ssg9O?s^SUHMq%pIfIcO-xOh8OFhTJtM22iF*|!|)0y!E{clh1q~|TI
zt=9+WTQ@GIH7J>wu)$6^Y<+S|b-OicYLdva{|ww+;e3yp&-quaaY`o2$kbF)Q}Zq0
z9+Ek2`*kg%K<$<td#^M-fbs6mJ3Q{@qD>zn<-w~VBI36ysI}e`>bJ7LzaNs8M#h;E
z79Aap24`z0OK_8r*EWl7<|KB|&Y1v1xkk<-meMn<7^i-r)o`d0o;q}!>cGXn6}tO$
zG*=xQr|uNE>~@FY0ir0?af_g$qC&CB;fqRb!Pys1cJ|RmdviucM)V8{X69(nS7cJ5
zg1kY_GHDZN13<7J22szqyVZ~?qV~2D_;TazzRo?wSGzWxEq8}p>W-Ule*E}R1~}HO
z!Gysiw(zB9=h~rR8k}?{@$Ug#3GMn^lxwdqi4Ji*I2EUtm&uydrN4g<%0R5M-X0Di
z<4YBbqWI!=@362a=^h>&9DH-!iRB2Y#ov*U-tKPSUNZM&4(ly22ZKA<%O}&Z^K%z<
zI0#ZHe%ZQzsc-n!*4C(~^3zyLRBFut5ij%W*H;LnAVZ-{p;SGDlqYd_7XmIpx3_|W
zhfe@i7BmxEAct+wo3{#=PGf^I0XBL4%~n!*Ma5c;ciO4iWtbfg>u2M)H#DwckmK)k
z1|ZM;Qb-y`!Ut*QeK>aqh#kMeLgS?Usdumq$He=~eb6YBjr-~t04#}EIoR15tTw(0
zX_^(z#t)*5HY7s+Sw~_t_uKLe>Va@Cs;<!YSeAQn{dM(-#IsfU!y?bEwqB0u)Z31&
z{=a`$G@fqpfRMxigoX3n$zVKn2T!iUI-!T93Cwrg%frc?A*y%mJSp5_w|53MX2JET
zAD?0I1`LKpjbchnR_1rw@MA7>+$_*{m4j4R>2RVQOQ!~@qvZu2lEa2QI-pdA<>l?*
zjSLPlqfGRW_w=Z%35;^#5HFG+x>{IxCh^G`Od5s;y$f*0`gc`eZnNxqMFJ?PjLb|g
zUo^baQul}8q9Ph%VqynqhQGsUA0@pZS3b{4qx}(Uv=IJP+10^3-v1c30P<iaDO=qs
zDS3-OtKsTU`{2?;i!$0n*eFw0UQW)<O$<_;R4R8L=stGM2_VcXM2HVM6Qak4W!7GU
zosjqvv`Na>6NH31DctGyK1SxYwJp1EzrOE*ZFnZ29Cvc!X37NA`lvx>e-4xsF8kA)
z?X)DC%#F3Rj)U6e;b(;G{&+_A+_>@6B01il^p;mwvkc;E-k(*C&wZ(yf0^kj^6G#o
z=x@oZ8RcrzbY6E{%oM3!U<h>Fj8R|*m4cO(b-{L8?K!qnd7Nv^3=VM8E5!!_xNH;a
zeiC%0a7uo6dl;GR>k9;^htvL0LPP{nOiYZTy`qF!=NvkQF+8CD<3|S=;Gh=*+TVeJ
z&e_?7UQ*`}U?`jT&{Zu8QX_lHAH*$q|CJT3@tdsX=4Ol>713}~mK`NBMqOl(1cVAF
zPr)yU5BXTX8LJY{VvNhn%iHvLyXOK*=fuRsXsx-%)=1h2ZS8{g4~kJb*oFo~%*{XW
zDA*N~x7)k7-SC1!LKr-Q!otE$06GzKJCqF>U(SEz$J-QIJG_lSm!7bno0**@=ttGj
z*2ZVl{njY4IjiVFS4y+>RsP){5rG@sOIRD0ToO#RMJMbPl0Xo-Zi^jC&Fk>5e?D!r
zppY?IUU%ZTEY={88Vo0|f{9<@d%D2_x-ggm@9i}ARHaJ&ghx}>m+91yf`Ho{EzCFB
zXLq)xOf}9#w4@(lZ%?grkPJQ2t~#)nJNWq&FJCGW*uF@_PTzZGWb6)<pBwMni@mYE
z3q=(bCasPa`a_A`z&cCRQ~|<hyImU5TGyP_;C04zQ^E}QorgQ~wU1|`#vZ5T8-&j~
zPdyb_Y_P|>g{7y1Oe30ldcuy598Zr|t&Z!VY=8Uu+QItylexuHraG);R_pxv^T*FB
zU?_==2(ir3*%_PD2Dx=Fh@6cqGK}Bo0p?v(Lw<1f*GvC^$KBrIi8&b4A^?&=lSmLY
z#I6WuvxogXcH}L)^@|_&g_RZ_xA40pN@V$W!rtDMw?iX5N<_J#UKjP_xm-2n3h?!^
z&w-thoG^F%Hd``@^^3tUag^+nZH_Q#eu#OTqCwXgf1Z1Ddt0H|4rkVKvxT-hRasei
zbJ3>}Fi+GQ8&aH&Go`V@p~%J1G|!OP7_J+A_3a(^`?4Cy;xGSDNgV|>wV`PZ_o2el
zmj1pdO8Wd+^9WYdQQ$D`$}yPm+ZreCD{PgqDXY;}v^RHZWdAwWOz9j>%2fZ26vQ7k
zrIaq9Cw%<;SUEj_O*TZD?;)CP*(C}fVNToKE|8$(x$F-Sn#V_30cd^8B#tp)8!nh7
z#jk(|okrrchr1ib25#Bl45T^Wt}WN;Wdy6&V6)S)J6S0CJvcKf3nNFkzP>(}B?lI^
z>HMEKFd!7STY0jgjB@f2v!V_njR?k!K~$Pm-*yD&%XVI03?ty#kU}q<<V#fhKmzBm
zfyyLJ?$?TW(0%gCUW6!0&A|3z5lP+|U_87d*JclwB<$s*d>mqc|K{A!h0y8}Yb#KC
znqd8Ddy?At<YGU(RXRIEVRo4dj?#IY^l_nMocme;nZ=1DZIHtwE-9_=a_R;28Rfa0
zN%G$9GXVnT_~c|2*oD_qoezz!*D^g}c+5u>?yaA!4iCA)=rgygipVh{04BN7KpqW(
zd%AwD9X07IcNkEz4A>k~8YZSF^OhSw0I7*62Cq&}wSfWS8BRa5s@%(Uhz;yWc5bkg
z_627gdm)Ig=WV^r;M*vF9pu>U>A9t+;xMO45o9H>ni2wm`nigNU)J+7Qe0g8=6qb*
z5m-CX%@@LXZ>d0$HMI&`y}x&vD%CKLe>Wn7Wn_*7kHoJ@f3g1FsO*U$dRkh`S6U&9
zboV7x+il1~1<_Ck0RcfGg);_#ei9-gB4A?P&Fi*4^E}>57dNM(_vutwxmLH`B$P3>
zx<?dz`T|MEBt*Cr@hoSr<&~IzNOJ75=~EAV)BfNkk*i{5XZI|>J|`-__F@2h{31Op
z*Ly6d)$C6|B|clUU(TQk>bvau9tQJ$iT11gFN_7H93>N;%1~MYmD$Mpm~~-F^bzX{
z%L0rE2$rp4oJ<wrSuUpugcMp(@EK%<1>GUH$|?r_H0n}KYbV4cW9N9Rtu2C@S)*$7
zgQZX{-lZ9%YGRJK5wc$CNR7s}aFAfN3hGg?DVS%s8qj8+Lm4b?r|<rHIG@{?u(k5*
z^f$t4hTZH8M7F8x3j~6<0;Q?Da1OoyCe)z~Iqi*h2aq~^;A7S94tfJx`Rw-gVnApZ
zHRameo(E6O)==?JQ&})eV+^1u{kFmV9J+h!%r4{;*AV}`4zKP{M#Ue;bYloavBiZ0
zP#hH*`cGk8hby3Rnad8j1;g1Wp?rAAxn!W+_nGj%@Kh@FO+UVsLl??ylcN1-80?r%
zlsUa8CMn61KxcN8>R>gTNH=C`zUQM+dwg5APZC(FfLyap4;c+uzH(_7Ej;`<2n?-u
z(}+>l^5Sskvy}#pcL%jM*X!Y*=;`wSZhEE)7VMW@@ml|~)AJt<NoqlNMR&GVdfJm7
zL&c5F@ra`S_dtf3^@V)$m%`dh^+~NJDK5qswFOMm_HfEG)AXE7{dPM1&<Q%!>=(O%
z(jElxoR|l%{cAU;Ul;>y?r<?VyKqzqVipvy`a=rd?WeRmdl2Tjc1(2tNF>bR{bb0C
zeb>?j_=R{7*nqD;-=B>FPGtrY#fJ%F+uAUIVNHiqxTc}S>pg&bDmUpgA5=d7(=aL)
zJH{CF0I#JOO_*3YBIr^vCuG`GpecQHyR?)4(pIUOqDs_bsn(qAXRZV}&=^)b{VQE=
ztXL*!Ze#TLO8k5-WYnEAG`Ep6rUjvj7FKed$L)$2%MJWR)re2+ka)~MPu}+v(*%Vy
zKlk#DoeNF>VUOlgCJrX3h!##V0$`OZk<bB{xNnk@n>%@FtP@-C*EcsOJw)afXG3gJ
z(a{sSeG!(=s9p?Fg|T*uw<bbewH4X3vY6?8lRs0x!G|&4Ec}&MUNKv1KS@~a$zy=k
zM1Oi;V<VWEnji0%d<6^7_-JQxbd<5;+kZX%LFBxP`0MlDz*Daq8-1!Whr*g?kY1Ps
z86NW_f_u1bz0+s1+fo}!jL0a7-hHmcg9%LR?Ci%Y-dLv_QPQl&H<w$8(EZ9>PmU2o
z$yg~Vh$beAe5duT(M%K!W$0AG@d2!N)=Hp3EjA%x4@lEWpkH|I!4uR^N*WK>0Amdp
zO;)v^b3XbAhRtxj!vqLh$+wfpur8d1o*W(G7=vnrEm`9WzQ1-Mml|$2dOv>tR8g&5
zE`m{0M?`w`>QaMXot4Aq;z3;z^9%9)u)Kc@tc)I)R||WT6cv5o;F#47+ynSkgW<pc
zFTCWsf#~Y9=-L74DD;=jPzDt$JzYzz&jbdy!@7H&*|Dpi>DCJ<(Dra?6x#r-`f6nK
z2^8ba1cOv9;8h+Xaxuh7FFfZ&eB$=+vhG!KK5#LEa%6w>pJE3}N7_>i492e2(g>Zu
zmgn*01LTva3WA{ypr-D@CV38Rz|im?q}<*uKZU)nN9k=zpmI`t_^^@NHaX(4d^3LZ
zc~|>$LLelcH!9oCY0A<<uCwY2hDE0F+Cm_`1Z_0E1?9P_-JhjBbphoJ0S))~@B8q0
z)mgkZZ%)QU$;v@lM@2=ou(j1YT&VYJ7su0y?~#&Sxi*@qL8>|O*3BXib!q7M>&ZX^
zXJSHrBIBA~!iZ|w+%SD5C3=Yzhc*&xN&r%p7Le4zLi>YLCR@WP1Av;*a_D5E$<8eF
z>P*r)f?Py6|9F5S4CTZ-Zxe$k(hdID4wMS>jk?>j&#v}t%MkH&s#EUKG7KOMuNqQy
zw|NB$K(iQ@72vm@xxfQfb&rmoo*tQoFGl}90Jtfwo2{Ir0F;SD7Gpvhj>IkA82kgd
z8ATQnGpRRu(@47^#k#Tjgf;bZ^^E_VeKPzT(o{wfJm(f-kcWzA*~Y$k&=nMDtC(so
zC8r7qUMij2THh)weJN8>=()bstTMOXubMJpw+yoV)GNlR!Ur$@M0Y5Gd~P6^PUhXz
zT=vf}kp;_U(YGA{P3h+b77tOH=OS8*uzSmphJQQN8y+BiG5@ywK6XyQZ!gi2{nB}5
zWbzTWcusboyyr4z`H&Gs3&KlMe>)NXtI&V0L<IVn1nd@XpShj6DkHzwQTo%-tQqYv
z#Z<o&HHhLX$-fYN4;{u+m)iZeWZOKLI;67F51vLvrE*!*qlCua`)=1bK77CQ=yKZd
zEkiad1C~FmcU!rmJ0I9?fbhROS~jo17qhZDMewGEhl8V24C%@(AmVB@y%SjHxS@17
z?u0dDmUlV*3(IXB*JUQ<F+4J|4F<OPaJ8J5o$W)X)^I%WQ~IP09*fy|PN3}re=G45
z?b(&<KuBBb$%UMB)p;2qB>>ZBtvKm#5cD6AYDjz)^4f|!=PCJ#06tg;7*HE{JpDY`
zQX0G}*>n$)S|cnj#)pi!ags~0CKAMCj3!>KK-K5DNKluwR`wC1ZYPnToO{!y2Dj%s
zKtc_BrouqoI`w+9kBWxo`1Ek)c+`AFMMKlIxR?U)vzI@qQ~ucIAW;euXDN(bSD~kq
z^bc2=!wqDi_Y?diEHF}f`C6Qh=?%K})w1D{?jyGhB&6uWDRVUZanaEX&X*=F50@&c
zs;bW<1z?0(9i_a*9jQ^1d1c6Qf7(x*pP|Zh+Dv?aE@6cog72(ovdQw@Q$te-Go3{R
zl*oziv4;yN_s+}%!LchRC(hQVN3Jy2%LJ{6m@|UXW87D-+864r6Sy52Q#k*aH(|sO
z+u9!MLY7tPcnk>z%!Y_)a?&lDj(udYF628QelVn$hau+#TKLg3)5Y|5S;Wd!(e}ZR
zOJjH&F*z14#$kM?_$$3lnV)wL<^K%vXNsv6ZZ20oFOz<Ca;&(MU^V{vPchCad4G=6
ztyn||L^Z{WTWcYH^LthLCLybGn8n<jPZLN`>ziDjK_wBL&-(|H4BT(=1-&QrTCNWK
zTi3Ln1Z;QAWxOxiELa#s{XC<x2=m#@{X>(;o9*UyK}0aP-r9g>2$-(owfHb5qlZDO
z3JeF0!thgOT-TqL^-TS-eP5=@k;MmxBBnZWP9bd}(?h2!Hdj>|qMy|(4cCh|K;uYM
zV!0bGr~E`m?JsmUVHNVgwQ9kQUmCVcrn=lo%=v<|+N8g!2d*lU>Y0=A`dAWb5Dsh0
zVEgS`fS3#>8k(_1^>|G@Qp1-$JGYM3Y3{DJmg$p!B(3RVhIEAKXsD>@M{1x0t!r#N
z14Za`R0xY;*i-)x#%SPn7}N2#=yBPe#Hr*{`(nw#fDVxR{2t@rnE6F`Ri9$#l<o$8
zu1uG@!rPDl2Fb*P#`=*Zp))D?&!3?P_4b!8B<G|VQvz&pnf#A^b5`LW{&K>^h-7W0
zG_?JWSRK4><>>0_ieG?m+wS53&x0PoEH4ldPr<GNG^k}e@32N&JFscwr8)&6k+~du
z6<+JpX8({yxx*~ba+>G;%eH$Pge>2A9zbPb@4KM`Nnf&miJ9XuoUcI97B^qgn5t4r
zB5h-16P%auT}o<eE6IFoG(+gCh{%7pc=^(z{frQObeozT{%)$Hh2{4-(y!@QP4))C
zjSB!TTBQ!Ud2x`C9zCvYE$BKo1{CX^zN|l$Ja-pb(9sM(_2VAMcR=VnVG^@z{)mcC
zPBy7X=HNeCY9`>c$?wyK2Qmvy@*SD{;e%%+<RZ>MX4-QBmpL>Ej&j{({+`Vo!8h9L
zSIv~)PI1CKyNdcU!rWwg)YxiPz5bRWXg!3MTU?WxKP#pHk^8y9iE6jop5Bj(i79uz
zv(siX(OhTKC}*SYl`+eJeP;}R22EWgm>u^If{?Ukr)rvhCnS>2{;5u+uextSu)UlF
zFi720DV~{Rk>WoY4_<eR?9k9qazk(^va?7BiKPe>kqWaRy!)GN5ysGdsRI-8I(l?y
zsh|Fn`tIBI*2_N`H~C{Ue(I?Zqn?#-EV{<kki6}shvRMii{(VZvI?(!28T=E7>Rg>
zv24ovZkC*9XF3piUJ(%uB=NZm+#U+a82(z@Jd`1dH5RIr-ey(O&B=jnkXrN|GcIlJ
z0USC}6GK5!5i^a^V}HTL1n~mWqzE2-^IamM{&wsLx%qzO3f;o?x~?LAKRXQ(Js#F>
zRgmqrPea*4tNWAwGKb*^s0mJ~ckle*DOt~;lHzroyTgd^HnOr2TEpiJjd{}CCoi!0
z?k$414-i-GI-yf}3?@T5`$Lke{)O8oS4#&tR?p1c<1`#|TgAcgzwZHuNaAT^0Y?W1
zWa)d{wHa6%+`l_!bs+RMvIk!FteCs#!~)a#;#Ns_h#An;1(dAFaBmq?=2s#h2r$M8
zLJ61<>FMd8ndc?<Cs)83UqiFPpU>M7qO&+UIzB4}&x!%?8UhKWZ(N#h2n4^mV$IZk
z4Dtmkv6P%~Zta2Zx=44N8_y&IkJa0AG?~T$Okz3by!EJ6i^NsdJYYNm;=P|va9EgG
zRUA0Yx$M#jG~Q3t)H(oVHC-(^hqj!sQS%cO(s;uYO{DI)8mH&i7|R$JU$PpTzU6t<
zZ){HPaQg=y2W^;N%2oPzR1~&1B1UFGK@gAgrHq<dXlW@uebX@<&_WHeSaJYy|AB(y
zbW%a1T)P`p=;7?f#l;2LmVp1a+RK;*)XUATGlj6*TFty_oy_GIxmS&$hH0{rQ*3F{
z0;}I1U)AWF6=F=xb12Aa`0TB(>nMg$QcxhegY`q&Ju|)}At50EJXqh=A1TFYIzQ4a
zSsqu5^2+xdY~cWKC?#1Xw=!0uvy3#WB`3Uu_WJO<i(H<!=LT->g%Th^8&n>EfcqKj
zJHSR+UoG07fs>A_dwcWyr3pz%#*7$S&!py+!B3)i?@3Y3gCZUztU}M5QrF)<ogXsC
z_|srm7_b7pNO>>dBgdI0Qn%_%FDfn0k3CyM{R&0$d8T`)U)40^;;=A~aoG*(idnD!
z!MCg~JyHvU=3>H`_#17i5dBAuodR9tc*OlN3TVmCqA|1KOQ08l`NWT<Mrva#vm4bo
z=*$W2vHT@h@N*`I*k?lbo+&)i^?C(CPh)9m36I9|;OJ<1<@MswWGXD-iiA2UmMe~5
z=v0{WHt|NRBd;HfnyRWv{B60}{V=GTj^|^dMEkUC^H=_k>&0x2b|bINPJ;DsmQzh1
zrcFY?F{_a$ZI3Bxi(Z2=YJ{a5QS+_GyN<wr^8+(TJhR6}bj^NkGma3Yo(PQJuhBNE
zv)0ra3egJA17*dTlm#{5$yu&`L{0ev$X#PHt6ec+#?KG)<;xeF-HB(=LCd_I{lB~b
zkw9D3AB=A|NOPSs44nautx}mLESdZL7d<^a2S-N&rxd9S<Hhej238Y^JEqR$t1_-6
z=}EzJ)I3({oQ?j?B+&}TS_#;B_J{S*AC0!u{sd&wfEBdf8e9d;L@!_j0pV>mZ;D6u
zThGpeH5aP&R9beZ#@_IzAk=A!)84Z^ihPUonCt~1bwjM)amfP((6)QQD77o#vYVTE
z0BgPq2uHKaAB|4Usepv`1j=7jaxyqbo8kWSSl7^Cwek0rm9?aTf=z{G$Ix)7S3jIk
zsZIDWPhIg*@pA%KPe~$!k(|qbitVKIVD_qY0lL17L3AXQ2ZOKj>?j_e>2T~0&=xcG
z)|8xCDl4Gt(L3Gf3%Q$L#;>u3Kg7KJwRd46KB5b9wDHd{>rB>95RZ1=#X9oQa1>9s
z^P{z~wPby!Gv4TycE5K0PoF*+Rx9fdC-;E~ST1VU2X*p29o_KD#Hh$f^m!0p7}mF#
z$WuRm-lDQj$^1Y)34|Cf#UWQS^=TQSy@CgE?76yIWn=q&Kyv`|+8Ql;LcNrOvn^li
zxYm%xgh*pRA~l1OuQeUyY+!$jjn#i0zo#DsYb!SjgxqJx_0-kiR-6>&oH>E{d(&6x
zwwxr+SBHAw@V&`GT|`|Sd*}BM7~m3@NM?@!?j-9{$+sVeX;rl6zKW4&QwZ&nvoPcl
z`~3WK;f*0HUJn4PwSc18<y%#zDtiyaYsYzKmy-ciKSb$wcH7d>YCRVqUpy1y8(Uld
zfcj4AdM-=qykAaX@SCwW<7)cVO_wg`JiEp5(Nk`r&^%4S_}l;b#cm+Ce8w_!Kr=C|
zBIdM1B_m6-TO{NE{vEX){#H9juI}dNxJ6N^EFLdM1MI)2AAD1W>An0c5?6$9TT-3I
z+yP|`4T-=IK)3{4OY{2iwQnZGuJ#b0XjE!KpipQ}3k^7%-voTp&*SrJ=xt1kA1KsP
zn=C1x`uxdT4F$jtf>JkhO=Ea(L9LCqZg;&pszjq$GR8IF-~oi4b^R9DgNy+aBPbz(
z?s|8*nPR_0{^d&qU>7T`Hotp$dncJl0isEg<-<K(60T4Or|qF3HSG%xgw*EA)UEd_
zd)v`=2Z#~a7AQnRDgC6lbgJKgR18cSs?G=1;15B7v&)jWWfqo}T7cd6Q(q`6D-&^!
z1VrKW<K0mLpF6K;7;$@le-Hta{ym?w0V(dJqx;DS`dzLU&AU2JD!6y)Uf|2pKmQ9w
zqNLdP9ED_}ez<HbjpTf`QKY`<yFghT?TLUpRImJiesC=n6?mcIVp>=xqqp|gf*Ffv
z-3dSA=0?3*x*FCyv>3T_9l0VV-KmTZ@<Nr8+xo3$8MGf{5@c>is)n{QH|^_KV(r+m
z$2OC+m>E$DKgaf}U=mVMS-E*tq{CdL-wj;)K5h@U5lKnn;NUah{yE?90IL%41=l&A
zQ2@Fx(l%V0R>kk92;p5YUHP5MD;7zwiFg7+Mm^};Ep&UW+&vj$8mB3XDUylEuJB9m
zfIID&`^L$c6k7_Ef9XIP*x1m$vHu+t^KrRWL`3Apix+>74t!9sH&D%Y+ubX6mq)Dj
zKWwBR%<7esg`B<0o}Cx^ZHVy!rNSD)h}TI~aS}G<^5x_f%DNAF_WRx4RaK>PQ{(SI
zd$%drj-xYDhzG(*??zwLrt#V7Y2?xT)3?`JM29~7g_96QCZ^53?_y$OmN{U(BBJo^
z!Jpm$i;0Vi%l&XZ?t)+TTls80Kj{1GU#by>{mBE^PwEFNl;(ynghV?mu*PhlyY0?Q
zUe{4~2YY)*FP>1~7Uv4Kx<|Y@oOh^#Ngd)J0-ZP9kV6fe><JL<;i{D-kMqlkY5%Ux
zm&G}Z>Yb{QXkazJxSjTYDt$q?QmE^2CK?bBpg9mrTU*@ML{e2#^O}@&7??&;9F>2s
zm5j>3`-){Ux;76%+q+N~1iLL+I~#awc;>L7VTNXA6g)g?ZZC;>T@-schd_b`_UG5Q
zxIN&}xh`^Osj1h2JV~AAcKauwwWiWjbgR8chT+yIgNX4Lgxywz`l7DrAo}#`DsFrc
z91PjGIs)^?*_G{jjEtOi2+SeC*x3V?#q7^)6D^!tBpT8>`6f}ZjZQijD2x9aRbK&=
z<=VB2q(};=v`8wUNOwyk2uLU?-Q6V(N(s^_0t!lllr+-aDc#*Eob`VH|D7|(nZ2Fa
z<3@Pi`&su|SFIRIU+-;7^W%|BJ}+v~wJ&64sDmvn0Z`Kq<Lj>iABhKi`$a1QNiH2@
zu?Y#%X1QwORt5%N#pOYXa&)`=O};UxP@sS-D=EGA9en?ShH}oWesyWXUcmu}H@3=9
zJ!L`l1un^Q;*$F65aaqrUVd2qUNJ?p;VORXxI24w{R7+;ZS~4(hk2!3RYOC=BA**?
zkSldU9du%JBgH0WeGlwC8_dr@_Bl;60VLuI9Oih1O*;7jik^-BQ#p1q41NRe_?;d#
znRi1I4!;xK!-isC+t0OvFT>WsVQA_G-t4TbtoGOAFnD(|F$4t#0llD8(13xC-WfY{
z5}E;kF&#IzI;BH9)uS%rU95uCfM?pZ^wJIskNP#<Oig@{bucaJ7k3G|Kp(HeAaK4z
z;Kih1CL<Ht>!qoxN^7$T`W1lzEF2u8?u6&s1u0j9w;e4keh3W6$;wjm@@kqa8-jA^
zV3mcLSzao0Xuw2Q-`bbM<q5MZIWs^0x*Dez-?O~VK&#iwO+(aXMS{#f8(zzOwmZfz
ziNgImDIAaIXmX~AzqS&GKFUzn8{}_;LSSrcY)`Y#LgBW94?qPa5W?wIF72pXirT&l
zt2+1B>H6!1u6*eGH`H9$9OPaujRD;_`S$7Z4Us1h#scR{F_mH53|HkBX)Lc~m0ms0
z8(hT|b4X6fum4Cj*vx?SZsuM!O+UrJ__#EP1$XOjgx>zL9Iy}@YMC*=?R05W0#lWa
zJ55BnZUMm#3GT1GHGzcZB-p#tSgLr@sDHAf+>v*?#a?DVjAwF&DJHh#2()WU=m{$;
z#)y6%F_2+nW82!=_G=1padM7)tl1tHaF9dR*2M{~zt>X5T5F}pbXyDci(DgT#wJn9
zr0n?h)2$1nqz$En1OqF5+1IgZ&&cudsdSUSefuWxDvCA~c%gOGivq0y=eUW<$&Ixz
z5cDh?yk@*-F;+mgQbqiM@gu7$>A_z6AA!|V=H9pq9BYLQ<6bS{=CqF<6VCel-+uJ=
z59aMYw6Za7Go2a;+Yu(UdArZ~vrp;s9iwlk7i#wefja8$?tc92+1R6DuzJvfx&(|i
zufRP5T5-paZ0L~2m>5mXtp1EnhpnQt9=uH-a0wKskR`xGTXp-VL|Sz(_VIfR<38$(
zR2MrJm-tiARnc&AzOMVV*cRE<qLZu2=DekGUW<u~JDR&TIx<phGb_Yvem%+goP?c?
z?VZ?rewyIacfSj%n&HDil^P2C#d{idmRS5;o}3}CN9SvG9R(Swew)sCiFmtschB6P
z_{{U6^t6$I&vCJ_>(Xq!sY8>m-@JLl26*TB-^1hMZ7zXq!Rsm1H_^2A4i3u}{H}W$
zy(R-49WueJcocj`ae)uQjc-OOu0J@IzR?&^NW&4Io|JY-lP0p<nEPkVFA~p;cZ!6u
zN=aw_uMmIj{%gXT$eUywj2p3$@Gj!VIIk-5z^!Ho#ZkCV=I)O_(4Qwf#_9NYh1$kH
zqWl-v9rLIV?<O6SLnL%EhNLjHxv9wy3R(IK8Z~&go~a0(xcQK>e6aSwc48;W#$?k@
z!ju*pd`>G8WoDwgHooe_k6JBwaBxt$TX)Ihu!4<(j@<xN6!aXL7I{jKfz%u@o345H
zv&HMuT%#22*F*eR11S$zOG^1<+Bg4psUt2GnyU%|x2Gy%qM{l=>SH7$n=a&q(65S$
ziXu3^6dpUVHmfC&?Xj)P%gQbQ9(_;=FCa+1JI1)cGyC@$$3{zQ!9DvEtihBz&CTk0
zlBan+i>sT#Ib~8aPUHJ{<L5#ge~H}JLLQLd?_)d_)_bX=LkTi9xTT?Eq*o*YMG`$j
zQ(!>A8V7-FG`N9GoI(^XBQ9^Dptom%h=~ZC0C}^yhTA7jSO1pywSK!{(q!L_tQ78c
z@b-C^@nWnbCvNoqiS~5gDz9#o{W-q5OgJ){B)TNW4=1L!`;zxvk560Lx`OTS!rI^g
z!`lnL+_eTQtgTnZiw$7p4F&~rRC4kU6(hCBL_GZb4t;{B2wZl4b6#})P;p=9LPm%;
zQNiou9S<xAofOuHwl>#yqvB#?lmtcja{@M5lr)=3XSGjlT@2i!g#Y%{$`2{2&m#?<
zTnamHCG?u)c$He%+7{`=41s=o;+GW}{P<h5^}0HDOF(Hg`co5-B{o3)taWC;tfH{U
z*q)@=4oE8uq5DJ-!_#cM)l^cz#KG0r>*dEYTZw6$!)n4O(fwX3uH1h6O;xXif`RYt
z1$K}>G3!pjPvk+AG)QtigzEm|G#$pDFkJnevqZDPo`*u7XIA;3zfkMMB`&m76pM8u
z%H!uaW+a;1%A{#*7+GZCB)*=+ER&t-{!94dOlVz!(Oj05jYB`@hHV91xGmUY<<loO
zKg8$*_`~i_8gcOk{{8!s_V#SLg69t?Eyx|l`Ual-&#|7xEDLSc7}c&$kJt511eT;o
zzd5LeuZ~s*g11tkT|NMA&3<5;f;V*U<*T>2Xc!rK!qcWqJV{gU-|~bdP$%Kf_hNFa
zUDEhK+kCV;tLq(-^X=QF>X3kEYZUvSrYk4^cvDbn(P4%uI{6xs+)DFsMvY_^|Ah!Q
z2Z!xk!!5VOnx8-A{~DW^w3YS=bwM-||40_16h!$)jBqf_>h<ykhlW@?SnwR-(E6ao
zT2;TCdc^9686ZGU&RH$)#kqCGu;r>K&Bbh`*Nerv{e-o=WdQqW--zH6aA*e-`Y^YZ
zpjX<=)^(@w#(-Q1#@8=n0yHuNfY9&-R2y)zj6VvY`XNHkbK$Udp#N9!oHDySo0utF
zb7B8+--SRk(#IWfp-wW!aiu*Ji{64o$5`7BMQp#`eJ<yRb4#T9KuP)S<a+VS%8IM2
zD?T(<3x7WDpPY=+JcgN=t#}%G0eNN?78VjNOJc^@4?szz-YGv_8wem~!X(-1e*BiK
zuDg4FD9n=dwK+Kpi>qA|PiCv@yFvfUSIExy@9#^RlQK+o$E}9jE(c;Rez&ropZi0g
zoHRn-Bb0qf`n%RhnS_8~aQ!kSHI-$y6ue0XT<YrTJGG}e2=DEkJA=7%zwnfEt8ry=
z{dg=aw=a$Z`>9@8IoG|gkWyja5<*I@z<hwKLYN>((d74nQNP#haov`vj!v3qSZ?kU
zUteFg;w{UOm6!Xagr=B}Yi}j0lekugj&?Kg{on#W2K*atnytI~akQZyA3WR0Ix#Xd
zTr;h|#maRz8Y>X_16F`Xkddy&+1tv~SFAPs1O4@C$oa|@MK??8p7BTM%Oek-bFdn>
zp3Ro%9dsS-(okBWG%4OCX1foU8C3k;>XUOY?k(hgOM3@;M@!v<rym4gxHw%s9gxfL
z@miznH?Y%k{TxcjV1IRH$7MO*0Ftkr$unR;>FUpWY;I#=835G!Q=0TV2D1zsvtTJS
z@911S!<^mKmKcp@*Owvwcazv`#g3<|%);!=H8)6YI205V5EMOPP9-KIb2wa4NNue>
zaew()R$q9POSRI1T$qk8H3g#~`ch9eV|<k%;|JBfSXs~}Fu4NL|B5jdVkkD&0t5vF
z40!|U>wUhxfxt5Y3Nf#FF@~os=s|N%c-_~f-}z=cSMukBqRcFP)T2xfy;(1L6ZSb0
z%s8`C@0zCSlah|t`_#)<bKv=)prC+0c*(-S&TfDo989qQsjsU}P1bnwyyYWY<GGbB
z*5MmXTQH*A+|m&6IFjzFW7TY{rzSEnE~@=;-%`w$;)G6F4Ek&Tzd}%4Kqj84ocoxG
zX~?c0>6QNJ@$nz<h9o7Cg3soKS{zr;z{tq!y4H*gs;a?>{{DtfG*U|zz-0}o`S?D^
zbq;fIoRhT@^ms}%^iyro(5Rz}u<dYs?G>?${X^-7zgT}-C}bC16N@yl>?uROY5Vl5
z=`d{DlY}!`-J2d7QjuC2p75`&s!H_7!0Y^hm6W3tC?*OTU4S|zBqY|_t@su^ug!)2
zjTvky-p3`6rd`Qi^g=T`q43&tGk0L$k_k|llCb?yT+CBo^YB%N^}}^5OUrJ=aA$#a
z`*x*O;cD7K0i4!KWhUx+_5!id*WdfO6sN|=*M4@E#Y;P(!AAv}2gi1MV`KWU!p)6X
z2+D9GXN@c8QCj7Zzg!s}ZD@U6WOk2ZSX9@d@}>g+{C2{1*~+LcnXg(!o^<zF&~a#?
z!|??gHcnWj$<`t>w%5I>y6Pv9DmS-%{z)+ddM0qS*X#*W1mW$a?SAia9&YY<QK`)(
z+NN}h|IB7O$Si)U_V)IBhlkC;P=itHW%7#k17LljjASV1@&IiD>0r#lbIzx?9r?>R
z8Uj^Ca|+AueU$GGbG4mEXiR0LrQ0dEWsD6CZGjS0fzUTltDi&FKgzU-)MYX9yt<s^
zXmt3q(1E+)JJ3duXA_c+V;}JS@{yxoyGlKZqhv-`XQxJ)>0JywO2iD{310>#WhE`G
zZot)Kq@}5zJdruU{8d_N(3`>wH=sS-x@+ThGSMv{=m2~AYO>APkk%|cl=J)gk4Y0w
z342K?zffm-?3rq_&1hiN2nTHMide-`b(LOYv)`bmIxfwA5UG%|uwY_%nG)5ZacK-T
z1jWQKG*nc_jovS6KX?zs7-ghlq@M`3%$LxVI8nS0&*j9)lO$EeNgz(z3us#u*glln
z>Ml#i!3;g*e6YmPBp$EgM1UFoS+YcJvb`as>TtCLLMUOd40@Bl0;w=CxY<H&0P7d8
z$H`Bf!;DX#?n2W~2N1;Gp6%7u70h9qqfLbcc^|T7rON5><LooH@?Ii;$*&@bJ;wP?
zhiSqjr-O5}`f1=Rt1{!K_b4yoLlk&Bi`kxw7IU&M8auEnY#Y+hT)rh_YY*-4Pfa{n
zV8XtAXZ@%gp_u{uAQ+%Q=O6!bpNpFt!RaD;35!wYt?ioh3`1k%67vzpx^w!p4fA~=
z91>I&F)CX8i=+3?%D?N`P+Q95^GeFnX=vV=ki3lik^IqiZO<|I!PiYz_RDo^I>AQm
zNB!!XsQRN4KDrjGftLEH1_AX)S^fCYJC`06l<iJVPKG8Xo2s!85<!=gGlg)8xZfQE
z7GUr^dCKfFcZg(TxNg%*VMa0K#c?o3Tw}@~y+L-H?TE8-d{d_%AN(~pHGHTdO7g_8
zKL^QE)O-&7(p9|6oi${TbA9(SeGftm2VyS7Tnr?ochkMw($Os-&4DTSA&|f%_5ZeQ
zoHBp=I3E$L={|JFEaM8@1M$Aea%Zf-QNQ?4&BLLb;b>Yg;WQ#7#t3rzV*T4KZ!Y8+
zO3v)4`iD>H-oA2D5cmn8xCR6tMGBrIP@ZoOBye&{ym(P;Hbj4My4#0fMxgUu-gXyd
z<Tih_iEa4n?ncwgp92@izHUQ$b^!&eYIyyhM1YVe0f$3@PHiXH5j1Muoxa^`Amg?!
z0%twqW2*D2(c)x#{#<`!I5!|504S24)^M_pzCMhGym}OF1|i0>9+a6E)6P27OTYd&
z=#FvGy_?X*-}od-ZNzHCIsFq~ArKCFX&Tzc0xwycU=bGybwV_7DY#{_I&G+?hKg+q
zn=37U{HL1FSi)Gcyca`bB$cZ^vYFDXWF{SDKqZQT^5nZv+n27~&m(5bY@_s}sCd}a
zu6p<}=;-s-i@nQ}T|fJ7(wzRx+FvJe<y`^%*9P)tu!?oucckJdhXDp;#fZrm?(mn3
z47U%~x3}|E=E(%z%YLy<-e1md>{&6!jW?yKWiN;unw<aCKh_dZ9C4CTqto*zDcLdL
zhm$VvXQKlu26PQ`pGb<B+rT3)oh|^oS~6`m*Whc9CdkaZ6&E`Vs+Z9MZ8>m9CMG5>
zIGvPJ*?Ll_i0O?-Q67J`E4@Ez6O=`iFZ2c_n;@v>eSD2Ho1+V%GA_;}M()*gB@MFn
zt69mosl0CW$rmL_WWE)(Tz~pU-lUbeSW+{?3>%}jw7NQ!#`ic@Rx+G4TRg5UjAV$0
zL^K}tjeip3_@2#p3272(L+eZYb<K|rODBxi^^ORrjQ)Jsee_|-*D~sd*@N$&)J5)#
zj9c(%xIFV8qP?3Y;6?y8Wf0T1_w>{*O1#0TJ6pg42{I-^Q3mtXpQUf{A}F<})pAD!
zrYyeJW@h?VXZwamMxVe`D<RQB5m{s0LrUqkbQi!%>6b6#evkt~m*1E&&-DCBT6^B>
z9|#j(PfyQdqB}V+Q)N|EiD2}v&Ck}5iFQRC?)kt6fF?2P)}??=9UBM7CRIYiK?OC-
zaH*-7`LR{JK#o!atI6%;#~i~u1r=qDL-=c>qiTQMH(4^j2993#aL+CQ0k!q2Bu+M)
znv*jYB-02X&AREMdw0-wc6O-qSuIHlmmEd+=r9Y2Zd0emqgG**oj4^~IFMuNNBZxQ
z@84Jx_bEQ=-+rZ{LZe9Jyfywvkw{)aA(79C5ne-tph;+x#SnSsE=xX(VVp_Hc9XNS
z_H<X(hJ(ABv!pD|?+Xt$hc$U)+{B3Bp6^dpu)_W1rCFz25q>{VJ3A#IT>?6a@Oi-0
zdm9gLtmavvc1;eUR7zrEu9P(bT5)lCkmY`4DxgUDXR7gj^2?q5q#R@Ar?DmbmbWlI
z$W~51D#mFT91o%T=G*P^uy^>2zg@}3<f+DayA6+;gNhE0>1@Kr;v!9qv9NGGr3@<2
zq2h250S5Y#osAYK)(90S<im$|TM3dSk%FV!#0ovO^t!IX4_`iTcHVw&$a?0E*(6{d
zU`o1mr=AB{dnALdJ5l{h*^FV^FuvrJg$nkJWkW`@g-wi8SnXhN;)wsUN8TcBPj`11
z@Fr`jv9-T{yQfr+->snSCskE%I!zaC;ufQsnY~4MCD`qbALg!RdX`OKDr3H^)Qdj4
z9r-2e3j0oAG;4aW8>=N+#J}KE)3B3Qf4kzJt%7iKVZZI+>c?kwmCa1}C4z7z4x4jA
zAnm*HA})eixeQ+&<QY0BopB=mQA|*9uNivG3$?2Jf`c(|aB$#oK7ZG8+il7xTBO2;
z8r+&o;I^TIzzu*Dk^Rb;-;{-9%f+)<PghdP2}F?dHo=?)#M2|dhafjBE~)N*MZAP-
zE&54TL*(V$!Rr~T9Qrv<WKY4yJ`~BywTXF()8y8uRy9iRGkwUB?s<O;MWfzZFoxlD
zdP}F2whp`kRaMpv?{6g}CcXk#B&O6m!atXi;+FTP%xX)+OkoW<nynjbg1<o<Okx}Q
z+}PG}SnEOP|H#Yj@gkLMi3`NP@80*d#_N%I1bH_w=__eyJhma@Z;vEF$zgYB#}>GV
z6!Y~fsb<rR{-o?K&7kpTxSt4194jw5@0FZ+>S5*E8upGS2{I`Dx9zc8T8GigvsPNl
zW44$l|MXha@Oo%&{>`u(?_YZl7^enMmLT(&PvXdxI?)3)H(v`M1A~;0TW#b$1M;oC
zUlo5Qhd*!Pgd4`IN7)yD-oU*iWsLpKr@=!?MvTvDixh)_RTXs8a4UhBSZQjysV?)Q
zrLc={YPR}!)8s5-5)z1}d1+`!3#8_{ou`Ypx6uCnK6ruO&^v%%$z**nTi{~tD-S=G
zN8^b?04WDI*0z1G8BZ(*+GLrB$n-n$lo$SU0d}_h=g)7PkhNzXmvxa2dcN9>j|jcx
zI_Yoo_UJU^T5sGM`I~v*0x6F=+G>HtP5xLa$0WdtGOX5ggw@(d2_hCsPB?X!*uU}v
z<Bat!D$ZmQij@MsIr{68`o?9c;jUW0o4F=_s?+u2ixTUE$8)@(!j=waiBk+05f9{<
z@$nGwzkO3oHwO1ScmuY7>DJTAAPNHozw;hMcpm`tV#a+!V44vx;u5NZoG+JgOUGlI
z6QS47%*;Fk8@`Q&fJ`(3_8R}~#_j|3s4G-H$6q%Uw;es;#3GE%pkn;~lPC`1MP}Bn
z{tRxAv$!|%?>~IFgWv@r3nCO8#y&Uh_ac*KDwpk2Gxw-HPx7T7qJA<SAKbZRVtnWC
zPA1>><}I3_Vp*%^PC7*;Wo0o}SDww${LN2j`X+J<;8$nXVJ=%px})#rR`v7k%r6J-
z9vrKCvMNl(>?p00etq(+#Q`nP^t#@Pb?O`5Ila0?trJj(+q6?atFHQu^YuoxbB`p~
zODjV@Y=LH5a)GS^!nG5?b0LttJ$4Xg-*!SYEE+8Ad+5U)7PC?g;tuqD==cJxh{H%?
zuo?bkFlsXHjZp9PXS`CO%wO5}Q37ufM86EIPL~lTeI3q`7iio7U2D-3{SvOX;>?GK
zhv&dxLt|1ty|)g=$!Iz`>$mp9gM-whzXykSuRDUtpRmwW{EiO7?fF1e)=x?j3=`Hh
zC8gLIzR8`!F?@%$+rlZl!L%~OWW@c`5#gjyVl@(pRffiCH}kA6PK;hATeKoF3`#ag
z+^Xyr(Lr)52qzT5@chQRf1g>aG75?X0Xg|k?oW@OT*Nhpe(bQJ>E`{0<|d)|qy4{H
z0BfJnpr8ehY~{7D4|;Z|*{d1OA_>?0vYZ3X8HvoSaG$dwDZb68ZLY09Be|{Yd0id%
z_<`1gHfP#=6lVQOPMY6>t}ZJo;2@k^S~@C4H*2K8H^pHrXmCkyI<XNQK|Vpz-&bL^
z^3{3&0ll6GnMH;`O7W(>`rmUR?G8yevt~yq0>pLEqRtgBeYdf<F*zCrzDfb!(FDy<
zN=gc)&k4sFh{9Y!Jdp^dbnu5&|A?<eB?ZJlJ2PHnlW;93haQ@Lu!k9d;Sd0?+O|cj
z@giSPAS~?e_Q8O2%t&01ikvU%;J`lh%gO*e6<gC)F>o2ft(_?7!39+WP^Hqh!efXF
z8l>{zM}Ac2bLFk!xl6%~uB?Q*FC5Y$io}4S@9UC_U0_<P`J78^fO@>i#`};=2&#a!
zu8E7>f?WL5Ww0}Et92eYtI-txlnaQ#q~=um>ZbO%p+RFjotU(FcvgH#uG-9>;!%0i
zAIp623wgF-i&R$Zt)RVj%lutS#UJq%{s;E?-*(e*JRbjiJUrDl-jEQAg6yi-g3?XQ
zhGP-a>omokWV5H)TYt`as5@<$`$tJ%n#o+ys~8zg$G<kg`G#YH|F=PAD$r)2Z~}Aw
zA;=s$qrZCWR0YF{_^WK6wdh~#BOJ=_qV%fJ33=VZ+%G&A^WHJewB=}gegY{e>tyeJ
z8xMX@@~xuhoXc%j4p+80NSLf_N`slB3?ex}=&0cj`#y(A!{1Xg7X=RuW-`PS2`_%s
zzb*!aBnjx2u<kSWf9$YMC$F|xb&tqi-RORf!Rs5&sm1KOIpW6?jMnVuQS*{)>xs8D
za=r|%0zlqBQpvveKD)<;g+H?N>X#q$B`*FAHy1uz#L2y7hu5~}IVU{$gW-?<=~Z%K
zSKPg+e&e4VB|0th`Bbryg3uJFvukcDQh&_loim3XzQB=G0-s`^^K!jH2AvG`zb&rO
zoye6dBmK;WOXcrh%nxAO?QIx-PWwvZi|%@TzUl@_r1R#7*9aB-@M1C82=<PSvc7!r
zg_-Cd>VOV14$M6Wu<95%1R$#bQRCs|1+VsynWvT1=lJ+V@YgFS;4mnrwE#9^ldenP
zYaI#s2jDs*Q_Fut7fcHouZJLCFLGQTd|CLyA58H(K6|2!h1&i|4_&uq=Bw?#@pw?G
zS-t;ib<eEuZsx?EF~v{0_ZHQns8IgwRc%40YAK$jNpJ)HZX?|_gc-xIg8*vm0Hr#Y
zQ}pfEJw@j@<Nthd_t7N$=f|b@J|b0wW^6Sc*!&><z-}a#&y(yooH629U0Jz?`VpJl
z6)?4zN}8G-Kq4D`tT+kvq7R}Et9aO!<=cpgz4xsyJFPbUq_ouaf%?WI^8pdMJTZPl
z_ZRkr=Z1U*UfNMn?YZ6e2E!J#p}Aa785-{YjMltsFqx5?!rV94zpTo}$o8t@ggLa3
zzXRF{BA9K^GKz|n+E%M;Z(MxX*)^T#_4IqWdB06~zZAf3<Hw!T^wGUYN{VtSPua3q
zHq^<ds^6%lJm0+yWI);xS@T?-m$bEotblHQ?E?-O&m;?Ami{y}ya&Yc0A^9bhYz2<
zrQze{Er!Pu$E?}3H~$9byKMFXm>mzCDEJ(u=_ip;FhV3m3tv$?du%>^%tGi;-63jw
zQ#<Dx&-P`9%eT&MF^JqO4OnRKk^jDv$*M$MbvkCAAb25e^c6~<s_49vXvS3Yuu|0G
zJlC8tP5~EdSs~Sw_MEy@2504)JG0)XHQZ(ATX;diX+e7)M%4disr#d8?WlbiHczBf
zlo-=~5DKmoyUJpVW4O-79y&>8Y;JCX+yfh-7y_vY#IC%J0AL$pFjm32x&#Uo@XPU4
zJ#fe5=$Ff=k$C32lOlrhd#?%W_hBzTLiwU7biNd4zyf{?L>mYh8va0jxK#H&ixsh;
zf`S`KiD)7r9;z$u?e+CF;*JKbZjp+%tN8%0-|7sL7WdYr_!j%zd``uz)uUbV7Mu45
zRwtINlKs=jIy7es4|gBk(~W!Bv>h6golQp%v$w#qn%qJ~6Tit^;3DJQjFNSy`6him
zT2d78xQ0N&Ot#ec@8y-HC1|q~eoB#P=sAMPN<5@L#<x7h>Ix!S%vwV3^3|22psxj|
zKoAosyO<+C0q{*oL_<SEp{QGE@Re0yHW8ylLd}hf|D#KBsp*4nj1Egy|Ku++qlbsN
zuqsfp#X{?wU5#;5MPb^8KeNA{+m~yR`?Wq7r|(Rl$420*5<0Pc(-ZMv9<$%g&|Z@W
zd;j*-2DUAs;>D^HJ@3X?;lpT%^8^{<TQKV|dJBt+V&CjA*>^Iz(Qt+b_#GmJQRww`
z&SA<skYtKuGLeyvJJltyoCj{PR7;Cb-3b<*s?FS4+8=0ZNmy?VD$@H*hVp<pv*-nj
zqUwhaAC~?Bi3|FkjFJ)pnP^tPB6hU&SGfT_fWb-W$B3b!sSh`_@57_!GqjKfT|Q4)
zFhZ2>75%kM{~vv&Dr3lDR%nrJ0o9lbp>V7STuz;7`*8muh(c>BPF7aE=}J5BmckX^
zeK|%r5$`1-m(aJl8M|)6n1nK=9+&#l-B@Z6xBd-`wXt$VC%Q#0F$p33c>clZLH)d*
z3+sCi>H34~Keh8z1Up@R<@I=igXD91`@Ujmqf^GLwxF&FDk&mz7(%PSQ_$YqJ8DO^
zflazYh>tG<iltTe)Bjf$Yv}vo5knSske?qiXstywG&JHDEhkC>L1?nDxhV%^aD23%
zE*})diwk$)Go@k73k?l@Z)=wTvM0DZ=>!C{Mn9%nRfb*;poYxep0&Im8#OD8ie)@~
z;;(=aO?x!3Sz#7#?j>+`s(aIFlhcFVGI#^QIA>rsav~t$EGjJ>%3XlWk<2#QH`&Vj
z?KPVkH#+26EJU{$W8#~pm#KY=1W9egPAEll*nf#zz0IEPC;pui(pjd9I?m(bRu`cV
z01VH;px3*1)SzXuCwSQK{nb~nD>gyE3ZUu{l!AFG1MD0eEwi(_^f?Gp<vE%*Sxv9f
zh{m74A1}vMD-*K@4Qmsu<!Iz`@Df|K;-#BX#EV%8@bWub8LwlHpQo>0CA+^CJTZ2@
za=T7<yO`~3e~J!?PkgcMz$H^;KE%g=4=p?JzFKvj{{hx<f=K~_0Iu2<Bo&&!EJ^6;
zGaK1gmt-;@2sh-EE|jh;>a~5yKr+U8NUzd(mA8qQVysr5+Gdr>7YOicYO<M}v4Yg<
zkzo-b`W-a1u^%13-@F_hu~(LR8sn6ovVv<L0TQnH#YO+f$jB|4w8X?GQ<%SIZ9olq
zV80KFZ#Z+GtgeuPa4GT#ALVPZe4-6~ChQZz{ns;cL2lveJtC@*!+d3oU7`v(L@k3`
za72b>Bw!K%A5SkYOl<5H@W;zU^LSlw0Kf!_!40%c)_z%hLvAvwz^a?nDm}-R!ZZ;J
z+bZHw@7tH&L<Gg61p?Qd7M|BNQXWf^M=DUngbRcl4&v4AFEac0BFT6^{+WMX$s2j2
z**kI17yWGU;Z1w}_3^Xoqi@3de?7DvLJkh>K|KV?Ea)U8B!I;`))_qWMR&(^j7M~i
zgr@His&@i{CZI~IQ$}>^y(#eU@W7rd4RJcK%BO4JTLBpVr&^U%&x9?cyCIekvFt^o
z#1NAu!uidEh%$NCMVfguiUqu9H#;Yj*PN!N0WNj#y>b`MUfiV~VO6311JOe45Qze4
z=(keZ>v4m~yYQxh=xI4GKF1z&Zf)(m_)B>2j}Z!|6|kK^<eJ)<C=-r^>gn$>Yl<<P
zOlnnIiKul#Or;#BbiIUhP&t2m|D~>CxoKc{SONrhpgCEF?@X)OQKqd^;fs=<jg2A@
zFmN3jgEzt5eez^2OeFoslh4^CX6g|g;l2^~Bf81i{8?8!IC7K6J<NnXKAhUu=Jv=<
zHNc?5@cf6u#=Pmgw3|5EL*suReOA3?zUJMPl9EECl_0=1K|u$I=Y&|C7082P07==i
ziJpcABnS)xj0}OqLLk6AWIe^PECMu#h+6>mbn9;|nM^dXl}Xo!IV9?7Ev1TI(az_0
zrh2vYrm7B!RLzoyuF%KXJ<eE-Yj6$1{AXC!Gf}sAb-T%O&9W-8iAc%FHUk?>|KdR&
zA(T$=OJ!A7YM2n9X39b+x+(CZV6rS8m6UT3l6eOmK7`Z8e0p>wLD<%ea;RrJc352F
z_xA2uYpY}}k#tX%=ocl(@bXPgCIcnk^lvw~eaL&$g$vn(fKW(FOPeM(Zaw#9dDCS1
zw|1&UvGE4mcQj)O`R6$j0ij4VbcG<T4}y>JhFRhQ9zHBJiUyG)($#cMsjDnA4F=lu
zu|v6(3u^VRUvV(Yr~nHCiuT*s*jtYEaP3CHC>?=tPOIyBm2buz|KHhrMi4qtSoL!T
zUeo{lzAAi^Tj2$kGlV?kr$(_~Wo4yx+eJBnDbAa>Z)N1=Y2j86C*$svXb?6eX!EPp
zvwvUorA7^NU{_qn*AzAEGA86?=gHI04V|ZR3r2QY)1|EaI>gfj=XZawU{p^0@+?$k
z#Kgl}1p{?_d<%cw3gi-`-rhV}>YkX;Wboq^6ifv*`O9=67DnT{cn;K#fki1bDxXsW
z3E#e;S(s3L^#u7<_{m3nWxPmM^2>y^yK7L|aUTjX*I!56tA+M-sNTm1&?->~F~oI)
z9w$sd^EM{tK+Np$_1B=B((o)@!KX1Re&*Z%cw0x46U9wf0VI-~a{sa>k_EuJon`Xy
zqE`5?M+0rv*zCc#*|znp`=plB4|cPjnOa)5fB**`Vk`(R+Q;HPf2Iu{druWhBN^5J
z=pK`L!CD73gx6r=;D8z^4O9@FIGIV?AEhb}4qEK8;<b&Ys<5^gnKSdYy*as9sVqa-
z9fE3DyPQlbDEnr<WS7TAMwskJs(h;&<&sf(OoUCsh6<g6tejkx{qh}(hr;OS=tq#)
z;^5!_KjIe^uU{1vW1rR`REmI*uz7!RVRtxeeKC<Tx;JX7PrRDOk+vWB=&bDQj@=31
zVI6;>VBnAjq^5!!JRjnTlwQ61!kz}HAC?ckosTs)qZk-0);wIE!MR?&U55Uk@A=bt
zhnJQ!A+$2Ckg5TH12_tHKMYBObb4T5;2_UXx}RvpwUy1h-6!bENh;$OI41ib6pMt{
zo)#(~1W&<RYkYlm4*qI8&>mZp^?P8Ji9j=n3`FKAYIQ+D1md(p4iJ1vQPIg{PiROX
zSpf^oF)veieD_%hj=i^`CPJzdf}9M2RBIN6wZE*~{(G3L>_n2-{(bd<IQ)T}GwhL#
zDSL0__yC8j(G@cwoaVom=84*?>z681w97h-sZazjA7iX+fe^?#lHrI|HU4D!{vTnn
zkyuhn%1?1UjszY%YMa{Q`)wEj=E=8r(ao7283a=eY1T+gbHC4}%c&bWXEjHM*t!$}
zH|4@ijP@)Z&^<;H&3>@FytqI}TY*S6H8s^+y_}keEe+0-oPTZglY_oRM)lJ-xSB;i
zP7R?EsiqsK?_1Yyl*eQfmNIlV*)(c+U+qhs?9BXND1Xk$N%#1%bloq|iEau^7id*c
zcOmmdTj2opCtmOU?<k8B>0gh0;xb<tuXS?*n;~mltBcM|cm8+7<J_U2U4r6n$wR?l
z)d9W85B0*0Z5smve{lB8CjdI+RJM%-0`os`RH3tgUq`LPFa&^k|16?l`Rq&Z&RO~F
z(-nbn3HJ-T^E5t)aPHlNME+I~&Vr?M)6Et#-8RmdAUWx0*9X&QzIFU(hM!p)NB)3N
z_MjP07-<f=E`;50rMrrT+6(*=h|CWV5F*{i!J*^gy5INZkqh~<7egiD0-f3QZiB3u
zUb$zrEdV;R-2BeVy94+&+=z%MA24LBduRN=cF#s3;}&#z4WFN=5I=v13fDS7#t=?{
z14|LBJkQx+<S@EAduvg59S<;}rEp?zN_|`74$<9cWL8(I?i6e<mie!oN@GYve)D~i
zE|gxQ!`Kf!(WVGq+`GiX6E!HM>U|*FBIK2?_l9-$`urYc3M-~e^mwTW9=z8}AiP?r
zrlAQ958qak{pw3AV+qR|)OFkfq4yM73!1Wceu?LtJ;*6f=Kap=!Jix!1zYoHwE2F&
zQ7(EZLqeONO*L$qjv3d(ndG+Jo+y=)GWqG7`{fBCIkIcJ*Io7?`L@9|2A=O_;zkDd
zgzsC=*ZA#rF!yC<zgB0=RAu}`3*QTfN~s_s6;j%En8d7je69r$t|oB$M{PeiJ}M2N
z7g0c&n-fwKSm691B^mk7boOn{hNFa&3gPVPaw*>`N86MknQ>4Mc3DXldLVfgvRHIR
z;?ddV<>6`lO)9k9ZV(8XlaV39fC{O+;)QRpGBanc1s%xy)9)6<hEdW*qi*l^&sM>E
zqok2_oY4D>W9-o-jAk6%+?z7NAVhF%FNus)jQl+@^)+5(mZHpM9$I+#13|zvp|FCW
z5CY60*~>UG<Os;^kcbFlz|Y=C_5S(ur$bQGg<$!gN<Sg8wu*{>e0==h((~drN&|>v
z`%m;cp~HV}E~M%{w4E2vtbKui_dcze-JdW?zYDbqvbfZBT<6hbWo2C<Y|=v_?I!-Y
z=jR$zQqKj^yCbzUn7HXgIPYcATqFH{EElcfr_@r>QixB5a?2Hn90|{BJ?A;X_;zdW
zdThb0Kowl;-FE^p>+}pmlH-d;pcW%!ZIHK$;N&3#XE`bgu46<^gn-av9UGIRLD0bA
zQ&99kmNKv1LV(t`yu1+VgL-`5CrA-AW%CT6ZD3t?&an1vI{Kv-E72~D-81<iEj5*O
zhvnIWzhjW%0xfq~c=-Bky)bn7GZ)wu^{o&E0|P)OaEr@x#Q67K!KhlGT`T!D?^D*G
zKRKh^pT6D-76sM>Bn%&U91qH4T%{qc=1(PG?LX+hG<vF>%v<O?h~$Ccv9Ym%XiK5&
zLxYTtP24c@LmkoABd~Ri;&*5W^W(QM4QLx6hoBKIxIo$LOvQ8|Vq)U<X!303P>}z^
zO`=urtxewtk^(wz3%A3)=YIt}!=f5r3UE#OSJ7w{uLTwTpbO1fAW%#1G;r|#LjL{T
zJ5!MuSp|RD7xTyFKQE8;Au^%lsHv;#Bcc_ipiqE@5{M>aV9Z<o{CRG-Ai?WA4aRin
zS@7S(CIcM^qj2Md0Nz{T;^MqolKv%3v7fbl{c{&QHtSd?<s=+DB9PxEPq}xQS^@a_
zhym}Vq4+iLC(G}F*4bLov;fjN=5+$EJ!8J@u|Mk%0e;C`Rtw-j0Ecmr)21419#jC&
z8l)y8VvJO%y%t(9;p3gC(?0rzuq}a!3CP`p(%{}JRk(TMAt62mva#;oyY~kw^hCXn
zFjU#ruC5;*GY-#=`rjh1wtH)7uQNo?;@#ccUcGtK2|r7NilX31c5fcjRiRo&AJ_D7
z<UyS`$_XXGy`{tWmOc(O$JW-p1wvazR&oJ!8!u*BvZQS!@e~O(j|<wOfsq|b;LtSW
zymDYBK`8q9_}<WmozD6Q!!7LU?oKU30Etet@Lxy&<0c1F($o0EYxlgwc<yYAX2k=%
zk89RADFO|p;YyK7z`sw5=M$*8e2ZxzHo+5BNp-$N;^qzQZ+!NdNiG0S6XH0ap``&f
z4I(|dx)vc)tvgNd?aNw=-B*7>8vuhU2wW{^s$bLV%4SEy#!X%{CBkRA2gc7ko^+5`
zr^t31iMn#Ks*$pfHb%-=JqcK(Xw`q0L1v&crTl}gJ09<oww-L86)_{ue)*x%U;`CB
zFB+kM4Z*KhC(_Xp#>P1U0|RFQ6^L65l9K4*Y_dY%0|Q4W<jWKRgFb9t{_^!GIF>+(
zh1gR8<Pv;<zB%;0LWlggI?RvO-(g~Wl&hNVz5f!05Cz4!8h1HyzfRuP{&Q^I>JN*v
zHHrJPD`~M#L)!yw;|X)3D2W2&iJ-pO_qe_n_M7jtN&#YcL1aP>Nn-Dzhr}l%`&eHu
zOhp>2UZC~FLiU?{qL`)SPv(0+Aitc6Nv=TcTQ&ui@C`w}HW7P1bHfPQ)jujdW(^nH
z)X%g$y4b-(R7~pQbLEcuE$@u)+#(cn%Io|p8NM;V*qEA8f$ZhdpNi)Q;Rhi8f}~cI
z_YDmq5QGTqx6~J<D42~OP>cM`fIAjc#78ef51$ny=lkXeL!zF}yPhOYtt$IKh)P6o
zv}rhS1phm63NrK~A?#3u7pjb}D^lxH=Wp)$j*3K6-S~Z%-IWWRbiT)e<)AslTnFux
zNCGE`B+jn#vl;j^8K4pbDiu-Az@(z|m<S*rWlc?3*RvNJ5fVZL&RpU89WPlYKUI<T
zr+0?R0;=aiMsG;;qT<ikIPT^KqDck*Nku<-%BF9u`aR^(T3r1pd3kyH)=W)21bqLp
znSHAMEj%m?V2*eF8KQt>gJV%lJC$k`1cu-=5vPw!l@aEvwpHa+`}D>+;d57R)@RCv
z?ql`?E<zN8*2zetq7l8}eDgatMA5I)!rz!<Nxgaur|HrEh+WwAMc!LlVCa!semHNw
zc-;vCO1I1+@EcInRs;s3!BZTtXdwgh?dh)<Ca?)uv;H(AH#8^qS^KPP!}%sBxh0m+
zgt}vBCoZ1>77dMMPfz8<wESyvjCgEuae5qkYR3@vNB${$c16uasS#tQ9X78bHu&@v
zqG=HwaAzzNue}6*3w(-jQnl<rH^K%z7N2rtZ${`rohmhBTi}B^$GY3f>k@zbLNSku
zk!i&BEgcro9SOdNWTRn|A@5kvu@ME~-am*skGu~4`r8BZNKsPJPjzr`5OHY&vkEvN
z|9vbhtclkKaAXi?!(#NOHB)cBA#(nJfNqU*IB+D0@_Bs_Uw?FEh@h`xZI2e7>cf-{
ziTNL3$>-wkuFu=0^BWeTJH+eMb>27lKhN*}+9Kp~Aw94(!@a9ASA3qj`PGIHUURp@
zl@9kc2%b_e*aN@gej~<1R-fxDezNIy)M@dSF|UElq?GFWgWm5C8RnED&ITL$iBkk6
zGGxbX8mcx`+Y0o|%sPI%9U}7Cp*O`L<JyA}FeWZeQdv3rcY1O0bFhi5LcY#2$QgF`
zJ44X9a=0Y0lf_$#c#*X7Smu-2Ehbcc$Cr+n@M<{+G#kC~GcYm?HStk2#!Ky3`1G;p
zd9A`fq$V^Y2(Lo`)DR{rbhRhWn_pN^@G=b_#JKDLgCY`DcOVLG>W!otQo*Q~UHOjO
zFtH--oU{yyZfd8&+eWuMQnFkNvo8gba({cUaS%f5uR32ti0A$G6^oD&jR1tLL1aDH
z^B*E!koo2&?y=U^po&@VWXMy=Y+EEDxf8v~MwC;wZzV%uc(G<D-S?hpOv0!t7qipJ
z<5KByLpQcoa8rO?rn1$L+Z&H(*HU-!@f%=wfkvq@4#*_QLEwngGMfa5`r)A5vwsWO
zq!oVQ;W+g-|8_$oB3g%s%hapwf&o;jtz?pxk(mdjBHYxraCd=?43g0Q+0_5k6?&>0
zn2MpVUSdsDAzAY3E#9p4z39mGM`G2_+xHn#OTH@R!gUBZpw@5KPlyeIQj~51P+{jb
zHrzc><nooGzm2e_3*9b%J-I!K(rEUXYVx&_?UT1#y3rUlLij@cq&VdswvnAr;Nu(~
zR@FSQ-AQr2*WA^GWny9iJ$!q6I~7dxM@No`BZv<H#-MDNpPzTH>q`A1Gb3>^vRFd9
z*fJ8sQDOdLb7z!>#Be?Ajk3eM<lP6>M0Q8Aq*8@h`|*oD2APYfRmg&^qh#XC`8P+S
z!^8ho-2JjjepwBg%umxrQ`6Jk5fuE;rei?bCa5IgVhBx7*Nvn+ip9a9)vfMs{7AP|
zZ~Mt_bA)dSCsy_%iCASeGgFkV&N=<ryVG-O#9MB%rbmzGOb@VS^Tl!Pc9Nx(LT*h>
z&9wa4)a-Q}__dvKCw6bk_#MbrDlsn+n;iVPvDiF1<zz*2Q2BIgco;i8eBtcEohO!|
zuI>!o^0K*op0(+<YGyEQ9sA{p&J~|6tb~Hnj*!4a)6Op}_$4NiK$zn1gH9%dWn^^g
z=FO|OZ@WMWhV;nh5>T0HdCZ*`V~CIXtmAuOAp;}ho%hg`#68n}3G^214Y-`(@U1uJ
z;1X3%>K8FdyDvQC%0c!xP;lnAnb_pM&TkE$g1TE46(NNKG-#i2lc~1$-9vUbvdzm8
zyE_OWg9DK@32F)i%!r82hsJE%5#$#cM`3ZSt8|S$&eoX!&_(8`tyI2<W30<8D)2Z+
zn`RsmyokNHvKv2fx!b?M_L6>hFo8;>of7V#M}%EqIqC|xM{qXi?x6m~Cjah)mJ8`o
zFx`VMsEA0yc>KRw0NCDwu*@O;>F()C6!PLj$jH(J$w2prlr++l6`O7MU0#|iLR(NW
zrdm+n*~Wqqo4WnEQel&4)lSLJ^p_y_y_Ulm_gP0mQfpO{CubG4mOZi<nZ&w(?_M#q
zc%Uwog&Gs#!g<@G41wK1Zp<AO6%}w+A!H?nGXQ+k!!!vE=={<WsK`;_JK!S>+zeO9
z(qfD?BI8|-bqrAcLe)R<NjT!z=w$B;xG|&~LQ8&Ju?zZduRiV1P=57lwxPn!<eW<p
zb%E!tCYxIk`xo&5^y-=?8t~QOUI$tNt$*Y*GW1Qcu}tbEpgo1<NLgPeFg{7T!J*JZ
z@j7aM%2t{|_>EtLM-taA=(!HQfGBEySv^H^@*{q{2y*}5rMCH*;Zf=#h2m1>my;7s
zNe>^anP#M>q<A04cHHzme*BmU_Fgz`jt)#tegJ_N{Jn^*K<N23OY7nKL%_n{CT=+A
zm*aJFDk~F=y5frM7A4_8LMCXz)x^e&zWyJG7M64^%_0M_BJ;1;skZmc<?JC;<wM8-
zDz={C)A9M|1cIG%%J6;RLf<&r_KpW5D=1O{YXkudj^Z?0j3^jyL)_2>CedPuh#EV6
zKFA)6V5}KnnoNrE$}3QDh`7o{+F=A{s^A;}D22GsKr<k6nxqWcda$CwVuMA{{}Gdt
z5<r)Iu-t`6$$;^w1gq`DOJ@3KWu{<dWNQCY3sMXUslPf}ZeBPk3IpcfKeCrUUq%nA
z#3sRRfT&I*3y2Em2j7P{lpWZ1s`xE5fnorbn%LWyXqAIL9vxMN0=_cD+h{RXOT@_5
z)a{P;DRAKMp%E?-honwyyJcxtyM;pU!It|pm*rE1X3`rr+dUC;3k!f7O0FCLnFC77
z)Xc0Aus%d=H$2XDi3khZjo~MVMO~G4G*ltZU@G2{votsV0NZg`dU~MP-tBqll`ZD0
z{6%DB)NR#lDxK~xCG!Dss1<iisbp+Y{%Y^xajw9LF8-nuIhI~RDWU#QPS;a%*a-<S
z<R3u!2?qxi!Pr1kk!mgr-^Mw7I+|7XbWllPh^(U4P*TcREu5yzp%<&z;zs{1*%&2S
z;lj_w)i{CN-WwP*VOuQW!S9R?|Er*l2^jEdk-qYa7NTG5WU%8n>gzv(^rGtN5PrzU
zhNW$lH`kZPe`|U<F163MZj?FuJ32b#(*#JsD+IQDgy9WOU=ImM-mILQllG49>L@G-
z(inmBk@yO^A3X;C6Kr=7G4BtdH1G?k5rjAz<evKn8`YuG{x`3YQ?*><I())?wK33S
z(C1U0dqvw0omr;ZE+@>PbfpuaRw119+kI5$3FCE^&$a644IkdKA_(<8VKrczO7jVy
z_!_X;nC~fDBQ7V01v41}bbz++vn(CF-X%Nt8-5q!9Mg!E2TXJ<o)F8rM7=@#B(%r7
zMA@HjNkg#7akmeyCdeqdMMxO)6n!Wd7<`x#$$4zwTTT7?l$KWTfZNK*@=w66TLJ<H
zkBo6r2WTc<DE4QV&jtI1|9p0q7PCykk>^Nu7#0>5MT;3L8wk<WkS@XoM7d!kJq8q*
z-zSq6CFD)R!=FJuBp7KylN}w$m1$vR-sDe7<K|d~AtV^Gxf2qdT>i6zw`W7;8?P>l
zA-!5_>v&SKeRgqTkh3*KTwL7nuPSeNg6;N6CoSjjzVHe4!xsm=^QN}Ha&A^Rg=PgS
zccYg?%$_eLUDkaU_S~v3uQ=#dmPC6=cyzM_>&3n_fSYZXl8u|YZEEUmu!e`5+srTY
z`u`v{4E+%LRdW3GO@3=nPdK7dnGR&Wed$gs^A5UB0AgI$OK;?8`;U*00Tc$~$m1z#
zuyulSY8)2R`9P^~17(YiSIPee+$AR^L0hs6eF_jW$pE8tOd<SMpd|%{Tp$?lVcqi2
z6Q58KZca{wRvKtpNGQLLrzwtdo17`@!D6kxHnJyObkSK4O3`EGGcgx58NcwFxyAA(
zl#BVQB10j=CCuEmFN^(|V!*WQ?kYMslx{70@(X1HvSeTK2I{|#eX4RUEre6R&lLTf
zHRVC+4z;+MHNN4w1aivsoe?WS-0WMeX^(zt5J$jA2LeVY+n*q&@>~fn3)tuI_}lUM
z<>e-i5&|_YDFIUx<TS`=l7l%9)}p~88f9Hw#f7m@+BAhM85YJA&<)u2&E<lEDmB-?
zY2Con!<S}^v=h<9rv9nnQHp*XW6Wg0WPGO3#?TZGbPj-<8H>MeWp)<i&IJVqK8Lq2
z^|Tl;a_Fwk1?3bK$K=?{ii;be*~EJAbj`|PeJuVq_k$JehpiF}QWOuur+iV#pH_9N
zPq<m~c|=tI+9U=b9!wkn^{-JZLdyEyd~<PG*%HJCAQ(~Ac~NPHi<50R6_uC>Tyj*q
z<HJLP^MmCx_pwiQXP@}pw_W`9+BdVZ$fI|2A+RWMiHw+7&&n#CRtC%|>rp=8V@_?h
z_sEh9$P{2enQ3JBYMD6Vo{AuvUjI=0qMVD!9Qj|bhTxC7U;eZ*C4F@{6H1Gv!bQ7l
zy5zpz(2Jk7312rpyxzkARU6IuDoogjaJQf!ba=>y{Cx=h)BkFikMc2(;66YkCWAhE
z0aiH-Pt`#jCouc%v)3ow0RIKPlr-fYy+QKrh3*rms&b1l<=1qK%`~;YdB`>}PEkec
zg_Il!|DHV{?+Uv9r?)&b$DX&!(82V6?es<Smq-kWn^&r3Hp5L9RvOlFS-&Ll#lb&!
z9a(#b{qaD4YL8cZ_XYkfGd~;&P+Q0rGs$L3suMZ|wj}7tvnzya3NO+p|1?(6eNJz_
zg-L2bq`E~-{etK-_3g*X+0EUji~4^Pzd)VuaHnV^k8@i!3Nl-D5FYcI0hy8!<bgse
zM8<2^-mb0~B$%-!NpEndcUa)a_V+6x@?qe$0JAyGvuAjKhM~)Q1kw2aCD2)|d9C?~
z_KGos*G+hsD?5@n(SMg`4-aG3ZhT6kJfxV3nRy8e7=vE*uKSsesd)!uxhf2>*hxlK
zHnU81hJ1BvOCGk+q2|Dy_-3{4*o<U3*QjNM@}1&|&WYjfvSj&pzs*i35_91Iy8;7}
zO*V8B_HXy`<U*rA(mj|+Ef3b7@Cbbt7LWdoCz<zouMzX2R@=f%F5ayAb$jQm(6_QO
zBC!7<0`H+Y=k>ZML;aH!a|W4bKx?+Nw}$|L9vvNB_It3dt`2a9!LL$yh}#p~G=@&B
z92|!3k=6p+n+js{PR_=L1TkK*N$9sm)Pm5%wpU~mg2}#4YZbAY-uE99oyq&L`lEDJ
zGcl$0IOo()>*+U0BLd-DAgCOm*?S`|GdwWRI5<cE^cfIJPaZ414S;H$^=YNCu@Mo`
z0oJd<p&<a(^fwHeZ<7#v%rfBr%9bfDGuVQ3WL}^{4DFQvi204bhJ)QWJ)HtB7C5AD
z#uZ^t4c%t&ouAN(MRgeB-MaAp{{8)&;4AW}q9EWWfpJHelM*2tUhs6|yZy-n$mjta
z4Jk4}!Kkc_I)(qn#Pd%D${M7%;}a8yK&t=(h=fG5hgZO=AVM}l2+;|Q?j$)Ao&f3^
z{mgR_9}o`;91p9UdUo$><K3$EdN1U0!hGx2EwJ!>ijMZn&NljJ0Fj^jASo;<i3a7L
zI~W~S&ab~CX{)NXK$Z%u>M1{JG&3^#17b~t?7X#BV9iUC!}XIV=KaF=WPj#%(Xe!h
z76*4!mn-IMe#|ZL+FgG8*1V6ZzmMd8x}y+DDahk~^c@v6*gDWRvjQ;=s!Y1IVflQ7
z^az&K5vX2VURHc>hph*Du;QTvHoL%<9MRd^l7bk7v#5u?74^HabW7PE?l^dI7^WzZ
z$ysTC(XD*HqSGwVtsOJj>8qBm2COqSLUjO#SgY3k^^ZZum=0K@#=y!Ng$OVOxs!7}
zNAju&)8{+AI>KEATeVxglV4*|wnsW*ik}(~FLOjEwUV9~?;5zAP?mI-W*c0%?6mQW
z6^A*Hh(FuZQ5{j=S2(EfOglrPE2m3HPG<MH_5|o;`TlM;pa_WS0**kCk~n@W9(8wj
zhe%9zsLhC}79tKLVc!-JhRM5kVz3Q%tWb9qE;%G@2YbVTeqU@S`>c|>G=EL?iNsGG
zh4Ji9@(Ak81glc0Or)9wrf)mAG-vx8)Wz5k?eBas*y`&R{i2##Sg0s2?jJXU+;Z76
zHufM<Ueen7DQrlI6_{d7OiGefONB_D`!q$#9w$zY5}q>#^VEWiqItX`-0Bc$IEq|&
z9s$BwJQ!AAsH?}r+XdWk`F%GkXIxOmRIM=6^wgmX{A}nzcBIrQT}a2xu({vVTBx_`
zZu%XOq6Wc$k-%=3b<5OC<qBNFT2u&nAxV^PW!+uxCa#Mr_*J{J*=JsGf>fQ8gAThx
z0Fy%a^#IQ_w+mCwHC$*3MXcLJXcQjkdKPa#Ji*WK#YHY#dGN{C_ZF;Kwz5V8cE`ZV
zYNo01L~2O<y)}g6ZEO&;>LIQ6rJKM^3w_v^p2MGSD<TjGFK`gR>rWlU{riMe4eD5+
z)nPJt0ju4i2^n15g4M!J&CS^@{vSu~9r2YDa_#yRa)q2Xlrl7?JUKP)SADEM5?SQ*
zKv^oG2J1$n`m_2s!3cT!uW+-%s8kFNN?3?BwIG<QiAcK1_PM#HC};tRAp5`O!qDpl
znpZElSc3+hH3f{lP2|ydmV@0Y^|9-iX=8I51(o9zE4aXG@<@pr!!%pD1Lh7`0+#oP
zkt(kY*1mpb><4)~I2L5ak|C~Y<QO<MNJT8uN$$B-T?I#RBBnN;B9Wi>m926T+xO_Y
z2%$$2jq<iw24RPf`N%&0dGT8O-fZ?y<K?QpDovHJ+_Wcskv<U>nfLXK=#Ycqp=E72
zz}o?j?5kw6)YkfXXhlVZs;`jul`ED7%I(b-VC)mQZPGYi@;#*wI;AltOlH|4`ttN8
zGdW*WV+1v6@vGb=dGAO3>;-!klJvPRHi*S+-jdO~R^DBi5AuocZ8u`O!}Nnk-PPY=
z%*IxLJKdma?6J)X3gox^0m<<N1w>FFx{<P?=(WvFQ0ct~njPT6urO>G{Aph28nKah
zdwUN$K)c5Zp~J{!b#-a6>9q~B!q-8YFQ|u&Rd35D-cH>S@T(E8rhqdH(+7b0%7NAA
z5gk#F#G!5c0}T)OI?Os3fVC=D%qm2&j_aR!gH&F_S^3tl@!$U1jTurkz%3V~&ISI}
zZ;IBN!p?07IB%->H{vVFo)Fy`^{kDBkH+EHoslgu1wXwlQONTVZfdn~RQd#^Nd5mx
zS`m4&AZGp5;{e^!!t(M9ad8wV!;atxIYxPi%+W9UespZiX1>cp5#aW?oU^M}$M_jj
z=xc~AxL^!LB=G@R(ZAycQwK~F2*p1lzUcSWL03NIsT9&0h@=oREi~;fj_>X5?NL!v
z`#~Ve5OgW`D|oe|<QmFkY?rV#5*j_8iHK}|4d?%vVo_xLL;6sO5sig|t4T!nm(M?z
z4_&+jp4nO7rIg?VAnL{I*diJ5<s;IhsHv%U80*$zI9wP14^3YgmDLumO9)82MFA->
z=<aR_X^@f*5h-bqZcq^rk?xf4PH7RP8!17gyCm<s&N=soV~^q3!Ut=vIp6u@le-tk
zG@CjEfB3!*-sx_-(UBJL0h*SR%Ww-48Ld4+7DVJxDvd~WT$x9v=Tv($MDOW-Evjz@
z#X!~{qHXwHhe`WrX~o`P0~kTq*woyd2G!-&EBcc26xD2N^i=0m+Bb74&7o%hdMs6M
ze&rhF$X1(@YnJrVB3mcBJ<Sv3z-or=h-YfCqO6sy*Bie6@PkWu!rx17wnOh&<%}Br
z)qF!dXiU-Yh!FQH!y%KBl+0$dadqWITqOWzM-}lwpcUL_i(dkS(~!T#C>3ISFlIr~
z-)q)86y0fX_t8(;2Ej`ERdF<msx5wP`My)0N6{(w5;_fjW|gmPw2RAP)ho)lFg?1$
zb8#WvT>WYW2i)bUjlAMd0oP4Z_^|+ulEbZE3C~9Y)}Sa5PO(AI553?_h;0>u8zeY&
z<Q7=s`qm~&(+-pC1^zl+y?Bfj`8C$)iEnWU>j#41UZU-5@y|jSt1D1HpLpm24DQ%8
zJ|~zfT?2uA)QP>P^kd5GqZ-~lnvc&)D6t<#V-N0@p%Gdir4hRky!U<M$J&gY)at)_
zKfMOCNT}n`!sNoa<<OyN>v8*ePLhDVv+W~?-lXTg#vJa_BvE}XH;+<!jU?xL=FSM5
zj_tB)Py}Jj^+(5o5+o8xzR8xh&^rS<|0ht6((tsf2q<1P>zw`17JLZ8)$FZWSaFb<
zp2&O20U;b97!=#RI~O<(+c0H1&@ihYQ3mNh<cd`ZV%FEg;IzBHp9y;<{EZft^MxU5
zLsL^SG~bj%w<<%;H_(Q_eeOS?;oxnZ{uPdxGKP9`KPJWCdwaZnD>&mxANq&*>xo|1
zq+RfB{l@k~?B8L|N5mWwl2UM(;*oL7q2Xz&cL5sk4osIFA@~{^Z`GzD><D=~fPAa6
z>akSP&1$lK-F8f9GT%BX`Z9)j^Q&)W-Qhikf-%aWa7Bh;<-d}6;d}?#h0uBp2~)9D
z5T0dmp4$9UoiPUZG3mbRV%<BaI!QWQ*-q)-++A<Nxm<vqc;c9G8sMD##bt-qx?AVM
z#@kYZ*HQTTMq}kknUFQ%)ZJ;Ncn?RBTo>ODy;RY1jPkVG(`w4U_}hxfTpHd1GJIGE
z*&0>*MG$rc6RQx|dH^y&0m2-PI;7GCuxV(OMuVdPe(O0scS%UT#H9W<F@IX-q#ck#
zITX{U<4zkz<o>%na&4tOKmwh8bcCjS7xNwLwk5lmPx;>0X-@e3*EKxoH-UMa6&cRs
zP(Q){Il5=6oqX6WG$MikNZL3mJy2lg=H@1g`KCfJ9wH}V=~IJ=4%YAIih7Br`*8bv
zyebSC2HpSu)HO2fZsRw5zFqD~^9}xeI|^0gWgUuTU!%QJb*?9}l0A+J#?e%qdY%l7
zemoOR`C5v5IxV5i^^G~OkLIHx^Vc(ZhEX29!zr@BM$Ko{4cJ9niZjhKzEDBM<j*qS
zJ}@t3%cJ9Zdx7GZpY$zdTf~R`7NcRsMOAt-0cRg{asnUmtDktwj+8Q|1128dJe_MC
zo7u!ldHhNW3$x((eN01B4!AIslD_t&f?pj}Z9|%g7!rIAb;#E4gHnE&+wJHgP)r0y
zzkli513NM>0n;GKK<uvusedmmO~F#94EHjGHc9|s9)mX&o=ZB$=0z%sZicEoR2?IQ
z)UhmjkBwG9jte<n%b+!huIbc^R?lAsHR%dKq{J~_TZ7t0xSrhZnO&WB(P_j#ruvWg
zmY(%4G_U16q|-ejj_noCf5IBaWj7}*3j2p^)<8i)!6ZOp$Vw2V5AgyR6A2`PZ5c#M
zPa0Lv90~&XYExg!|9m@6HNrB1DxH-l7urm)WQe*n^HD)9;#f8IVRfj^OPjXNs5bSO
zddMGgyYZtIKGtK<z6Qjhw^F6AYN>@SEprN73Z0mJ>dAsRHY_Z`32r~OMP|0sgcl-L
zMDGj2KK3oM*ibPF%<yxW{=SJFpORdabUQ|#2B3{tViCqhnd_^|9O=mKS3`;G8yjP>
zAJMV+J(kNxxo<AmypG6kq0;VS%6s*)Rke5aJ6c+Q&FSRSZ=2WMZAFC_pFN^Hg3%}w
z6blrsuS;H>db;1{bfw71$bf?1J5C2R-lG6JAm<;fMsWQqnVCImXlMuu4o3C~P@?GS
zd;n3SfaCH_*n~G{8pB6%r`A{PDOcz4Kj1xhM8x>ozeV%|Id)`3D1mOn%;+WN>I&GS
zJ`v;S9dD%<q(AY_Nf14(CSu>RremhWp{~_8$mypiwl^5s6yO;nQ7wL6>{^3ju;+8u
zf$A*HU~|c56B)41SXLHdd0_mT$T+WA=h_anN<REOrs+WO(wBIeJ9msEWJ~(;e4-R_
z-AVNN#>Er_3&v4A_I`IEN!&=?6P(hnr~2Ufg$W-E#&XET>hJ%#J%g-9)7u$<Z9+q+
zF>GLjzpsQxcXX}+*#O0&1&CYRZVVtz<wz_p2=Mm6h`$TPX2R*7?k85Z3Qcgkxi~8I
zHHVQ2kiB0kc&1O*fh$E=A~=}5&N|2pR~_SwDlqt-n7d#9r{*v|5$KWgO0r3fS$MR0
zR+=m{jpWUVv3U;1ec?+AB)<qMN5MDT@RLQz%*+fKOh6;i`{T!Fs&mL=gNqqDpvl8#
z+U|?6Tu<R<hVW+OwL7=)1*umaoZNdrhH|cSgUEWa@wU$;mW=_PrxpYuupiSLeerTg
zjhD<LOU_&i=jBG9-+O2@&GebI+3TPAelubAD+SI3JT5M-Wk;kG@mE3q$C8og4Qv&6
zGfJGCfM>y9dP}}7Dw;Hrmi%VF!-uQ=;#=Oc(LJ+iK~cd24!#0A>bPExL5`2Vcg1yt
z*2y<4HE`at6f@>DG>{n>7!>!PLIxI64+uWtY)o!|uMNGAG~wM6<enz$RPF?6G+^Rw
zJtNU=E-@6}dmsIi`HzODdiho7KYtV>X`)~P=hb%E45@pYuT)IT@7|{r>ln`Zl4ou6
z#7I^aH`UxPDwRr8$yF>^{i(zAdtfesw|9>BT*t^L4UQ)+_x%@ed(G*&0_15I{tQA2
zF&v~h<^mN}RiATNR$k&$zZ_V~&X(B6aCoYqon|D<I_})M8DMa3azP(|2Xp26UTT9`
zeM7@-U_*guI_fQ)w0|v@%oOB=@p~Xky9?v%FLC@4QCId~O2QB1)AQv0s$NzLmaFP(
z43D7l;o8^I&eVP~iwRttskLhi4!CZ39n(=W_FLTr_XqQ0FA235FX^SIMe)WrmwSuX
z+5z5FbSu9IFYX*X@KpBgu(QY_o%LmX76|p~nK^fNUprr2xI-4udk_R)U;Gu@S*<~0
zgCJc3X#fPmzZ}+O=rdluAxiCZ@EHC8`UqO6?EVYfw;1lgsdB{-WMRQKGBSev2gHCe
zt+kb>QK;(&P-gLJ*|io%Lo>@BT+U=7D5e&lnj)+LQ0G-u+!vRZQ(){qI<7IjZ6d4r
zjOk*S+)YxZre;8mHE_EGE%vghQ_haamR?{<EU}*^2BcZxfQ$?l>+wO%cmS}vp*~HL
z8Hg4VrqoK{`10M0LH-#ZA8)_dj<I(A;Rkd6F%{j#4^gwWy1N^)nhcv~iuzWJPo$@Z
zh+fYIwfC(r;VGv7Rpt1c<94&|h)8uVdnU*JS2zsrF!6(+A`seFEveD`tm(A7g>bC9
zzp|pPesh*%s;3WM5l!^{Fs$oh8c-a?(%m4ltbJE?BF|v2lB*^<mz2PX^nGDHZZ&>2
z#do0}+nXYr^)-?0oTyx|{~d16lIN&1@Jz3*uCl`w4O|2Oe_z-}ql4W5!i?-VK0K-H
zz1>+B-I8LvTT|Wk#3*t&k&N}uT_?d-RkDiAPZ=ew)#2eRWNPP=t`EQEGqtfs0MyVq
zJ@~QxJEN;X|MJ8KMS^9>%9LTIVrFJ01#lfMA|Ms1k<xmQqdryehn||+51t7Wi?XKB
z@mZrozH5og`96-^)4*&E)Y1Cs|3*ylr-^7o{cFGMD!OS9HlP*f6LYxAKdBZdzjZs&
zuR+6%Ix;7NEYG`9T-#}@QCE%5Pq$Jx#ekN0UFlQr3+OR#?sD#G{I_YR;->FkQUp=R
z#n;5XeARyWPK!gW0cz(RAy+eZS68fF6)n|PT2GQ5@V)<Dl3AEFQ0M0^Ok9Ox?G$=!
z_e(MHE~4$n#3w^SBN3Vjn(P~v&P!G4NkzT*LkTD!GBUmgs3Nb_OUEB`8x2XM%3cXU
zu3RTB<VPpjp(w?ehy?>F#RawAbJ$K{hpyW=0Bk2t7&t+sbQZ$*<qD^-AB(GOmcL~+
zUjkH&N9^)2x{7A5)JO~-kzFIrTq1<k^udNL1^p`nqUX#ITFd~KlXbN@+x!wLb=G(O
z^mIgv!*`3>I0W4S(ovvE@$49F<d})6Ky40=LP(DAgMWtySTdrBhFiIihO{bU%wsyb
zN??K^G3!eVK61~PZRCRyCyVn1l^tCzy3-%NC&2<d#GiZvFBHhMzo&;`;ki>Z$p)UI
z3|b1IF8iQ5dzv+YrvC{97r=w&-*QdoXf_X|c>Ce+;U{O_Tssq^exG6a;cOYnm=Dh3
zyW7iMZzdcR_u>_`M`{Y?8>9}WGR~QxL>jPiLniVm@SL??{9ywLR-m7sA8}eOz%}q0
zt^FNMO*6HK9{NEWA4+<Q_g=%tx1}j8CB!S=lA5X3$;bs_lLpR4xlJpa>U9kzoSwtf
zQ&lK3;&EbJ)X1~~vh&Nssm|pUA`r}gN`szRgY6Ckn<wx))Wr<6{eLaML--Nu>g%iB
z31X%zDWqj&hO=f|en*N9Z41cv-B3+a2pGXeE7KevLLH!5SYn~Q*U<R>+FIdS)9*Da
z+^%v<l+&1k9K(~>lx390)@TW<1E2Rb8cy6y$azB&&sp>>7E+m9jT919w^mkMCzuCK
z{7OphT&?_^Z!hL#=Llp_Y0tj7cS5jwf;q@I#y46a=EIX@8s}IkOz3SO)r}$C@j1P=
zjB`V$zlBzu4J1a;d=ZLBf<c7d=yl^`RFnk71#+;ngMm?cHgXT{T#YEJ2pa;Y7A&B!
z`g#EfeXyy1xG!7#M)<}n9&z8}DT^OzVrht$8E7-l>F;Xvci_SxS>P!1<O$04J--sx
z;18sAuAQ-k_cF;WV2-lIb#qqG-$))CLu$&r00o9^Zx@*;PM6jhb#VPkN<&BSP=8?O
zfBeLHrSeQBgxw?WZA0a@j#QdFqo0=L`SBK##sOwEWG;a8&Z+ep8v_GMNP&;~sy_BL
z8?6sX{srYKD3TuEtZ><C3zzth-G*03;A^pIoxR+a+4$j;jeFuJZ408e-a9QGy5<$$
z#t&J<HSQFSVwert#z+Up!Bn#+pOEXOF|dZiHzZ*Sk%^>If@KO^nQjcvPf{TodCtCy
zw|!#Yqnh|5^uw~s3RE{NzP;dVDiYb-`4wqhyU%ZEAMoA7WhphOo~eT(gmcv7Cj4i(
zD;pPbHf<W4Qe5iA9C#0ysed2QD#pXw6ajN5Na5tgZLMKhTwOw73N8qoU<KE{4IC*V
zE1>R~n9xp1)7Mkjk~~t)!VC@wYTxL5`PYG-F3!w-BClY=5RY_qeZj=MxB9)_=h0V$
z*Mm{LyF0`6`3&Vufs|VuW7=#%hbzM2O|_FQ>hyP#MZI+kPU&c9p7tEAjfBE;1g3Cu
zT{5Gdn?<%j`eRlgzCZeVQc=83L7m6VS|}T}U)L%JH%-i5&=-t&djz`K**gs>r(PJK
zIy*7MeAGnNFx%*{<P2HZ_o@Q<S?SA!^fmw2Ha3qzkZ<iD`|;zTBVWf4MeBl05CDL4
zjPm>Q78Y@BZHYr&)c!v-ncs!~JdLNo&UJhF;<_2FO5O#ZXH(>oMUQ|btM1`_Z;?ms
zm?8t8P3%lfOlP)jHBr3bzyw^FgO87nhv%-|%(0!DQ1#j6HL`eZ9@>bC6!3ss;cv;#
z=#B&-sxFFi64RjPASlyd1Yv{wJD>vKDm?v)qC0d`Tcsmsp878y{axybC?a0b4+t{b
z-o7oUC<a0>P?CO%kAE8%M}j0RX|rvWT`X55y(!hhz-B;;k7ko7?j}j7XVB1+$dXi~
zO%}P=F8Rdft3jTCc;KlwB+$ixauiqx<kE)3BtXB~Y@-`&D6?qHvf*Sj{y8hV@8P@v
zlH*ZR{*aI=H}BgOX(=v~p^t)fhw10kCOx}2cXx{{@>-t#nK1d=Q)U-Kx@c=j7}<p1
z@z?79L^gmp5UHdP<_|Ba+(%10J2p)bpr+??+4zo{ADc#W%g2UW$E-K|ni-h=SFUCB
zPM_AoBhs>I@mjEC#Y#cFIWx60^PtlpY)NiTErKLDy+K0cai_+>=k*Y)UMvxaS%OJ5
z9s)(6Q~(ghPY}usr;>pE0ty&^b)S+^yP`w-2=qDM1pPl)re?=$r*$;DdMzFae2(s0
zOcst<F&wd?sN&>a(DD{J3ddo_$9v<gDwevp5eBz4oSaX^BwRYT-2x7OtH1Dz$#_7+
z!=rwuzC>Ty*Hb7Ug#z1k?R#16T-Ru+Z{Z0Il-fcdPZ`)5ih~1S0wkhUq!t!V21Smk
z&9CE<u-red`eKzv$KOk<q|LE=k>Zh|giziLDkum@IS&$RXpqqnz&q^Hdq!hswrhp9
zF50{p<3{$e?|w3y(X$ibcd8jdOf>83&ZZAEESY~b#^dZ%l~Q&{g*qAk2(?X>&)4{r
zmNt)807+DtF3G8>R6rr=s%|YUjbuhuDfyKYhpgV0j&*uozRRQX{?@LIt3@zIjL!~A
zvATgbPos+O&y`8@&zgoks`EkEfriGzWfL5Huf0cCzL7`_)I|YP<_Ka)L}CDUwVlQP
zZdNy-U&4|GDecUtl;3vf(&01U8XZq}7h|dNJV%lARY>2DXE#9u`mc#(WT^$s`6;iy
zAj%qt>Ik=4`!KdgM9|+<kCZ9@`|~6JYHg+@0bT(u3rN+H*ji4}zzPC}D!A^&4O0;G
zP-f)Hn>8dxi@fm=cc<y3!^RD*rupG<hZc_{Jm4gK9EGe?>zb*wsU5_3;VmtqR8&;@
zO>g!hzN&LcG6U{diH%L4<ZeF17tms6{96w#-R{dU+T=gC$!nU?3BT$`!ute4!VCrD
zN(f}m%j>wm{InMc;G>^CJ+c)q!O>HzJKF>(j#hWo%b@s#mV5pNm?oC5y50puzlgJV
z6xW+D?#*a-+A6~rHZw!W8k2i45EOHfSRJ<CuunW8#LtghZD1(v1u9ymt?-4VB|<4d
zWqU4Qe;`LFrr3b|Zn?-I(2q>hAp4-0ui>SKu$PYY>if!i_;!?emai7X$2A$(*UMnR
z#-RA79_)lYXA0|+$@%n?Wcazq=Hs=loYikHZU_kde(}1yzMVX|wM3b1Xkx_7xRPFa
z?BfF^uqKvU3{WA449LLbjv5#o{HT^CgZM=9Ip<?pA73WLq=|I3J+jYV%j*1gD+KR0
z`|L>8PXUonjD^NMR1WzipyyfLU?`#>VQo_i|I!7FEtC9EU*CrRbF!Utdb?nC?mGpE
zx1&p>svFqH4L_UeF)}ehB!(Jp(KB^*-8;X)jZ*C@ccpIL$AXJbUsUXGG!eGuPa|z?
z{P^N$<WGz+Puzp6ER-sF9+#L+(SJ00krycy-^oeWlWtR;az+xQL?%QX|GoG|vKVG3
z_NRKna0xWd=3uZmDc)15#A2kO!D#HDAw*3Kg>FGAG-ol_&ocWL2LoIXt#>{bJOhRc
zRn=QS@GL<v$Kx@dJrQ*0EtWbj9gU>0MCBX{ylsg=v&YWc1mCu8Z=1j|Zr8Mj35K;!
z_*H>fg#!G`mM?bTt<ILAL!sl-F7asaPy6yVb;kEj&Gvsk1GH9lJF{+yN`7wK4opOM
zh5-Yz<31IXCmK(IOaN&n1sN2;-$r3<*Wf*|#4m?6K0#)>IQd0P+d#67ixf|Bu~X))
z-WRIG-@kF&{$oGHW^Yz9eACPG_Ua~q7e2nrU@tapGnRytxOS^1QW*OJPG%4iRRY9k
zsrOt(B@*NfK;R{?n*?Z87RV<`F(aZgujjLaudJ-(jVIdh%JE+0c~ei_t@rXeCStu~
zBp&jir$J>A|9WEe^^^~z0<dJQ`tMeHs*Wi7NOrSIvs8@8?TYY=dKm|P%e%kCz2?5j
zBwK3O3@$`-Fj7Kt)T`!exa&b7*zkb|1ia?A4S>f8-1x{vOi3vx`w6A*_}EYswE`EQ
zv}<;DJ^ddPuiV#}0Zq^$T|F2^5i?x<8zzlq7>SRAk;JP1+fG)+A*Tw@pgwtOuGzi2
zlnv+zz-jxU-ZOq08QZaj$sd!Din(p-t3SHlCd$~v1Q7`dJXWDo02-?plYqD%mXx?#
zne*u-k;yg>{~J=%NSt{tzbug;g20atYVQ{~7T(Tu3oCF|3YS}d!;N=whQqJYg0hxi
zpi8-u^<Ede6(`>3Kbm$>X7%xL>K+dZ*gRv*SNt)yFA9<2Qn-0Zks5XiV8Q^Eo|VKA
zL}b*tY{VY>Yw<3OS9!XvHOWn~vj^=yXFsxxTJySpD2x6o#%ld_`9UXR4ds>4-LWoG
z@^~WA<-1gr`;4?9nPfYHqA93>=cX@SfO}WIG3;mK5QuY-Rxb#cHXj*97B$w5!9uO#
z-=Oqwo&Rzfx$aAuZuhPE-qH450Ei%mgr&lACzRE&3Dy)LDE5AwbqB~_3i(Pg-nC~@
z6L%WN=HMaw+J@!v0cD~seZ~g&<Ta|GkPspua{!U0jg1X(6Tjg$rF&Nd@&Jn260P)@
zJSZn`Cb_o7&Sn;wn_xzUc4TD<*0g$I2?Cg4$cc_PG!UUUba26{#=4;s5)^a`<o7`F
z%Rvsq#6%o;6_*wlB_MPFo=^1GeQ-1vN2n3N+5zCq%ju`Sh@5oJ85#cpF2(zRY;UFL
z+ehiwkI-_PWvz3pt5h9WFDmc|e3()d^|a#;sgAQOoNV7yL1sLHhXSmCB$e0}?=FL%
zUICjBT&|YoN=0N@MQAGuaijR*phpD|Rl}nn6dJNWzsqNy$X0Ita8!St`%tR;@rmr4
z%K44Ms`aUrS?$bajW)}xl!ipggjV&D!OuFG)=!kA@a0GE*E%Lp5GoLQdH!4AY9sOX
z_J+n{e)s*yfZ}}4ZT<d&6Ikf%d2dzKk9U;vu!mtFTLj6i?z5Xf*D5P3BmNL5Y9sb#
zV|zdE6H|J-|6nb@@{aa|ac=2=!lk#ngCAS08wX0EM24~Vcj;IvA!*m=0v?V6PG=tb
z;|r*D&HQPZ$|*S*FM*s!UNT!dK4rrgnM_@JBuPe>9@^C88<7?Y{nnalX&hOzN68{=
z#%DhS_UsZA{dV6{ukK)ZY&B8zF)hT+mRYCS8$Q>4Yr>o*Wup7yQ5<WI^_G6#oHcb3
zec}iQC<T2c#4o?1>PnfUcukF;cg;jZPb5cJnvFCRr}F^)%WBlZ{9H_Q&i8Z3XywEA
z%Fi%m7Usx>n>xE@iS#}F=4MIAK^0T1?qZ7}6ye){T&Gf}X%O%(10#|kVIZiDZND@y
z(*AJrzn1GOPcRI6`}(q*5+GnBfDd?jhI(zEhq1UmH3eVUpm_gpTD0C1GXKdHgEC`@
zP_N-ui%T~2Ds^m?5{|!e>H<c-mpqUZe~5nm{D0(<pk`aW`{MiHgx%~FrV=BY-ywQ?
zuG&<Una%jUhR9nam9B|zw{N6FLMyy57aK;`0DBa4$$3IV#u_dsH@60pQaZqoC!&V^
z1$UX!D|Am&EEI#DW_8W&UuvjfQ;78qW>YBlP<{G343@dFGIqG3d;?15@h)^~z5qj}
z&14xWY)`<Lbi-U<y-+An#S=+6lK5mY@wH*_BkRRMGJ2Z=NnL@rxg85P%XWtPmwNrv
z;@UID1}FV~jpC5#DHK1s1)C?5YzrGXq^~^DeiT@gmnVli&CRp}Knc{h%17El%BT8W
zjeLy?*!kytk40`D^rP>})xK*Q+rjtHJ9+KE7eimK8`YQ5BxTb(DRXgDsrxgK@!l%w
z;FoBUxs<cP2`Fg?gKCio$1sniZ{CA$u4?Z-7!rn0H$WB*p%KNJmu`-Z?f<JP>)l4l
z9AXFg9JKDLg5(Jf!8%u^(^?m2XL}ddTx!0!?`Iy0pkvKQV`IbL?P|EI*OgLWn6c2a
zjC(O(KBCW3DsChj`sIo<-0kX$BO)SV*x}#5f4T!o9u0y;Q=q8Q{Q*NHB>t5iH9?w7
z3Ov~uc)<ZCrPF*0F}~OY`|Zse4j0~Et75n-_i^~W@CYbEHfc}75*Th^wC_^eTU$g(
zKNq-Y!gamIba62@$G*?}9Quw`6k_k>s6Vx`DmbhopcL%<nY6x^PC!7g9<`p4f#?xB
z#4po;5((4*F}i^Je)i#H#C@@nqvwlTTPz|LAZMeL=KzVD85lc+53Wo<PaH^S6Jp*7
zRB)?{#gBXTfKB0CMo@5JGXv9VEGO`1kgv0!R(?h92g)9;-Gd>O{33=9@qzg1Ge$^H
z?gIDmeVvz;R=3Mz$_4U?w1;{ZsJ{H?^2C?cZw_EwsQZ(GwODwrM*}ULqL+uy3_aFz
z&RyxQ$<K^1SAWtiri8P5o!a|Ay#W?k`rP24IIxvSKxn7l!g&9Ab_OPAWLwvz>c@vy
zD%$t;LvPwS#fjFQ@JpVf6c?;oYQ<<Q1xt9Tg$4&dH6n(TViq4UC>uq>NqjGs?q>?x
zzI-V2T7?R|dy6o7wd}KeT{61tQrDkMkEEH74yoxCD$W;gZ%_jY$I@~`GabKun8S9O
z8$ju&_V&BBJyr9>={!3dUHob>qUG3DNhp=F=_qRwBVXl~0_$3|vfY!zLq=xOpNN>w
zz_LO-QlO9`gqVUgHzRfRObqixY{Ij%GZq_k%%6<KAbkNqvCMu^vOoF1H7=8qrRRg!
zN@vcurG(b+9%aY9v~O7ze##du;PxoWiSKs4)nLTG;pmN_kpe7<h>qRry|$TLLMmda
z)$e+2E;X2gi6<LiCqzWkAVYV6*@bv!9vBul?BECJHH#GK231@R_Gg^n0PxQpK)k9W
zRHGEP%eYxHQLW8|V^h7xvtJG9_x4De;Nji(9>x$A^hs=buLpiHgn}vq`D}x;<-OjR
zq$H*-0sv2}p7TP67&tROrH=qu^4%Q#b1SgdecCbLd2hgqHIT~wl@Rk(D(+>NWzOxA
z7hwg5QA_P5sA%b-SlH>j)>(v=Rd_X9LCsjwZz8g?9zpaBE3n`!hR@(L0yV&rKCZtB
zF@}~~T0n%<{1dvgLw9!4)Yf)_t#eIaV>JT%W>`PyXud+UBfST{vx~_u=lDH}Np4SX
z2uuEhi<Zl!-kJ8CB*WdmBrF=ml2C<5->t%WFL>7i^;Ha`wfyswnTM_v*A%KCS?SWV
zYwm)2%<Hcq3(U_ef~3QUIY!L*#cs4-iTi}aUMZr8Kq>)}-yN+Y`V<0pL97IEDv-DU
zY}pIMS=rh7N|WO1?U@nAYJfWFiQ6lUZ%?20t*VM{G&6!Q?tW{_Y7)J>&0SI}R;w6Q
zdkGzG&ZQtJlyoOHlc7~!yu%&))z46JaDS14oSdnv3gW;<CniAr_7jk|UKlNA^kHw(
zK+KQ_y>w|1LK^@M)k-6;)B*I_wq;Sp?gBeMr+oX(BEOY(+OoLbbQMC<;Yb!IBhv~a
z6wZVX0C^xvE0G$|B*~lgqV}3?tgLi{@=iiaix{H%O_#dES!8Q#G6~SDa;;vs_5MW5
z9f^2*&Bcy+{cf<H{EmJ=r}x7l^iIE+7Blt}iq6yNO31drhF%TOcY+7`3rOxDF9lSD
zIg}hCA|hyLXxVJL5Z_pvP?Mp-h<M;Z0JaR*1Qf(EPhyZNn@~udPyg;f`e|}&LTB)A
z+)-ZgPnqf#W_sO%k&F4FD#Ng!R|vi3vnCh>I(68In20EAj2WCX@Jg~nvj-A50(Tg&
zUH8G9n%z@cSf~IleV|seK+1}QJVPrVBqvp-zCrt|kl(qo>_|_iO%^}g`|a)Cc8TKZ
zcK1yZi(>kQMon9@N7muY&ucvJA!(um-e8z(VgZw_G$6Efb*+h=rKO{j-Ppszz<4@P
z41P$*dJ0GYPc@1wUzy)=5JasQ<!kFz)8WB9l^}H6qHh3mpNp184pB6Q{>J)yGua#d
zDrw`D>0t`g981zIS7+7`wP5e)_`CUJ4k^B{%{}Ah<ptb}-n$=($(3(GHF-Mp>*xBk
zW(wt%Cpw5NW+UE#w;c9tuxUZTs0=(d;EKS(Sn>Llg0WaA)&);oN@V#3H|h}vj}Bk0
zpN<ZXW}YVT$GIXh0l&Zc+pKL%%13D^i2m?gElaM+%dgz5o5<~u9r{Jv$H$SMHX5p>
zOJ?sRHL7`gi=a20&~ONaQ)Ur!wFxY=q_zgbYhY_@s|1lrRnyQAGxj}5QCayVgw}ci
zQF3yUdKG%qdUH^oM);Dl)6kFbrBM0qdGgHsswe}0CQ<0boh2%~j8cg|N>Ni@Ykc(9
z^0&Au0=6^Il38}fKNSa|EX94aIEfM6)YSV$Z8#5F45+xN03#WQs<EMCXKT7zN3T|~
z7G=n$7Q~pKqcD3t*YpmYkx0A=@+TNm(2e(9S^amrws%fw@hUU{j6)rRC}>{PgZDAd
z{c{AkdH0)tMO+X)#CXigs?2m&1j=vS7U`d~(U69k>b3tA6lCwg)gg9u`Uu_!>ywh%
zyW-Z2mSP}}vve3tk7BcVL2ya+pOlucl=#3$R*kZ0&Is6Xkw*36VuyPD5{w@J%Bf@2
zurS2P6Hp0%%ajfS?K3?J?Dh~fmjH>aARVt}wi4u|wD};YHlstQL|w&RW~_UXFFfRg
z>22Gzb2E|0rQ?8>u1NW*agOCPfLtdEzMh?)Rtes#1T+&8zL7i;c(JM(cc~e3z@TmM
zd$Rm=doCiQgy4^h%SCKL2V6ss_4_mHdY7!yv!wFEh=_!Lk?t(M?-0?b09ab7xKL=`
z>j3}0_u+qFnFG`FK|3znE{sjT#)S$Qf8_b_f#72K&!49gK%Of+B~2A6WMF!On}GNF
z&N~SPi`&#|0lT7suP`;*s#NZ85eH?Oz6}qIfAn#_Dg?Sel|nlLzh>N9QcBFB=HrV5
zzuO*Y2f+6dSy;$~R4Ib4Y<S<XHW%C*(1?*ekO5HYt7ao8Gs-^Mco}Nrqm^CbRCE5N
z5p{K?ATJMbE5g<SzfVYJY#KlBh1CkyYIYFJ<Z=r-D4bBSJdvllqaXZ119c?sS?9c+
z)!*t3@_UOrOIbWOfWs86rR4JEJ5xb9oEw?nFHKA9?(x3SGqC-oVw*P2YabgdjED{a
zB=O~N(1Gv8&FKEpRIA-1<2H@W4@Ghb`*&l{wtVO2qLEkN^=oBI%ly&u`;eKJD&m<<
z2k{FFzkg?H^89<W?)zW5uPC}*QRvm{vOgU@b;1rS$9xSc9-e7f^TJLEzZJ_ILX)&R
zEi(17lF`%u(%T;rUAH`&Z6KKVL@kD0lA`UID`7D7%v%j^g5{utCuTGJph#;56h+fs
zH+gaQGx(f)fCW@|{c{8ttv@tqA@(tdWQ{XEuZg~?_w2Pl%8J>Z{nf*~%-=Re#&Ti~
zdk1I+lyU>a_eTDzKj2Wpp9J-7(l~MXOX)%$Sf>cp2vR0x(6R0`9T432+Vzjc(tPHI
zH~3ygPK;o0==UdIHE#nF!lA7M->2M!-(NiJ%8DlXydt&o+`;n(JNUweU|Q4FC6mv&
zZD(f(ANdCqYE8fV9RHo5+~?Ax<$PPH_<L?Hh#vPUjxr<P(ScYF=tllj-%lRWB8~~k
zgC!4sY#!ZM((g7+clwL36s#HCX&}VNnOoZBkfVtI11|k|<h+U=bD))+zPN;u{8<Co
zFreLN@P8D>(bLnqtcDubCHSD=Ts2k!*_>qJF#j^=J-Yq<e@BkQLc^sW)iP_yPTkfN
z!3_eh!|Ke`aYQV3VxR`1Z9xLjTf79|s^0^L3gXh70Pkvz>w_4;e;}kx_pZlJ%12II
zF3Od)U;O6Q6LWh5Egp!f*-j0PVA!$U+zq{IS${#0qyA)^QW&jA`TbW4Kf_rJIl_DN
zx}@A#`OC}8h;IbcWJ>UBo^5b310Rp@UMCTy&lRYusvok0^sdVFgFWX2m5-@=+Btoq
zvlt55lP+)Ic@bq1q_aJU?gd8UGno|4;ow+ss^|h2sTZdImsYujy8DdH`6kzOezRlT
zB}To2h7=pCgN52`EP2OAN4Ay5!)hY0{8$+qzdJECsWRTZ*vNUDs<TR~^sve2QV=Ou
z2lOQq#4})~Mnc8`&YQ9NuuN&~M7G0f@Z*~1_2Y`3S63pNsGP%T*y#{!(he}dJ<cb>
z=>SZpJdLJ(dR<ad(ha4GeR}_7N3Y4m=kxp=?n@|%Y`#JFI8sx2=rsscQVVM|8Oih7
zG6vBpqu=j&M4(14yo`+?$g9Z1j!jChm(04EVR1tg-ILwBeT0S4gYWe(W2i+0#EJwy
zn9tc79rU5-CbvS*HW*J6K{xpp7*kkO?hxeU4nc%^8l!?Y262Mp<?{Bq{^l|Lc6Ph_
zm10t*bHUuuHqkU5?H$QNY{>O^e+c#?+5O49av*xE$mYlYOW}72K)|A-bxqJ|f}YB|
zFiMURZ8Z^xzf}drslC*`FyFDLkkGIr!-)S>35AOsb>gq<JBDCB&M~RyDE0ODBXl2L
z!q}~w`u+3gJ4n7W;6x}ON;d~#80acuVf2wpxdLPRBT><oF813H$pHgWRfq@(ms>pd
z^v2UBsg)?4YgfC8L@d~wdyLBX>MP&mk&j8=X74cwd(^B!x$!d3TF$ssTt-wBIH8Zk
z40kBioK*6G6oK}aKd`~Z5-Y*_CAs@+-e+lHoJF5eOLpeo#v143kYOV2>U<7YLXP?7
zk8dw5WN6{7P~#RMG5D#NWZXb5(H}ut#kWCbi`?!iJ^@b=05^OEA>M*u3)#d0?wqO0
zvh#^*3Q<ROYvJ0QU`*8$Og-|t_y+|eUj_=8exhu@R$|}6Dp5AY*M5fHql_J|Y^xhc
zy4taTqI0+?0Z*-(*d`$%0ZDfTZ_^-v)j;<?ud{yuI*V5>f73;PeRWv>Jei<VeIScN
z#46P+Gapr$!+mk%zED}6*gMe^9C)k&#d@_BZ}v=!^#8X0e=UIevu9C#e-KS}6WEUr
zS09jt0;&9eia{?3R(;R3{({)~D82UnDKDyvU8|@3+~8<jvu(UN&0hq|k}?KYnJ#-+
zWu;SC)E@w~D7UdaKB5$BF#a>PCEbg2<+A(zSCRBEV4U7A2H-Jdmp~Mzpm^BwINp&<
z8p)e=!EQ}6T991RJdU^|&!W8C*^;Qv5hD`6FX;RhtQ@qUi2~_{0Q9W^tL+5@J`%TH
z+j@x)%Nhv(^YrZl-zV1!@J-ust&NRtH}U+JiZe#_=aL#m{!Rc~z{dEHgX6zjYV9k>
z_Kysfd9hr|2&)U8_*E3N>z8Z`NKM>ocoQIHRCW@gJiDMa5UUgfn55Zsl_i|vNO22%
z`-%21&!0W(fc%lZ3iG~i0|PxU<su3$2<4Wt#;eMmzOlMa#_w&2a<F%U-ui5OV&rbZ
zkknZRwUL9qBxABI+KMB7T3R^Isb@gxmpr+d8OPlDcrC2_U->YG^+JIvkT1}D;|e&D
zKS1@xX7B}n>+?5Ci93Clh}yt=h&$~Yuq)(=*)ev9V_%2{lT$e_65W0qS9wnL839S;
z(}gWCQ-ItJQ6s0Of$#tVfcjw$r)`dX!TG$9L4vv|7j{E4*c15q`KhCOp@!H3#x&^Y
za=575O-x)(VDG*BFLm`FKYHU0=gCZm%G~&@i3lFl1GQoKQG4U`EEPWkln#ne&L1Z9
zS1Fh5_?&i?@5)4YxVg(<&O}~YaQ!}3S%bvenVQgM%g=6Y#qtCOcb6Io|13M!KcVb;
zRZuDm6BUta7%D@n`JXbbOPq5~r#u@gVy>|Xv#KOqWYC)Vd3QtWlesQH;}CFiI?;O7
zMVIC!>EOTt9iE7e0|Ei13)hOWzW<#5n%@!W#G|szO_w7npV;slFZ286EO}F6dP9Z;
z{s;-zs5LLS^^hO`*ZWN#?`q9`P$_IG-sFwnM~)D<I3Rk^v9lB-NMuG@(25KiiQz^D
z90t<42LOAs*~buC!oxBCI@N3{pHI4v_mec&U*?o3pH!n)HU{=LV!VQT3|xGFK&dEL
zJtb+QIl#e2%&I?WDp{EO!>l5|&Yh?~EvR$p;~TV!v0mm?a&bG~UCPG=qs?A7Ud#ti
zTRki)E&U7sTvrvs0n^70vD-P9H#Q0}#$f2Rafw@$Ks#&;s`je+Q~IW!z-;PS`L`GQ
zTBy1?ARF8RmTNZ3qSb&$>7lUu(nEitm65T|kmX(@lU%p!Vs>9fKg{f*eE2`5^ELLD
zqgB8cgYnjO3Bus!4la(aYp&u@JSGI*xU(F(gebB3M_ja_IKu%8b3F~Hz5r!|!xc6N
zP;ilDq5ds9WI(xR<BK6*+CJGeSREs6>yp6B9tWuntLuR&U;NCm?8?^WhL4Au-tCU4
zh?%zqV8FF(ADT1D!MIb*Ru5waA_xZ)1yG(Hz@1s7x`))xc@&Ompek9S9)_Em^znNT
z_tS>uir?CE4ZM_&nVQ2KX6=~LFzKToi`TU<N4+vcE4ys^T5cZat`~iCM+P=FB*6#1
z32Nr~8M2QvzP>Ha6!S5LTAk~6R@b~{#STNkkCr9zTy)FO8|+{1Yr6%Z@q4_FnorFb
z6w=Yi*0jp&U^J(bQqP<OHBESU__sffFdjV&zdQb!R+yZOO#WXr<h7VXUhlA^IOy6}
z2Ooem7Fxu@stl;^-055*7)dY|rXp&d@jREeRuG8w)vtJw{Y%4B^XSjnVeLi_duRrp
z6NUVB|0Vn2P{~DAT>DTWa9~rQ5LSp~ejp+OkRBz3Xmr4v<9VJtPwV%(1kfYU*3#5i
zcDiXLoVi$(ln&gYLmdTm3`2gt`NSVZpP%0zmr63P!juMonJ}o4vi+Kcta=(=8}J<?
zK^2HJnx2KFYk2q*jAP+2mgP&Qx_J&$WM_U-XLz{riR~tvvwR}yOz|fr_|o3O(qjS}
z1%7KzPpvjWq8^-jgsoS#My_IsrlDC*jhq5_l3!RT1xp-!nIN)2np<cTmU0y0PIo&9
z5#1k7Fd9I1$N&rRq1&&lObV5Ps>44zN`3evs!hGn0CpjL%sM^X_~Q+<xs<kSvqASS
z$XPK3x@fRKA{-M6B11(QUjF=XnVG2-{N8w&i?Y^9WWUY7Z1n+~j~%VlXC24I&CUdU
z&2Rg0uoP$<psK_4g}akyK}N;{NFmf6TZb@Ib%Pfd;wTMu57Yi$tv&UY!cS{;rV3Sl
z7+Ejkr^Ud>cTys9t~*Tn=dwHB_u%QUR_YdIH;F%?@l{V6e!AWW&{O$EMNL^O5%ZN-
z+F*Q(m7C4W3t63L%gSH%5JNqKHx_9(gk~*lS7Bjcq@x@}F{XNJ#%`9k7sC~9U1E~D
z1@v6e6qWw#MT<R%PJhzTP>@A;TR`LPO20^Ycoq9oMn6^-7DSH@-Zi$#$;n-?2y?#t
zLAqAAlBy1T5?rNvAoBrq*IMxY9`t4@UcC&?<2D*5y|c7QO@`uG_wM#{)5JV(5^D`%
z=ATIgeM%%2p))Iovnud2f^Zj%Jpkf-1>N~S04>?MxR!#eIlO8abb;4)>T2T@IHbVw
zF)_U<zi6&z-r{`6isxHb%NJgDJ#GbaRmM)+%G=aTsDXul%bH@-|F~&_dX^AEgK3o#
zJ0XKNuekUR(u@XKoWmR-lJWwX9l~ir>NdjRz&Z1XXVYbC8i{dcx9Fz;lD{5&@Az@+
zdKb)o`wfl}f73qt3FOm1C;OPmG?8Ou;Ts}|XA<9`infVqoIT^CnVW?&`|BBFTpxJ4
ziIJNx;L^3tB7=bRX#*7&TDtOI5)0#K{oVBxLIrg65CNgv%fIe=5}C37Vt%PDLyH}}
z{&JscTuX7+ekm%>@a4BdIzbH=XaX>YCSWfpv4l)K1SEv(TR(KWO=ifIjH5Kv-@*Cz
z!+%*rPC0OtySz8+lYzzR2(HJ;tpgv{+_29CnqALni6$^kJB04*XFsEZWY|e@NQnXu
zs$Ut#PY`uN?gunpMk>-iVqpmf+aj_Vx=dD<<`d-Dk8+Fe<D-uJ#aaBV#8IK=(&MR9
zC^G78(^ltVyiO1Mmb~lMbQj<r**&75gePXveFAbvJ5SFTUBol={N_8lMK2;%qdqMo
zix+s_m^b5(?=j!zn|{+g8%f#-+Y<4kzLDW!Rr=9L3D&r!?QwKZBC&?4^WWPGvV2%#
zKf;AX?Q@Z9YivnUXD@E+ErD$!T6|406-t10x2GygQ3uDuT~2L0BM$eiJ3m~%y>%WT
zc0!3<Tw38Cn(DEh?<u-+QgVIQ&*unhG6ChFr=#-G*tKsd%ItJ(9o(mZsdxmQvIkNj
zuUM+MQ5%klDV{UX;*!Nn6o-6KE5S|i<VAJ5zjq?EwS==%Pa66j8xgES@&!bCLz++4
zGRfB4FI6mK<KqB^J~AI!S>g51(Z}@m2yjzj=e#BQw=zHT$cV=2+3s3#2`5A^>XcQ7
zP~8v@+#ObeF9fk&`3<x@{Pz+i=gSvHu~a3ypGKYQXPNMoAKZFJmYVTfg()0E&YNX9
zSKJltMGDbia&(@zXXgt(%^2XNN$U0cuP)D#4{`_VC8TuPHWHAs`ym#1gmM6#%-R{C
zU2v_1v`na|dxuExKagf%xF@1`7hVz5;VNSr_lGU%t3AZYIjtCtT3fK#=u`cl1E4*f
z6`f<expFdZ2Gv<sIEPXT3sv`-Jawfd_X!BkOr0EiCw8TE$nV`=UVa;y!UE#6-?%vM
zdi!%q7|t?b96_u-P+$&WxEx&8*Lp30X936g8RFz1WZWN4t~<SBh7Izvu!hOVrD>fV
z3-VOI7^rKfqHZ?cOs}a7ey?XQdY6-Dve;rERehxao)%&Q1-vfnGv)j*$IqWWwIe~Q
zu<xKQPF+JV8iaM5HU@ubd|zHzFoC(Q&}9|*ULUhr&=S4cALYvoY86<~L5hwP7{cd6
z-9=iE;En-W$DK{lVF@_L$p0zbU0<@_zHw)zpp=UjW5+KS=SX~YTyHwWOI2VH@E9<n
zSsv*>rcM2lo2$D~r@Rv3?6li4J4aYcTefYpHocg+w0I!4x7U5|(3$Mpe46NceKXG^
zWfa|<#^X(E6mD^S6C4Sl`Ocp|A`YTkzb+d-)8syq#d`JAp8=i`gd$Y^3o2y9A~cpf
z_0_n>MQIkpP7Om-k)1qyf73;E?KR5(IVh++H_Nq{o<gsIaB~44iSpm*v>Vde;%2`&
zwE7n~<#ad&PpkwI&m*#LGzYO?UHYEGX}APhBQwZCTwPn6P5E+1EedLcNkH6oJuv4T
zDyx}mM{%v}f%@s=4Y`|-t8VE^Dn5CCSU~+AU-d@0?RZVed8PluJf;%ct&+E2D{kg7
z(|pEI6MXyGPr_p-WpJ1)Vg5edJx+#?(Tdaym-Qv}-+m+{9OSY)4S1#=G=bXvYs;W?
z45Q{->xqaQs;O@Ipd>fL^;x{`@`YHX4~`!ub5&bQ_PVtQ`tb&Nlfoe{Zbl5RNVodC
zNh}R_0gbGr-o46!8EtV7f5+&Bw@qcqB=P76X9T$^ftT(N1g_^h^uFr8G6_n=fAEm_
zp|0c`W}YlvN1IoBbUVccjj@Q48_ZhMc1~_?0DNy}0FrwFWw#JoHD!YO2{u6_84|Rl
z2Fw2-z2J21-~HzmFB$qGi;O$o;p^DCQ;IXQcHWR45k1s;jn9MTq#3#2{2=a5x-Bm`
z++miX_K8pPna6YUn}gEb7@LAl<A!3*hYKdk*@l7*w>FoQ8p~7znwkafX$0OS_pPj@
zJlx#{^s65U5Ek053lwyb-<EYhJG2x%T|kE(B_g=&=Bs64vNa#xhG^Zc;~EwPKh_^s
z!6}}`PKLcbH7VWS1OHitig%pJC^?r{lR#OViBS|AP}{)GI?dSDnpEp1Km#jWqB^Xq
zf3K{_z{<ZT=u8inLLpFXnUChFL4GwQk{pOIx-g3xZh96zI}<nh5uilJI;Og#Tz@FI
zv4W;1F<3!cEL3GSluj5{B~m$Ma{*#BP^7nuG7LAm*}=b-hGL$CP|aKvgULOq2%7tJ
zxSw4#_X1b7ZjB^rpU-yZmDI*FrZF6RJYHtMjHcWaXo#AX<tzL9)}j8RPxqovi=mH9
zJnOB3j)0_vmGw743;OF%Q!nxuEEE3zTVBH-%^w;y@qCWvEIPt-9*oBQwX*dhOr`em
zzJ6unsf9INiJ@0B)+<87$cGA~M#?}@K`Ez1J*I2%>?-P+KH0BhLjGtRuC&lR$-CGu
z(GFq9EPW%_ln&PW3ezqe&&FJcJ)8+l=)wpfmz3jvbKbV@&|UG%%@=3;&xLAI3M!Lm
z=&V9`N741wNozeM1N3P;OT5nv$UJ3$Iryo;_}n!%76Z~*{=>tr9&*!6)|A5u`d<#3
zSakGue*L|@Ccv8<?-Va^_5u5z1<rNged6Jt3HZnn6qyD6c?s#4bT@HAKR=47{C>_$
zI?IFO=)(81q(g57V?}4_hNNB3^u#SYe#f{aDbSg=Lmkj2B|k{w-mK?;L40}$z|X*d
zA*ghE%5MA>MbLHg8_Ya9O>a0*AoK$fDWM#Tb`TU6m}IycbDcl+wUyLe`fGA^qqdKM
zB0a2-p&2;YHvp?CAd6V}=3wxY-`1CbvP1E!r|&`q4AL)kA<6_kd<XoXp!pOm;szqY
zI;{=1aTZ=-h_n3!;nQj0XZZkTE~~Ku^Br5gGMYFpG-s13fikCBHL1&Ut*q=vjbAOA
z0WL$#Zh#?!|0`L*Ea(vkMc@u22{OYv`$aT_<p4ng3arq;P%4kNAl>~G6VnBg8^pfc
zNQWFLWN-sCUhsmM@x_A}+~*&?e%%};k7%1y_+tjMUDfN^={aSNM~?y#o(m2f6a*ke
z)@h;EAEPH745$s%N%#>|KZG7|*oB<^z&>ArU|!FS;r<jMp`usoid`vBUR6%<dtcr1
zjClQGAPQfmOtc}2h~b=K`|rxS#G9Wh$5ZQP4_Fe)_oaNMgj3o?=2luFO7Bb=xuZF^
zs84K?5T!qJ;(Wz%`SjR;zkgqJXQ<@mv*h0%goE=w=fxt_cvQmi*hI9SPq9nDE$9rS
zvKNawqdz%~cv{QNLN^JGo80uOFVBYW#J3QqZln6HX&yRg;|#8?ts$B7AQ7*G2(~_|
z+bRkQW{{?J+;t1EC2J#fe%~0ziL=JtNI`|-+yz5p$F;aplLMPqu2@y>7sFc?#}eNq
zZiewRxamLr?sNL@zWwnnZX4q70iAz>LPO&ka;uEXlA($cOd4f&keEjW1Q)2Y+B0g)
zN=w<wZFMU%CGa1y%gTJ#N&3rdjy;Hkv%tmqF)b}8*quXf4d$smg8K!8zobb~7sLLQ
zKUJTdo6OmMAa(y<SjL@X15w{$2idu?{QyTp9RfScWZ_lUKV-=Ps15camatddgBpj|
zuU|vji$Bo3(<<Chj`^qa9ntagc2MZ;L);_;8MOmY)D3TO*k&P+_rKQXBV0voUy%=o
zpr|+%kNx;gIM`pn&|HKA%lOZ<jg5dyH~qh_;pR`tDXv^$ZAt^XSpR14Ata<%wbqW7
z7$hM(Fx<k1;afhuw2Hhh)(>~vK7=rrk-<bP%;dZudZ;|{R8$6Vum^+9Ey!mDedcjw
zRcbAuihw{thws%O3gq6vM_egeK<X~|W}S%;ZPww4%JHYHB)gUbSj9-t2Qsnzs_(|`
zv7N434Ck9m2fRavIkvF!_Tm(toT$VpFH=jB`ea4UYC8))8TUdkiJ|f8V21g|oGRWI
ztkrfKUj8ORpYz~T5pw?B4M*kVx;uH+qj@Gqd3kW007WcFqnfnV?!X}uEF#NSak46-
zWtY&~>x9MczD-?UuN8V;-{?NfxK2P8(miEScv(nlx7mMl@L!2erIDf^=IOt0G7Xn@
zq)|TAQ<FPRO$X~tD%ok~|Mq@DIB%}Bcu;+t{tjwGx~TUV+>(;tld0r5WZhVe5n5RE
zZHt~=EOp22W0}Q?^u~{R(7jpd$&pY}hhkefg~S+x8U-;vCkt6rwY&!tH2he26#O6k
z{7`@=MUtnyL#9k*223B&(so~&j<tAuf-98LDjoNGuiv#OI*GD)+K`Nbpr58uTWr1k
z-}74(*yOong(SR<9rXX`U;TZk^DlUarg&|OeSK?j(zR2+Jogad1dojc+_v>S%beZZ
z=pH`~g5)oFOfOqof!T+JP|2q6c<__2l8!FRi|M~ajcF8bP|F}401(CnCq&J)Pk~0h
zB#02;$1T71X?qjB_N$%?a3CgNt&>H)6ZgVC=`>Rh>8xg#`3*Mt{klPLNzis=y!yBM
zK}dT)6_@hErXNXKZ~2&QYocL48oBas0scXF|Jx(Tl79>7GUF`|t>7NA<?^N4YOG?w
zSUG_+XA&b7^q6ve);|%8E8>$Ht&pKfeG<R(WE$J*^v-+>`}|(j(&!<B`;-;RiEC<6
zUitg~9?48v%Y&J#5B5r^C28I?Q)<M4&F&;L&97VOsj_lSD-{G`|5&tRf<b{%v%r3H
zlEbs-4j6X9B=Z#RZuOVPGn5U2lp-FZb2GeP2mx_m2;qI@fsz&*eUa0OB>_)@i-H8v
z>U5n;s_!%`EVC^^M^GvZ&o{hdS^TQE8|*PP_sjDQ8hMoxxfekKa$FMzPL9KkiAaz#
zaKhF4{J&qIhdTfUnw|DdUkv(Tf^(ivFt2G&oyyMFSUd&W4lq*Ni2YZ)a2<8Afgo;|
z)vQ3JjN*3R(ON`p@MBB5vG-+KbF#su@Y1aYN;W-z3X;ky;$W_WGxRfqYFu^rtjHtp
zoB4BoglYPHS4*vxF4yw3SNtD}Hl2UX3ip2RwXgF%DtIu8Yx#7drTmvym_vlz<=Ih+
zdak;D+9mi29Asr4F^g?89dr2V{$nt!xs3dhc)VE2tAZ96a8wq5bs^&OwqD|w9BKnU
z469(^2cMz@ET<HX66<Xl?7;)yK#MEy{MLhDye#LXPJ#GSi&wLGr#*M^1`)J=IQuE|
zuFr`w{Rg~qx<nu;3}&E+LPD~R`M!aJS%^EwZYVSOGPnkE0J}<dHPrnuv(CGlwA)H~
z56>s#%k=rMd3^|Ln@xh&d!^bE&AIK~ZV_^vkeOGvLV*nl*=4$&_xCrO{nz6!t|OMZ
z#y=C|#R7a+KA`X6>A3`AIriuksEUQ^i%qtwV%=&+$K@PFrvG9)Uj~X`Wk~cMP10bj
zLJ=1*CYw3;{@ML9{#>b3tQAVO2uI~~jjckn#}T5mhLwhp05DDqBop~}p<QLzUj>y1
z7+13Hp6cX9X`g*c!5$d(FTy+hZ9l~DX%p{FeGZ4`<@D{W!>lqQ&s}A&=%SqKHmmC4
zG83ukT00fsd87KV0K$ZFpjhbp2)#Rt)AA3(&Qc&i>f9E7OI6kr{XS^7zS{ou*mE2W
zdxBL>&zW46%<a+NloCinSiW+lNfWX<uS_%Iv|`&taeaHBnt6KkS4enHXxv_ZjoUw$
zxc`y?HAZ?_z{A3JEYP<_c4M{OcC$#U&>MRL$!-ERKD4zN4vVOyB<t3nrsab*wtAlV
zvOqzU>oCLkW_fQ5c9M20tfAV!41%?6SK<ThH0%zHcuiQZcMPv!*pm<X2Qv0oV$(Hn
z#{^QwBF)4Nk_hJK(_mYfHqwUo8>Cc!kvHU)_7`PJL~IK?(`Mg~iMsm#=&16rhPp0_
zb6$4firu3jH987upsBhNO2vun519*J7@!$twiA)wU2aa+H!mWKzP}*kvX1CkH_esd
z>mz*t$Dx|}>N($V5_7iPyr)oXs|Anp3>@hG17f1W{ov*KNj0$4p%ee$6+b`M2EMZN
z7bE9D3|bWrAJ&8<OO|8N)FJtzMD(s}(|M2@>%n4G8}r=meN0#P6x0->rPTz3P0<Sf
zW(=#sfgC}qtwP=E0;$$@@Zt?a10Uyo9yw)Y*P^M?`44LNJ4?rcd?ag;zM?|myZMAp
zyc-YAUVF#(ke`^&=BNA;-e4ohXc&E6g(Bj*se?4XL&eSepD#hbMui-qerngr@$q@L
zBXCq>Q6PeieN*Vofhg%9rr|ki?uycj8=ArY*!+z+l+g79U;nNxDCGbJ-}yUsAgaM&
zpM(sN>FLzx<XsS=P7#}IJDmOQGPnxftekahpL!JhvfsFqy!SJALbM2__MLUOP#aoa
zyeQ4Nf3Fa4<pm1yE-sItYw{Ep>f+!s@8nfh`4fKq0tWah*pLixJeX|cgcBhW)E4i~
zszU;s(gqW2*!n2S<=&0>a+HRu;Qt2hc*X^w3WDJg#+9h}_;1MqPW+CYI*Ws*<<J4R
zxci`IU?38JN-!OW!^Givyvd+Gw*apvOhj9_H*K(A_5Qs#BtsIh)VR2G{$W3ST4|x)
zQpO$g>q%+*k5+Q7!CBnqH_F}^x92%pj!Q6iu6)*lI-<{wyY<R!(3W+fd@cuW9mtCc
z&=oT=E^;aO_VK}Jq4t0%Q60!jLdD0^CD6~wjbR4QgS{L29Dux<pNkfzqw6tL2CT)D
z5X*c0XywoUYws<is_ee@Q9?pe6r@2xX-TC^RFFobr9rwwx&=f)M5Lreq@<*~I|b?P
zZlvSPt?zr@-#O>Q|9m-L&iFs$8RL1z-Z%GN_gd?oYsNLNtBlz6wsU6UcJ()WXSW=V
zC4sd+dir6z&ph!*%SRnzrxK>+t&ODeBoDB*SMxoZ!y~Jzw<taND#+N)q@bU;{Xyx3
z;>Px%Dbr(bRu+2t^tPKrXd*j=mtGL0JaLP$8qTGJRY%mgjW$N|+S>~hxk$DuZ}FE@
zey*N6BlJ|qrrK&T#h?xiAX{8#P_gi4-GJIIP*hA%2?AHu2G=iOtbmsB(Z-@?{ky!p
zF;JvGBS?X5`6)KmG%soCyT*<U&cjYO+3ueWLIWWm$dP1yXCe}x$%~gDvt8k=e`wIL
z_7V0sk}F9H@S<<b6Cag?-Yl2-P@Wa21qBu6Vc=6as3}&@touX1Eyy|N2lha}^-bfa
z(l~t*hmAWaI~%WVjF*hGeo_9yN&y7i=ZE9ghtuk_v~;l8hezHJW<)#FY%HQ2{9MSC
zd3A_qEm&2j?Tc(9yh(_EDHK!IHcrB4NM&ByyBHee@h)T3*Gsq<uT)<{Zq$NXk(DCN
z>>RzsqP7TArrsWRMzONjH@FT!sWEHCpc^fYKmo%Gh-5-E5%Yov_p9-UtXlp@^TOS*
z1?sjUBQCSXQs=S?`*WZ6_~#<h^QK{38G6K*1G`sJ)gm)3Sbd0e20~<Op)FkS;$Q$0
zUIfsV1l<Af_}B>i5peMx-m=eC<&pBHDkT)}^$kOR;?0J0ve#e2q|Lr6X`+(MHGY`7
zGyUjeS$*iha_EzAa0OBz>7bOO5efW?L@s{8At6PzF%IAkwqbdu*`^^SB%lu@1{4k0
zw7S_Cg%M<pi=e)`!bsnvGAkgtReW+4&_aTNM8y8zSOCkNJwN-W<jzah@;4Se(mlAj
zYk3@J2iHMx%w#Yt@<KKlPNe3<3okx7>^_6<4W)&980cIz567m6_6!o@c#90vs%f8J
zVvqxDIuMTx^dSHr0>G|@pj-45Q+Mi2j#O+8Y2X9`wxU3zvu7$aT|M4Gz4g61KI!;^
zdu%3GH{yjcd&NafJ`;T0pZ<818>sq<G!QqRBs;j8y>*2^Xx&n6FVvQ#ewdm#7`vP}
z^yU}WKWo1Vjpc#l01~zavC}hq34$*DHB3gp8ib@Qm_74)cR?xOm(X0#snd?)1op$s
zskbubx1G9bB8po>X&>II2%Cqmcm<cwB&qj~ic`zeXD358Pa`i5bw1=SZ8~n}&%NU)
z9Lanmq3w3?s3K&|b(Di&_C5|oEU#Z9X!_mQUzOXXJ@IBezWNR45IFIDrPewq7l*q&
zE5zzW6ES5tnb(h|(KfAoUTl*EXDkF;q36}A)hK@Ghw~9AF%fWrjp<kQw2%FDys$iN
zZ1(r=w2l)g?$I$>;rZYFbz=cFteb=1V!@4*YK!rA+UrXz=WSwx_HFzMt<x5#1fT2e
zJJQ<ZWps~=To0{Ou2yxvciskPaQQ~Y<^w?=2n7Of9f9(M5&rM{=qt+Dv)%Ne0TX#w
z*j7-W6KrObCa+|s<S^|^_Ic7S;W^p^uTXT5Jm>Z>TPEf2m|BrjE}y{=+9QeHjin%b
zvi`fSNb*xBY>B63q>RL2yN?_49{%t+k)!uFPwq_+Fy2;8@N%#O8+&*&)pwq)%FqQI
zA=a}qovW~`z0>7w&ZA`8g&(q^bX5dA?@~<}akNi`L$mxp8Faqh_BSZX8GXB|rIR|e
zfKisZ_2Awcuj;1d2ScHR$*|MAA^oxuSfvAkwEzjRxH@nFVjX=014_JYO#q(YAr#I4
z6*gkQ%@4;mVCgpl$-whLw1Q*bW~qb4fk5#iAk_zFcNa&sKS5@|OORUljG%@DD^J#M
zz(puB+qE78zy9lVilim2#ibLjMF%s%pIwK${3S$S0TEh+07QCZz<TT(E|RLM-Hf*K
zU_V3|-CzqX35oafwS5U+FG>>^HkOn07yDH%Z>CULM#eeI%Y3ZwXtup_qTQfNH9p9k
zv~fr!A654s`>yx4>-gCD3|kM17*fUFjf=eZMX!wMLN0oa$GVB8E=9`j=S^pIqo3iE
zV;8er5Osd=XfF8(K^N#42SQ$(hAkNoF^9xG=fuB&=aRX`C#vl8QHQnOM$#QFq6$Gg
zKaqf5jQX+Foi&=&R}yl!t41h3EIOXp?7tNW<+`n@E;ux@we?yjr*Cdc(!J9JBp!h=
z$i5&@ZJ!Nzw9-xsZW|SjB6zI8N1GqFownniT#=etTzeMq2o?MG>7%lU=sE4W>4BtI
z>5g-Tr;=GNN(PN}#Y#t<Mwg}O7U_*etV;VC9c!1_m<L%sU*^q8s}HGJ_^JrK4^v9_
zq|T*_uA5rA-fyrg$AhdEK%DUI>(`#|1$zvY+)-5}{_=FGp9Kq#DqwIu#Q&w3IVo2^
zb~G7w`hwdU$I4jB#yk@wa5td$9M9(za%F6Xo_Mokje{{R=w)f#6TdL0)Eyb0R+iVj
zt-0~(h{utik}4K=RE1<5LoB-?k<g&U<=GUr=e#KG$I>j>?c|>i7PUTh8yO(nygo=~
z=m!!#@$>cF@HxFd-SRD9i0yKqE6aap=%qKBxI1!A=h5Q<Qjg?3ho*a8VvIG<jH~dR
z>|*qKj0+d{2IGxmD^KmkvUp~;!&qrXYFR78?(I%2+a8$xp|`WoIWMzfIeOzBdl%GJ
zA`>8>@NBNX^~(7)m*ZJt;X`@?gqXtX&f9?s)o}O(;$PE^-&f>S;;|Un>3{LIDgtvj
zexhexG9t#SjUf$t=ZBcgpHmjTFEZZ7L_M$5!p%Ehiv=#bIFAN8WDLfqRZl-ULahx)
zDS~_xFk<Il-2s)+UQ~R+I(Qde*M`N#y$a`9{e%F3DP^mKLhsGB?7<G`gAVC7ng{VS
zlo?>h23_<dsOW)AT00Q=kvuf{SS_AkH7;!jnTk@QhT7=IXAEkIM@ui7v8_Vi`q%nB
z8s;5~uGLlcX8UIRSlnkn9jj->aqai&PoeP6tmob(r1ZRrW0owR`@hjYoyK0jL>2J#
zJ>AsnwM!MGO;eynQ^ilEA`7W7eT5g#R^d%~JJ=r}%x;yLVA<mM<CV!4@%mma*F{g0
z>A1Ny7^jAysd!G~iRAe%uHRYI3=dokoc6t{*3#58nJ6o0#5Mp%_v=_p;{_WxQ6JXL
z=Is2c77ttAiRwFZIo4+{s%T}{j@hH>^6V|kiwi8Un+@Rm7~!$HgDe^;nkt@N_mv6K
z2zHoVuZ7x|?c}Q6;zzNUcQy~(S@_FH$hle6Hm;@y@~J)6r|W4&gyiI`@Pr!O*o%it
z)htz9yEnQkZmpkuA0PgmqY>ydbm5nTPiDD4_O@s#@;KUr+I96w6-p5wn!KVmV2)~`
zx1nI0$7is<6$6iw`!e1vkj(bz<BRsU>GbZGtV(PK5k$staDZ-aJ-AMr)8jB9(nob8
z8u^=q-<B^Z)-dW~s;HLp*!&|gLMxWbZp6gkbnaE4J^F!%FY#oZ{j5~~4dZry{-_0H
zNE@m6qgmAF`mXzV-+5<+``{HhGuBY96Qe?&I!fxzko%^A9$1^frsT>}L<JwEpk)Fw
zYsFp)4DL<e<ZNzf{N9#vxacX$m~O8jOMy3M;_5OimRaPRxac)&k&1WojAta5vczMh
zWl4*1HS{dJ?~g8v-Y4;#$6c|WVK5SZm|fj!9QGQQ6&=*Tn`ba5{s6%-U?^T1w(t4U
zxA}!%Ksr9;C^TN9iE2MpT~w<I(j|*cDpVS$!j-+`)SuiZ1SNah8=Cd&%{rSq2mLf(
zvC5vW|16j`UP5=KMO%S9@Gg}0$9liuUB9jenhS^OlVHI6icAOed*86JKrZHNt1=9L
z1ykD6u0Ul06+jLIbOGq*VT<2@3{qk5b8Daz1Q%v3+U-H_VQ2XZ6bH}B3U@13akZWY
zJFnPCl9A0$o{TT~98jI+W=7C;q?VDInBTTIBk(*!?pF-ar5GhT%<&Duj{b1|vP&mF
zID6OiRg8}1%1F+mu&E0y@OI&BsuinAa$m&io9^}R-T#2Bl8u5tW!m-7dh|M9v%5W-
zvx^Jpcy8W^dR1hiS8;98j+H#QKE{XY$XM-?dOmWpwRdG?cTUrJnY%7(7-M|yxbtV(
z!iC+wQnrt3u-v)*NBeOZExGzg%&>1%dTSrHPu`pv9>!Pxu4wo5UzsuS6fmU-pdNR`
zX&v~$bsjxzLfn6w+KKu$6?+e^sq(&2z~nh6rCBMPU4JNw!wIyulRcm__Ux%DFC1~|
zgBY!CusMbWiI#Uc`by%FZ*8qciQ<R%kSd1yO$Y-2u;wLVaMX`$!@0SK&v2|Se2;AH
zv_7sg<uTIlUB9%mkm1f#WN{m-SLtn~3Ixa%>~Y;K5L_U%vU)v4qT@reXV248rLGM<
zR3Qj4HPHM(={=mpeu`OhaMz(Gb%DM%<h@2ve)2gjoxL%u(NtHng}@Fr-;lI~42@Yy
z@-yuiE0sJ{u4u11d})bW3&brRl!3di=OT_>Pa>jRJ_wAR`!B|Q|715yLbWwDF}2n^
zF)-V#Z(N)(+JYBSTYb)+K>Y(d7aM72X-?`^Eq@a__Z_-?*n!lq`?r$#l3sr%Ta@QV
zaxN<)Si99Wb@{B9wETw@MY8Nsg5>k($-uOV0IDPSX`v=_7Q}|gVNYLOP(y>Gb!Or^
z9AChcz#IQA(rYiJSzixm8^CYojpri+rEB1Z%GbzQoYEdEe4A&?nE7EinE4U*50uh}
zzC*a7(vzY)SwxQ$FsRD;{aAdmF&?g?hPu@qqUGCud9l4Qr*cOpa%Y!KBurg1q+%JJ
zG9;~5C|hj0rMisv=eF_*Rfk1~#qqG=yg4y;e66!7T?IR}wTUJ*4XtzG8Rf*}cOl3O
zy|!XjHe!^}b*?JF(hGIHL5LhW{$x=1c$dqdS<L%rtL`$GIo<)u9d$42V$1Ne+ixo=
z?=b(jmG7tim|_0GT|60CJ~rRbQS&%ZrL(+>GOhd`@w)qGpt1McX8AdN+xqV2&`1eY
z<QqD)AEM`VzDg&^J@YN(4TQ4_c!$y@Q(rom3I&JYj8vd<-)kHTwK;D}+rf!$;38qh
zy=3pqh|CRvj8d~VGU{^++knZ|_4>?*2D1^$fg#QEJ;waP$ulx-wGs8hlg<|-5&Ay_
zNoK9DtsgqZ4|L{*K+uk;9fHtoReupa@lSi*?GIJW@+wdE^70vel_6aVB<ZxjDj}OX
z$$l|6c#&w%ec6f7d4NRP!{8Osv*h9S4msoT`az4a!k7i2;8EKZ)kDhZrVSn8di?^0
zJQbW5j^p{8b!_Kn5(Yk>e}@!0yBvQiTcZ`6s46R*)G1HD2!3Xiz^n)^^2%T~akZ<~
zm)%;Z)<F%=f`lgtmw70E_khY%ezf)bXOH}K<FgAC!EBqm0txXbw`kSN`QkfRT#F7|
zjZP^Kp{em6BV+rN0wg*{4{rsV4gHW*Q;VrC{v8R;)108m7&s2xXU47vD?aJz^sQmk
zY>KR)BaTgywAWEtJ!26+XtoGw(qbJEkJE>b%rt9S*Z=4`ax5>NIM?NbhDxso+#qn5
z)sfC?>u$_&w3@h8^1yrg`(xLRNNS9kS=7)<@<)Bcq4%@d{EcngUmhI#2F$jY1;~<#
zreb2|T3w+{y?KfMh2NCXdVn>NyQN^CcwX_QdCFx6tCa#RO2d~Y&LO|5Q^h?pbROqS
zbm%URCRhj+CSZ>en^29!-<ZBJMDa?xza96R5qXlLy&YR5!8Pa8Td-#`)?Qc$mWNL+
zzj-;Bilsc6eUp6R)0nbV$eX0d@Mz*D@h1^2eP1(wow*zd8z6r+pr?TI|MzNN!Gqsb
zRN7m9?#Zu&OsH@Kdg}_5MqE1>rg>wRjx~_Qg9VVNo!()+4M5JniYI9Dm<W~jL(0%9
zDEz5vXzO$Oq9tuw!2sQ`w?tan-dY27JxZ(M%8kuip2%dlaAEn!J;90UpdRK~=NoNp
zQm@B-l<zI0ttA8wj}gmAc9s#c=aj2#S)5TEIXJIop)Yr4Lb4b#-Kmq|qWL5VFMitD
zt9foX-NgT4G(P%dqsHd*lgxCA2K$=eQPPD^!PrE#pHByM`9&)$Dma1QvgUkG8qfk1
zMoz;yuOuM$>ik_Nz|0&BJzh4ubI+hLPcb{Yq0q#iSq`z(oBr}MHbYX<!=J*^Ukqp}
zXhap%bp-9_pL{G3(~A>-7QLTqF(H_ILV06w0P`35ckL67SmIE-59=K%2G6YRZG9?I
z`pO?okm3JX=k*}nvCAGn5f-sH@DwhdHhJP?x1Ht{wjEF{`C*Y1shrQoBHC7+XP_)c
zotWg(epX16OXuEI@by>pQKy%TvBv7oJI1%ESWbo9M^CuWpB_3b?R+C4W@}gplX99K
zU>uh8?kTlllR4h(IO>d!!j)_dk=6@qZm6?kCD^uL&F#x4{f+5gN2V}H?WuK*nDqgV
zTu0_7YMme#m*blQ$>fhvXfQK0u&H~q&TE_)M-Msiv2DXkcQ+;tGPNHqeECrPP*8KC
z>|;U?ecY`)XC30sv`-(%AIFC6-X6Xx`O(>(t5N(Oq{#O+3VZv1zLq}Qg#J+qkuB##
zRj9Pz4o0)K-nO%5YXb={$92}6s%RZlkLp)K9hmJw>w1H7BaRacNwVE<L(f97Ma70Y
zsFU~0T|_5PZ@7tUXqairZF4sKM8$SKB}=<A9QoMkaqBDHE<&{{fp6c*)9}vl>W@uZ
zn!9l|xP9=M^eUe_$zujRWNqflzaAy;Eh|zVZrL(B($G;`6Y<ieoUcavnk(8(?P6Bk
z3-UhRWRC7^9$GqHbT4@?RwJDS_aPTnL#-LHL$vwns=QuftvlRhudlli81Ef&t56Dh
zrK}`HL=!*#?ZeIUq*SiOCt<xKv-7vuS=r-iq?Jh|{=vX53tG|i+m}3@$2?>1<G-*R
zb%|LL%)5B4&o=upwwOHo3h*Z;kmm$M|E#ax3cG8%j51^q*L_NPH9rz7?eXOa3i-|9
z3}Tr4Usc)?Rf6t$vkcwl1UKqejsweMA|CINrV6y~R<VjJPgbwKJi1L~*5|)fY+$?S
z8uyE7RGIY?i&@8KMqn7+2M{vx^<qf+LAU;RiuMT*{Q%36L4jIF^+niM^^r8~-dMhT
z&#;2{qfa^c^xIEYDpa-CzxCVNa>+4;bgf%_w(VKzDlonuKW4h{vr7rn!p5>yoz78g
zrh7RwPd=^gfLiZT(I5-w4n0OmRIajN+RIWj!DFA@;z#6fas#H98e7WqMXXhFBv0IG
z_y-KSsqWHcpoJrcKRxC7`rfUb{=%g~g}!f{Eb`a8ggpFj>EyGc-M9^n$&P{!CK&ZT
z!>$gCzbm-A+67WNTPM}j(#(ZSh>bMa+6HcAX-QzmU6hbr?iakkyFH3BCqO2YOy0`t
zsN7(F*LJY=$MmiU*p>2fTWhQL8_zUTwclRjt{Qs69Q`iqQEZ7d+WP)&!yUUnGc~T8
zF3FGSUHzM99Vff>BYPg<_CcxjE-0;mBg@YMoK+CcK|F=<brCxCC;Q8Q6vpwoSl3DT
zc@fq&fIB8m7OlGVmX<N!f_%b7Xb>~Ht`q;uAqOirdHfWvF+M5R&~R}wlJnPETRuCA
z+bn@s5B#+kP4r)>msLDba+~kJ+O|0)u{b5rnDWe#SW5QM@gO@MtQ~5+5pj7)czh~0
zA2953X!o10OYH@F#rI!nGvZ(H%?5DYv&A&sb}9>m-%Sng5^b3sseTbJYHi$f71(&e
zsTzzut8Av&*qMYu<}lZ@!>#0aRI`VgXFvQ<)EX2&fii8Y_Ue*m9P&dUP>Y66c7usB
z+k@@tB&e_<oO?1JD~nnkm(r_W{^%sq?$ahLw=vC3G_4p?9_g)A@9YTOe{-ii_hi`j
z&@b-~!vebPJ}x_l!q<jQ+i3=0v&Yc_@F~;N!gq(mu9!qifvk--C#aO8f#wv+<HhC@
z`g4oez0igG{Q2`cckiZ%l>}=>vqX`b4@o|Kn)}MDXb0zWUi5QNAWxUmN-}PE4+=x?
zK0k{^IRIxDxS0@HotYVF<0Q}q4FCLjp{O3frSIM;aeqfW0BIi3B>oO%RA7y5Oj)+C
z^}Ki&9Bj~=C=~624#^sV@b&0bvK1)p1_MzlhjYH|SNZzIe8~ID0<s=K<@@rxjO`Qy
zbO=3t%5;5zs}G_ZgPA&?`&KYzXCPmrU5<VqQ2jKD>aloz02sN?8i)(&TQSJ3{pU9z
z88HJv6|3zTVEk7ASoA+%=K4?p)&$;71Z#){wFhDLAZWtggjB!5-a>m9r079ZNTo^b
z?-+9V29WuICK;eloQ7>TDx9|QMBxxa0?q1se0;G0t4+sKL=5)ZPlf@hx6t@zczJ#j
zFW~e6s0#qR^bF8b02k<?N<vK1^EAcWw5p5q^PT;D`yn;N-#ulj=0sZoXcK~n1bU%9
z44OSS6oyV;w&A*e-_s6g@d7)mB&6SB0AQ}Ju1>^3_jfL(vSqY*p^N7^H1{*<3nHe4
zCi2TF=?XZd1JLmBo+S9+uNw1<0($HdOoM9EfBmY^0J1AE9)jozWJ}Zu<`Fj}$ows(
zjC#>@OMax%5?CogV%`CYl`A6!w2&03@6K5na0kdl4`t@x+um2mLaK&(BBU=jhF=w%
zp^2KAF~Kxu$Atge5@KmseL&stW5pJ6{X0!!nPJf|N`WE9qrbmaP8G={`YK1|?Ec-m
zJ^;rL9LLfI)e`6P-BxDTiN8-X{4`UFx2n3jy{D%{$H~d5cH_eBVC6;4i+?|$2tNSe
z5_%pUVsGe`=K>K3P0ja7Rk`YgbP&}!Okw`pNa2cTE09+0yZi$RxUHZ~!ek-(Z?<Ar
zBQTe{VqX%>gD5otK{;y-Uf<tuybQ7a{SO_3=-*k4Nk)2r_(tCsrReYDT!%mya9N<k
z+M#an_kl7GItY^ybdVs<2ZaTYQyI4ogRVYsc}@LpAuh?Zkil3%H1|_d%7RE1)K&r-
z-@gR_0xfjE19F?!@#DX7J<C*~VOfE`1${6L02(7f0)^GHl3%G317DwE12Cfp_H{gu
z6ym1gQVj<<2aqlx;kD5X8iz!ICqScBb07bkB$e-|MIgk)9hb8}?_hfuDsP&80K<d9
z{D8p#b>j|TSfk54;OK%V??1kT#UkPG1sYp`nAStD_xnM!`;3f1P+#Ey%s+w-5D=%#
z`ak}TcsPC)i4v&Opu|NAtfjC6F`%O#&l~u*KgGw(Jbn81B?cqnN=0>;Vx<4|i}KxH
z2?Wd$67im%o)4|3;^V9;f&i~+4_H~_FnPqg@lBC~GFd)RP!h11@Xo#|yV$DSrUQTr
zbUa@E)Z+j9{^S~iVG`J`dIBLU(D^x*2|>F9uoZAX-x{!AtrQ{uwQx#w-_4WLW#bSb
z{m$;Ltvd}551<lyfyQoiteE*p0FLe0oqxLt8_W_Q)0;axvzs;{<)#n7n?<tEh)wbQ
z^=p37qqV_c0F*_Q34w)Bfy$S;!oNk0T)p`Lv4T)=P4qZaN5`Xh4jq2GJ3BJ~<vc&0
z55u7rjt8!N<^M1~6`mKTP>Hi0)^gf~@QK%Ae{T=5CmgmWMom!(|E{A<N6|+qz@EW4
z41n1L!4L#W>%g}AE-0wJv9WQ~;NNu*VS6G&JQ+&%=quRh9*VHPV@^}~j#Le}hs(3+
ztDok>tUpxD3yxw%ygbX%>He-5DVSQo&A$da{5wx42`Xce7Kp!Wqkg}NBxLrR;Wmq!
zD4=iu!v+cWf!GkD9zw)%8!^Uksa^-~4<H8wxUB#Dhzd+T0`kGkoK4dL4H>XMGGlXn
zE`j{uxs1#`u)I*deo^h}2=DdMj&;pR3*&!3G;ChzY7Yk(GZLUDTEGL27iReTTOAz^
z-rD7Mub+iHL2QAA3f&5HjlcgF_CtxQFHCR`6Uy8`Q1k2ZPbB=OVB){uPAQ_~QtN3e
z{BE?`zr#LBWmNWNRXh4YW`el$39Y>pZV}`V0a6W6kP%XO`oBXqjbua^Heh7FQHR+p
z<a&?{s|IZlh^ika7gqxR^MB)NaPUs#v!_GUQo#sdAoM{HD2EVIt^g>^zsS*egg}9(
z<0~Ns`;C)*65}+%X-W^;fS`#B8QRxG6iThMsJzZ?>(qSMp-8FM7{KTB4KW6qNN+Iw
z04)Yp71P?wNvG*uz)X(jsY0?GWPt>X9sXSv0<Y;m46ZKAVztw@HWVjcPkRFMvzy0X
zyVe%k;@~a!FC~OfIsv%t-0W;URPRwlrKGTWg^nm7CzLmU!GyS>B56#nkLO-HJ+zh1
z<yLMrG&lPLL;w8N7M>`iEfAD0VBO@XGyd+h*Z+@j5DG&1*04<|p_8q(Iql%qI0XI6
z*?BekaPWfxMF84=rOQ*pu0=wu8MLyXVc!N--(Ej_;Wn_3W?t?trT+}H9>3@^yd6w-
zuN)kTGMC$WdX{316ZPPfL#QnSw*m)jRRF23j26-(As9hGrW$cWKt_O-3IH-{XpWK_
zWko!fe!xpGz}YZ!K;<h1%!Lp~0ay+caywf>$&q=zE(HNprPr7I80ZT@W2h0<rV9`~
z2|u+EYuEI{oGckyuG+xAFsw}s53h&TY18qN2fEibV3P!Lh6DBnTB!Wtv_E=-*c=3f
z9hwcZ7C_-n4HYKHh9JO+sNx<DU?C{CmLUfs3iX!q(zD~^5$jTooT|&y^~3eEqwRE9
zK)KT`#IG)hzc58PfEtrLcyDknE)S~eEM6Q<W7Ktb+rHzD>z;LWDJ8D*01pMPln7@8
z@t`ycK`4Nis0XhOQFw$?Gy=Hm0rd#$;8dpbe~BOg0uE`Y(L5-jMnhjkNBeXdVp%n}
zJBX481_r>o-e+LQkO>1FKstK4_?aO3#Qf=~xVT`r|Kwx>oFt(C6FO)Hpm`#1kVL#e
zNg3L#&v<m`3(E^l8YMbM`}?$TK*>x;BW^rV*5Pn-_CMWD-7~+CBenm=0&Ie?64=Jb
zQx^>=yn<!na|aZB{8L`Q62VFZ0}%}@G4Pn}!6X;}uG?vg0ys7itA?~V^>3Y{R*^hF
z0Sah^O@C()09ApF+QT%RTF<8DW}q8RR1{K$yqEKzLDdtm&*hMaKk3xD6a)1MK<;H+
z2*VpZ5B;DB=N&k_VgV~00{cV2vJx@=cckB?d8b_0AHYYgQ0hhANFzW3Xw!E<10hU3
zq&_~coj~~j>@=foWgw8IUB3<<nNr)oEx{$zD=H6-raq_`QDqm<v|<b#d;fXTRF;=V
zj$q-{H8($_pGKVaZMdBd-c%r-3G{=GLE}K;AqK_%jz|BeUBo?jM?eecGh?tRf~PB~
zfb0XHA3~%GQJ4k-NFd@UrD^&5ZrgQ(-3KmO1Xn3k)%vR4oRH?`=8#NA^6x_O4>sKY
zxGS-@6*Ix<0>5<`rh`Jz`6GbjfMx^p`gPmGW5nis$`omr^K@l-IWQ?Hsj%hm-cwu!
z!^x&u_7TC$0jJAN0XqO6)oP;tbCyD6oMuS`-mri{)deeJXk-K)werT_Q+Pq#&=nog
zNNog$0^EbD|2aMqeE{F^5KLjtyMn*Vpyc`;)d3Dh=zx-b7V`Jczx6Y}ye<I;<_Xx}
z^40HgK)S+m=m}B5B2_%%kDk9a;1{J9a{UYkABW9MUFTHZw=IglqaNlOKt63I)>-hu
zpqg1tPx=nIT|jcbC;7W+pQjn}i$bij4dKPtrheSNmY*EqXM*kl5Xt@ck3k_8!x@Ye
zn5x`&Addt8D4D^YL3Gr9gUTuAYb<bM@3FAp5J!LII|nqmZ3ognC*&6R?=UA#t*xjw
zfH3&;7I9vMix9%c&=)|&f%o@!*L_$2om}A}KNY)0MF4vYBsJ{7nSC~_%4s{-WQQ+@
zSetJ@;hdWSnPs7e#OMEYRDM@NuF#z<G)qSO;6hlJ1MUI1DG7Ge|L*)E6QyX-tXK!Q
zo!Z*lZOd|I?%%)f4+Yt(udRO_Wjoo=S#KN1$H(;+A3Hf6bxzeH;9caAh^KshUpWg&
z2(p59_V&mmg3j!~IM49r&tq?Zfb<s_NGyjc{2kU!7?vjx<8106B#v=NIlncaBlOc~
z6k}tBgKVHD`G1~Nh32o9!|@pT?*lYlL;HJX6Wk*F`vgyA6uc30MDV{a_<uVODG)Y+
zqrL-OUjh?R5fP1mZGjl$4nkE&y+Ggp<$oQ#vu|=RKC*#ZjnE4QK~sT=OK?>Yhcs;B
zHsDSKa0a3_NKL){dGqfpWTf<kbOq5A3RJCu<2^|ty#OUg6f*(133a9&aCHz0Tj>_W
ze^)G9@!AUX&)^0(H-IJIwmE|y5HOVp-`4Y$ni>g&#tml0QEvaNuKIA*TUzJ=3KO}E
z0n&?j@Gto+1U%1q0fFXN)(zdW-oR>UXMME0TQHp%7uN}4*v2rA^^^=a^BpIXe+_Bq
z(ie;s;G267N)r%(22XA+E%82$>t6c3c><f3RN&2zM$3O8qT&c-t|0D390WyR3r7AW
zU_B>hOQ0tTjyCA_hvL}B;9%?|kHgrDQz(g(^4S^k)d(P-uAD$?00{x17|PR;hlB}|
z!R-;F!e1abglGo@4qb$Lib^&S1b|2gzG%=oz8#!4F^+9#YnuX(^w$fZr=^`YJ10TB
zGz<K8vhp+(a0mbcn|@0ODMG;qv1ITayFm*MK=1s3WDvl7?w~^sKqLs+5GX&y^oy4N
z*OEGRn<pcFpuV}8V*nqzU?5%d3)E1d``KlqpcR2v7%2JYc7Y*q074g_nG0{U1n~ky
zO5x*2g$;v_%74C4h}18Pe>TN6<^LW-|1X2=|9t_%UcS5j|MPhmFQ)STocL(C<r~pX
zV_yDZzU`|AUduwY-RQzpeZ|0lgPgseSVq|2`|9L1g+-#J<`J&l9-hU_OUkdVJ`ehr
zpuFx&>83a5pjOjUMx{A<a%Mt^j6aFS^T+1WKDy4Ru>QUEY-_LHu-SaenqJQYYMG0e
zgS;KJJ7vad({fGK?=Gb&t$nKb5}6r`C3W9wKO3iAo3Z`ze1a9_1my4wm+#$Na&A*1
z6^`V3LxLk!-DOlO3s+{jtbGr$PLJ%j_O&Ig-r&iM;_N#$rtHsbVO=KnjQim3I0nki
zswG)0RV_d0&XuS1+r4^w<zYN=L#gvbAd?8)NeD+4UEr#mZT`rk^H{Ha{NAP}lIX`I
zE*<K6ecQX7KZsB(tMDE=t=SVB7LJne9Fo_ZuGkM*jbe(%3`~EMp3$#A%*0l9&$c=8
zIAXV0FjvY{TwL2Y!y-)1y9NCzwIPYpYmCSG48L=|anNm8DQmQlkE+Wf2QSzZz}USX
zeSr+*%S+Q<e9u#@*bSM=&XnrpENr$Ievo2Z@kczLmZW9YrGIu&;XoR4dM%L>B{dOM
zRdsp0$;?0WmEJ+E8Wn})gwRDoZJ7Q&PhN_ld-X{zk4*^GSbC-<Tb`*MTFc?B*A-|>
zXDVl-Ole{RGe`OnvB#=XD(X7Mxs+qx`;*=`$8&m9+DCu_bcrGLGuF*~mo!b?c6@z5
zh8vrT8s$$T^HMK|0J%zzfB2PmPj%@)neFF&C&hPzX0Ly7>4^Q&JkA%m=BnEDp~tB1
zclVFeeO1G}OLCy#ASz5kLD<VR`->DW6hS@|LFO5^!cpDrH^+OpM#hPEW+z|tOiOSW
zb)WdK8HdjGI(gQ&-QCYui}$J%Y5f9ap?zob{)D2k{a5@Brke*INP<Vp#v`T9w9_qu
zvW?t#)yhuul+(F-_jiAeI#>-Prq$*)wWibf`tmW&mdPQ=1|Y`SsZV~7-HYTowPWbC
zGI1~XXv!sf?u}AiHm%P07d*@lwz86UaCDJt6;@8M&P%MALF_s+nNmx?aMXD=)_69j
zw*J8bW;#K?-74GeS!%Z3mTB!f0s@u#hc6NQ2kJ)y6vyK%8I#}i`?0)LiB@(AOU9Fy
z+1RQzIL;Ur2Ev5X9Qz(rx$fs<!61sWe=ilvC4F?^_5;*7)_5;<J#>_w466>1?xH+=
zmr*7n;pDV3a+}KKMxy6h`1?`GyO55*+f0jn*`GDnz~x;_e6J~-Y^^=opeC#Ag7D)`
zDRDloC!W>14#zBMh)BTci=81JmP@Of&}5Sn0~aJF_L!$?>;7)sH>LWOA_ljJ%N8ho
zSRO%);oh2b6|YM{<nR;Du!1$O;A88`Xj7#J2VdSAdyDp(RS{g=Nf4wtS!fK)Z70Ec
z>0KjtujlxK+dP+Txs+Gsg>?ak@la9r&pbZ+Tc#89YoC?N%EYQ!W4jUsvKx+yfvlA?
z`$SA=nW}{Og$b@*qgi--yyk4>7p`qD!(tE2i3Deo6HW`t0O03tK-}djp>d`SBd?=|
zMa?<mmtR*b_j1<8w)h_)jJ`!}zGx;Z)m-{kHqMhCQdJE{e>Q(mx0UC3`ft-G{W(09
zk^FHy><NIN_phZ>43e;>iIEbg3>`Nbfft9#T@@BJ<wb&u9NQ>^Wa`-N!cv*tLK{K4
z$WeT35DxCTZ=5umEK=-&^Tk$If;Jx=1>)^B?u}1ARc1NM)J9Y6NvLHylalYOn`fd5
z-OVzZUsyFDs!1v|?$zJTQ<*(uY+F)|{QUVd)^u+WmvMW}ZOz%U0{bJHR#(Thz1~Ct
z>Et8jtZh`^D!=D^TD!G{<fD+I$<(;NoVD4j@57`#Q3gO#|LygzQNraR0iB@QVPzNG
z_U>*p)zbF@oT@nH#cT2W_TL+h><>58vs`qVhb8KIdCJ5*t&Qc=ZfrO@PPeksk93B8
zE{vT2qlkR!s+V2X_^G-43x;W~>*n&{#ZvBEPu&~kBd207YgKD6o*nt!7DvBJzA>u>
z;w$qV=}L|Ck2;%2#qY<HHort#5~gz|Fn^7J?EPhJwLn;KmMb;yWh|N@(Tx|6o0JH-
zUcTtCX~%the_#UXezcu*<3vsQ!YA`Dk<yAr7aCm?V@1}lHKy%KAHKBjB(g;t#Xu*C
zSv+cejQniEh_-ip*n_XDoa?B|+-M7gAD4wCK2f2MeO;&3q*w@#yC-xf4x<kwRgK#}
z^<SR_hFoZ1#7C;u?p+(qh5!;9+RSl)YK7Z!Y(nkm@i?{kb;kTX)tElqERCABfaX=l
z{AJYVBdDy}eHr8r-(SPlD%I0s9;%BLiK<~c=5RKDZ+U9KI$evlo*p&HZI5wHW68|U
zyp5?%e{XB}Sa=r0tYXm6eTVrVdwO_i^3zswGLlM8;AUlgHxUj0<0H&$BNDZt0z#6P
zpRdkjb$0UgHQuF>;O1YVjuc`2HhbrJo#yP-MOV`OZ|1o>qeG16p6;|;tW!@*D4vB4
z%`1#avhH2*Mmom&dwI2=f2>z>y8Xzx5PP5Gs#U5+YUJ5#3F4NAYxQL<y(vGCwo|%M
zSHS%RJ0k(biVh2%5(`pzd=L{ROOR^?zH(bTJDPj<npigRZ$(`1FWyOdWW!4!?r`pO
ztxu%azca9T^M~fh*<C`DyS!s17o5R`LCLj@M@LJw%rw<_IHj4l=Qi{vYhuEuGSVki
zDJ8pVkm}k)B0`QG3HhHO_wp3=_y_f0sOJ;nQ4*9Em*<;N-n-XO7Hu)kR>&Ts+assI
z{Mfq4(YnicwP^127gFbUzcJ*BgF>yOt~!&4>_#r^>SN|e3u_PE&j-BP22DwBaF4fC
zS31$jd|P<UVRQECJlyV6yX@H{W^yC9{A-;KNhD6}BAJOJ0x5rA+t3_Miwofq_mC|}
z-5sBvMvZ>|Wr9eQfa9E@!~{?lMBEB*Hiz_CW0RpT`mLEvQX1#W(zbFs-(~UeojLK9
zvzoH2t~cX{ofbWs-WktFYD%2d^%Lx@ZV&E1QZr!ER{8FeO?(i0zhh2W0VDU~nf@W#
zSMsOfoEuE^7*ws3zY;Rtz6WWlw6nf8d_Qv%q^owt;ZBAx7pc3f-8MX0;bgSaEi$u=
zr?oa(LBAA{!05Z&n)FsyVm9u{%mx<obiQLRm741z(a)k`lYufe@@2xGu4|7zOA(FZ
zdFid^PI3Ve!4<LOpGOxbsSFA=TVJ{TJs9b)=NvM)`0>~Yuf0SvqW(d`@ugcL88r8r
zAvPbWbYg=_D}uZPvU!A9Xdto(b2GQ1CWsmXK?T(B>vh=$Z7Q>x%K1`%wZYc0KKFoL
z<4T6#9z1(k$tUK0XV)>!ap?0J!UA=5U#<Lk%Nv)bP`B#TbeiXdtFuvJnY0}xOs&!^
z>c^fjd9%N_v0Q6sgsU$(LTx$7=f=q6Zx%7{(N$5#$Fo|U7<Dkh*~FsX^wC}G(ej@5
z44fjE6ZS_55r5wMcHtq(9@DqiV$-=d_E|DEj~^z=VLz!(p42t@h003hGtDFVltNJG
zcxWJ-?iR}`s<!7&yazSK!kWFyLWJlk#<PLV@LURck)Bdx9*(MAcKF1e=823C@p!(7
zS=tw~9$^Q!Z_MlRmUn1H8oxTGrTtu7pE(H!$d=$J#Z9*QEs;Avxg1QGg+G|CDrD0=
z4e;|LqnJq1^Wb$*au?iGsiAVU@P1d(mv#G|VMEDNXI4y_wIIRe5O+0Oq#;2J9ja2t
zux&X{)&n#2S&utKg)x|~Tk-o0ZoEBWJ$)>|G=yt~#uZS%O8eC^#v{FSk=P_CcKuoU
zlAq~`3I5H95Bd8m=oKPF67UJAh&q>_DUPNwNu^<>i3Mrgrz>`;kPHcR)A^#36gIqk
zL82)VLRFkJ_x#j_``*UA++T$_RyP8^t|YBK3lHFR?RBei`dVhVyo+P|-O5DanjDja
zS*7y4WYV3(wD_klW29dP%i`!!apNXqV_~&*cBVq=4Tw6hLd6?Fxcs@4ls+U}0TpxW
z#tnrv4v6rPRvaoYtk_8JkcWyxWn0~Bl1Z7VCoEwK=~Y<WdUc8i?aB}ie-CPo+_nDu
znHHW?vr{CvHlA!nWY>$ulC8+qL@2QKpWT@(3RBxoHEzCIlj`ep<<f}Go?v0EMt-vu
zXv-#Q!tQ|SAESCl9;2G(4!!D+(#@l}R`(+0o2>OAKhwQRR1$UsI0+jrt@WM@UXy5e
zd7*0|#8If*B<nGKg-daTJH@MZ_vSEmsK0yC4r=fi%RRwlE}`|qd(M`aV`2{mwVE8}
zrxVWDQXVDJRI<k<^#oT9I3!SB>oa?l^<n7Djv-4U@3X24wTB<_B`;DV#^Ye4zEeqO
z)32Dhm5u>*75PO?b!l|uc+`F^VZMbC&6StB2jU}+1TBKs<p<v$7T<kCyRiD~PPGZ!
zfeT?)<yb}fk$7I<<g1y*(Y~LkMeFE(V^0!qyW;s485T@s-5gf1)IiH@XW#dTb146|
zr|A4AK<tp}EGm3*-i3zY6L;8=!N~ZZHs-bzHxmk`v@g+#p~kZn3~$b>@w}z*V3P^_
zlB%wC^A{JIrrWw(Gsr+=E+ko7A^+045qn`}|DE}PMW&-N&*pUA!uwCY_tVj=XI?~r
zt!QWJ>p-dy1UZ~EqnW*(9R!IxR5)*ulVf{pXjKIB^cL+WZw>>I8-QKh=eNNj&UZiA
zt8~AAnBu6gTW9#d+D4&%FQ*i}kLX3Wk^W6Wq$9FqrN(+k8t<B|X4&L&6rE6C>hF(}
zoW~D}?=2><I^w%j9n0=#(*@KDUuXakG#btda@g?`HS_1Pff*b-wL<Ks*Oct47FV~*
zgv?#5xClez6>?ZJC}-nF(@;(PsDvrd|2#=~!!>@eFI;qF9p<N4RU*MO-11_zEe_-M
zu4<Pj#b>-9f~N$`VV+8B98CU{@sAWT6(rP5=diMwuL`G&Y<JOo2D5|Ad$=tKksTz1
zi1k{?Z_MCFv}$BzFAX)QF7TipR|JhGUEw+k%_?wqWh9HBbdSH|!K&jCUEuL5kh#`(
z`n6(ulSZ#!zt3!@xrb(@ryH4=SuS27N?_xC9NL>2j-Q&TQd+;6B!({d<F-<4znQWY
z+Gr6{`eEEWnvhsnNaNbrbnQB2`NShJv2;*?>vWM|&fWT$aVouH>yeF8R;-6j_XFgE
z?%ed^$SU2P(+8Ep`!Ac87%6!64-e6sFbhCq4N~}NWA^s;U>|q|1Vn7yBtftSGUpS=
z<GSo>Y0qo;BIAu4Oi=s?IE&`|)R0+C9z+*3nRCU~8G5_pext7|eyOiC_KQf+Of0m;
z<`3q|&y-KqI`>Q&Y#)dHx;NGI!(x|)nNRG1Yi5ZiAS&DC$$-S{3<rmUia3&`D$aen
z5_Zz&>CwzIk&j;1$d>4W0`1%$Rr2TtyXIV9LRMs5`}YQHog9p<YqhE>%sI!|w(pea
zdF3vq<)56(AFJ8g-Ck7*h&#NC?nh$EdSZKS#pPac+f@jaqncR(4ev>iNu%LSFXPpx
zi?~`ZzF>e(;K(xb@d{~%<udU`l<Ppqv$Iy5U9#%;HS)bUHy)Qr5II+nyI-L#m{?GY
zPWu!z2{##X>a(x?TKKNd89Q<n_S_!Roa0(?2eN$WT9HX#<fKF`jQ|^^lkpk9KrY)Z
zs^!DjzBrt3OXlp8DFH=E!S9Z8mu}04*a=H$@~-N2j}C4{$o579FLaVjtRlY-;Onmg
z(mp;R0l}DofD<UpNjpFp_RB+4U$GSDnM1EQoeyFviDlJi-`=&FIjfw{klxwF?<1n@
z)f8qpOv+V$9EV;xH+zuxHHV#AQ7&lQ%uVDER{oiM+tB-QFDe4-FY-jY-!E)2EAaiS
z-%H8E`P4*e<B0Bbc4><=Q%6TYfl6_revYKF%XZ;2q2UlkpP?W6@xu9%p3^ZF>HLkU
zfOrIfh^k2Lf~MX<39rUkF^!}1P5Z<z&56uznTy)kJb_p{i5J0}XV`Y{Ys><J?c%Dl
z3B4$WJs<ckpA6hB(LuvXM1G_5cH^sQ^Hmj-cD#HL(o<^|EssW%ERrYgZk_@0EVgHL
ztOfX)8c(WB|GZ~%+B_~eez%gRMT7g}y`wR56?JXW$Wyt*5fLiEeJ;9)@GPoJ2EzB`
zX=`{1C!|Q#Dc-5q1&|cR^uCOLZ>lyVyb!%|KVb`@yLFkXyv>W;nVMLPq~>LXK__EK
zyi`Dc!W&7cxnomw^_e#AgD5c1$3QI^!IS%i+w#3iHiB*wiiKYl7@!z*4;sh;<b^0v
zil^#9NhO4w2XDq_{L1n-lL;RsN|u-}@4NZe0_|v?3Q~`LeDLq6`RJ%VK}><>DcjVx
zKZQTL;zX0oy*690)jW#&MmEjYIMB}vx6J0*1W#G%@l~dg{=vzl@=mu(SrnGP>pOB2
ziT0jF1-?2=>I^DO)!^Kktw*V4lP2DcB0+29R)>$i+R=mGQP)eRIGl<HSqG#Yv*4iP
z@{{n)PW_%5mmmDQe5%Egge2I<-}WV(oW<Mwe>0H?lDTI$E@scwcnURNnXc|E(%HJ5
zp7)2H*<5!XI}v?gcK_h?rddVfv(U2b6S^Cum>1;4d8HQA7W`;Be~|OA*c+1Hcnja1
zQ6*<hxOLS0f#B-wSZqMyc-w5Gnh0yJxy?S0xM_YPUSqM&`$YIy<@$nM)`*C>(~Yy%
zKJnfH_jycii@09vK$VDuIB`#(D3!234TJ3iQIo6@eu|;J#J{eo9L^gK-A6LIc2qk*
z4{k-OWF-H&jV*QY5Z=8)(cpSRLpazTn-4r8rKhKdLOUV^1FVlx!`th?Ty={|o8k|G
zPZO$*U%!5h5p>~1ph*}Q+AKD&=q&waaxRnG(L~r{28d6KiyZU18bwIP!UBCMZ1#P)
zQ=9yKnIu;BseYS2@HM7)!VL0{VjHPwby^LgW4C;0xGef7i(QZ5>6;stczF41mMdd&
z*Lsx&_0EI)DuXT0+)KZ_ZOrIxW^fopa<)ZtAkmmeq-DL6B8NSGu#N2SW6%emzeeIa
z>n$%HuPM{B+o;r`&wP9z2zmXrV`dQ<+eB@rx#lY~(^ErQo-2<maq>cLr1mLS>n+^C
zG^ZcC<kPB3tXFW&ydmbUCuPoorPWQ;QvBA+r!?2;qOOvL!W*^71Uq7iAhmDsd?vEI
z?@?Bn$5bngZci_f9|7vzMh#Z0W4+I;%*q)bp9|@+Gdg=kD^|mkrpsTG&2ftcm)~73
zFgbrl*w`QHi>9l~Y#1jkPBxfxFYmv()y~j%x2`*oC`ZT0M4?Gb=#9ieoFMf`c!)2>
z1Z8|IG3UGMp>D6rT9@?pDOMmb<}&Ty+@}V5XsB!41D%(m=V!I=Q4&pAqxNLH<XDJg
ztJ9n9xQN+SlnNA0f*A?>G9%SG<~FZKNW4#Lcp;iiQt&zLXIKWZ7BZtkpFlowHLBZW
z|I+MAedT7<o>gTU_e!jT02W#iE>UN<Mk%4Rf0}}5tAuV6EnopK%_W~cR%%jAj83{C
z{VWBE-+@?Kor1Z(%~U^A7eW+yo}h9|(+Y31Fb?w_d6lif3pnwbew)k7R39$pma;h9
z>lBl|igMwt_?b2NRHeAE7K_$N->dzhB7nEXn`#P&#uFbke);{4h1m)>ySfYt_@54I
z6NW}d9925ZHLGldH`L7|Ehlx;hTkl75&Xn@s4{w6?Wpq5`p>9wQo%09Iv(*WW`E}^
zw#xSdoImQN@;VJU!#)0da^zO2dXY_-jZl3ZDVpqfoy>CJqj~yRa9Zp5qK8?deUmA3
z)E9!Uw8~q*%7q?Pmqd;(CEr{nE3WD;t(JM<tW=_XJ-9*8#-`?vXBuh|Del%_cj*JI
z5)^A@O+%>5`ld!>puK*SHOOwuaIC_$&K*hiATcqjH=KO8pFu@CS0*+BC7V)j1!LxA
zt2_B`;$%gQW#{R^Q6i5Z<^^dx1^)cw)zZvODg~pni|4)atpnqidt}U?0RP49dBOA1
zOFGU^<D0n>&%ElCAxe%4^yn_K;p7bD3t!JIZC_PpepV5XM|zg>Yj#%e<l8p+eWTET
zIkGe!GD^J-!HgH3KgJ!p+MC9+ebkc*S)fJye9vR_ak|E0+e#v}LStvhnS^LWoKg7E
z18yUvumr}+xrkK7Zj7AyjCt+Gjq&P)hY9HN_-7+J9Rw>_`WwfR7ALK#U4AmfSYMx{
z6q6odPgB@@GdINF7t-U?rV##ESR3WRlCJc)5%UGc7=?^@S<!%BTkhg;`olnAV3}E5
zM2r4pWvP^%>_!h4XnuNmclRdHS8z<*PqnnoHrj~PsLRCIaYdp!P_V9v#s%3O5f^Ru
z)m!wcCNf5(=X2|6#aiM?VW5lf_gV0D)5-jh)s&B>BD$A1Dc9W8WOFHYf6qEJ=|$$9
zD!(=#?&!wWI{=pX%~Bf0xjD#KcHx?~AzpklF_R5(@6KN0Rr1jR#J%&WdQ(P<7l+V`
z$ts%P@%Jq$s(J7CPy@1x3q~dm>^<#0@^`uDPolVsJA7E>anjY3PY%PDBQ9Gq+FGZl
zr!9J2CNs)@SxBAS@>?C6y@`?bWfr%xzuYK)wA4)?De)@u-PYuflIp!SBr*mTmIVNT
zI%yLi(BZW|s*C)g`mc~<#r)l4UCOLvu_QtjDQQco@WjxOPBy`186&3toG-X+$6i^^
z@`V9|ec6y;k!ikh5V_^6r+rz2mUbVyp93kl<oTJij1EemyxLZLUi-CHxtAx!%~e-D
zH?z+(cr3?0*UWC(gl%JF21Sj3tU+2ll;62D72dAl>$Mwa)Z2Jla2{2i&@2)4^}%lq
z3jx!URhcm!wCW@u?P}}pw5Y8_=)+5ukk_64*}=nPq2^nURC|><6Y)r!CG4unj+{wv
zX218=w?B0!jkL{aYb<w5KT2hPczIUkTNW+;%1Yq=4`(+5WtD1w{c*2gdhvp{%)~ZH
zT%U-)o+^MyOs8c!K-jUZS>ih<inaLcPrRG^@qjctu;;SjYo1%-T;t+x#$Mpx?=Y}@
zb))i#lGXHO_vN`0b2@vNe{0Q>Ten_IHGjR)1%ssC&-Ax9$zhJ6MEVv`XV4yPYv8&K
zpjN;GcLP;Bsep=%#A|jE;CMtM3k?7WxyIr{RY}+CM5if@wpQwvh_rSNH;?6P#!-ey
zEO8vlw)hk4>6qUwWB}qh?)_X~F<C)WU0)@j81d_=xD=iv*>A(HjNOnR^5A}{RU{;6
zyPJ(PeaYu?{&v(jxI^ydAFQ}y%kH^Rd(-KfY8M^_yomOuGiDLP*-y@#{fgb%HiBng
zGE_&njP4*7gC>=^Gy(`YyBN7fi)?T4C-iKH$f)+`-dLe-t0P9|TrHMV;ahPb`Jq$$
z0nJ)wxg*lB&<k)kw`hrb-)uL<1<&PZbThKCg^}(X6_|CKagfHu>S(^P<ImNOB;9Mk
zuG&4N$o}?SGy3C4(TRo3jj;mM)YwJ>EjezqF$yY|D#n4<u-9$Lk5v*-^`TY1O4x&H
z`DV|T3n6JaqO;a~JT3Kn6TAl+YPny+>vd<UouUi(CXXb0yV%dn%Og5h+j&*(^p8KL
zR9&{f&oEAf1`2zlBk)Di(X6k52fi!j?iza(k!YS~`6po11J|#6tXQSSTk6w2N%;3C
zHm5&o3_#nUoDZzV-axh$RW4^_M3+~6NY>NS!=hQ%*CF)z@3o&c{Be?zfrK995}s&l
zPg$BPLJve7w0>xN{82_h>2gA=`FEjGK^x%*8dzaSmDxj40Kuyc(0!|KHV&)(2slkV
z*&{O>XiOk?>Hjy5{zi!pqQoz<2SLN~nm4G7r2SSd1hk(--Y}4Z(vS*D|Gf@W0+F=D
zfOBpYnl=H9s|#Q!B-n7ifV3Q_a5M*Z$ztLO;=w-MYacSiYLx{X03y!|iiI@LmxrK0
z1IA52T0GF86aV+qStJXX2eNBJIi1J5^X(lSC^LZe094LX^xl6L+`J=^-1#{nK^7Dm
z0mi75r4$5^u$QR;f0kUM|7&KVDSTM{Dk}wmwGvHtt|^GqIqKgxRv&5m7J-wB69C^T
z38ydr-SVe3r}uwd@XhVcOWuFi-dm+=L(=$n#iuPFoaFoep5*`AkIzNI8~2h9-su1x
Q5fYN*Gr6a^V!FQn2M%N{LjV8(

diff --git a/docs/articles/vignette-03-interactions.html b/docs/articles/vignette-03-interactions.html
index 66b05748f..d3e58411a 100644
--- a/docs/articles/vignette-03-interactions.html
+++ b/docs/articles/vignette-03-interactions.html
@@ -33,7 +33,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/articles/vignette-04-nonspatial-models.html b/docs/articles/vignette-04-nonspatial-models.html
index a19ed7fcd..f6bcbe6c1 100644
--- a/docs/articles/vignette-04-nonspatial-models.html
+++ b/docs/articles/vignette-04-nonspatial-models.html
@@ -33,7 +33,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -193,9 +193,28 @@
 <div class="sourceCode" id="cb6"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="fu"><a href="../reference/plot_model.html">plot_model</a></span><span class="op">(</span><span class="va">model</span><span class="op">)</span></span></code></pre></div>
 <p><img src="vignette-04-nonspatial-models_files/figure-html/non-spatial_graph-1.png" width="360"></p>
+<p>Let’s say we also want to schedule specific sampling events at
+specific times (which record only specified individuals in a tree
+sequence). We can use <code><a href="../reference/schedule_sampling.html">schedule_sampling()</a></code> to do just
+that:</p>
+<div class="sourceCode" id="cb7"><pre class="downlit sourceCode r">
+<code class="sourceCode R"><span><span class="va">samples</span> <span class="op">&lt;-</span> <span class="fu"><a href="../reference/schedule_sampling.html">schedule_sampling</a></span><span class="op">(</span></span>
+<span>  <span class="va">model</span>,</span>
+<span>  times <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">0</span>, <span class="fl">5000</span>, <span class="fl">12000</span>, <span class="fl">20000</span>, <span class="fl">35000</span>, <span class="fl">39000</span>, <span class="fl">43000</span><span class="op">)</span>,</span>
+<span>  <span class="fu"><a href="https://rdrr.io/r/base/list.html" class="external-link">list</a></span><span class="op">(</span><span class="va">eur</span>, <span class="fl">3</span><span class="op">)</span>, <span class="fu"><a href="https://rdrr.io/r/base/list.html" class="external-link">list</a></span><span class="op">(</span><span class="va">ehg</span>, <span class="fl">1</span><span class="op">)</span>, <span class="fu"><a href="https://rdrr.io/r/base/list.html" class="external-link">list</a></span><span class="op">(</span><span class="va">yam</span>, <span class="fl">1</span><span class="op">)</span>, <span class="fu"><a href="https://rdrr.io/r/base/list.html" class="external-link">list</a></span><span class="op">(</span><span class="va">ana</span>, <span class="fl">3</span><span class="op">)</span>, <span class="fu"><a href="https://rdrr.io/r/base/list.html" class="external-link">list</a></span><span class="op">(</span><span class="va">ooa</span>, <span class="fl">1</span><span class="op">)</span>, <span class="fu"><a href="https://rdrr.io/r/base/list.html" class="external-link">list</a></span><span class="op">(</span><span class="va">afr</span>, <span class="fl">1</span><span class="op">)</span></span>
+<span><span class="op">)</span></span></code></pre></div>
+<p>Because it’s not just the model itself that’s useful to visually
+verify (which is the main purpose of <code><a href="../reference/plot_model.html">plot_model()</a></code>) but also
+the sampling scheme, <em>slendr</em> makes it possible to overlay the
+numbers of individuals scheduled for tree-sequence recording from each
+lineage at each timepoint. Here we drop the gene-flow arrows to make the
+figure a bit less cluttered:</p>
+<div class="sourceCode" id="cb8"><pre class="downlit sourceCode r">
+<code class="sourceCode R"><span><span class="fu"><a href="../reference/plot_model.html">plot_model</a></span><span class="op">(</span><span class="va">model</span>, samples <span class="op">=</span> <span class="va">samples</span>, gene_flow <span class="op">=</span> <span class="cn">FALSE</span><span class="op">)</span></span></code></pre></div>
+<p><img src="vignette-04-nonspatial-models_files/figure-html/non-spatial-graph_sampling-1.png" width="672"></p>
 <p>Even the final step—execution of the model in SLiM—is the same, using
 the built-in <code><a href="../reference/slim.html">slim()</a></code> function:</p>
-<div class="sourceCode" id="cb7"><pre class="downlit sourceCode r">
+<div class="sourceCode" id="cb9"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="va">ts_slim</span> <span class="op">&lt;-</span> <span class="fu"><a href="../reference/slim.html">slim</a></span><span class="op">(</span><span class="va">model</span>, sequence_length <span class="op">=</span> <span class="fl">100000</span>, recombination_rate <span class="op">=</span> <span class="fl">0</span><span class="op">)</span></span></code></pre></div>
 <p>Even for non-spatial models, this function still uses the same SLiM
 back end script used for spatial models. The only difference is that all
@@ -206,7 +225,7 @@
 <em>slendr</em> also provides an alternative <em>msprime</em> back end
 just for this purpose. We could run the exact same simulation with
 <em>msprime</em> like this:</p>
-<div class="sourceCode" id="cb8"><pre class="downlit sourceCode r">
+<div class="sourceCode" id="cb10"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="va">ts_msprime</span> <span class="op">&lt;-</span> <span class="fu"><a href="../reference/msprime.html">msprime</a></span><span class="op">(</span><span class="va">model</span>, sequence_length <span class="op">=</span> <span class="fl">100000</span>, recombination_rate <span class="op">=</span> <span class="fl">0</span><span class="op">)</span></span></code></pre></div>
 <p>In fact, because both SLiM and <em>msprime</em> back ends save
 outputs in a tree sequence format, we can analyse them using the same
diff --git a/docs/articles/vignette-04-nonspatial-models_files/figure-html/non-spatial-graph_sampling-1.png b/docs/articles/vignette-04-nonspatial-models_files/figure-html/non-spatial-graph_sampling-1.png
new file mode 100644
index 0000000000000000000000000000000000000000..717de92611ed5cb4998f025979e6953857158ec1
GIT binary patch
literal 49838
zcmeFZby$>N`#m~<fRY9wtq4-m(y8Ro(v668cZWd;BGO&bjdX(|BAo-$-JL^o_JDrg
z_x+yVIp<gBpL3n(y3`S7p1q$P_r315)}9XvauV1W#263=1RE+TstAFg3PT_$keg`W
zCkvmtEWtl$ucReJ!N1@-7OC|H|J<^X)Uby@$ovt1kU|!A3?YyQ5UA)gWtW8Y31=64
z!;6cZ(}ex^@83UAaYGlQqx@1#DRw9Q7xH)c*Z3Wj1P+WHg#m$qv-AunVrt5w&j{1=
z%K`>b_3=Lw(4~sH-Ms(6_x*gt@%bHS!0nrDvFvU0=QV}e-JP*Ldmfb9w9ZMcKlShX
zJ-q4@1XBDk3x@sYwU5S4VWjKVzX_0N|NL}}5`udD@}<^|ls`YT2!h~Uzx1)Z<AZ$t
zvX~KB;m_|Y#2|=q0asJSfB#JxdRNh222)E%EpN{>hovK-pwZIPx6gjMz7u51{=T2S
z`oxX1B8d?4lfyA<YUt_VhSK@T0psS)>-TI6kiS8{e)<1;Z~vc$0N=zB<+>}h6n1uY
zWlkHKU5R{ZTaA|ux-K)QVq#(%6H)R!iE=7=s<Mf^j*11E<@r|Ve}~(r4Xe3#Fh}X&
z`_x5d`uu#|T2aGE?sMb~e0kS8-RgQ>9#Wi%iK67m#HUaW^P#eu0=q8V#^&o56>{(U
zrM^fLH|<Y{t`1ba5qI3k=7i<a(moKz$%|)i`}&bRF1%Q6F8axfv7v&4DGu|&DqDlR
zG5X5iD1Rn4GnUgb?y+oK`cCC!eyr!#pe#oo!|$InSyU={w^Ktc3kO2r*veI784dOB
z419cJs}=44%zw=H4X3rOG>NrB@dJ)r9rL`go!Yoi{C(FRtI^ocDa;K=-&dXf?d`MP
zm#4{}$oVp9Ms=;%@ROtaE8J8ZvsIKzjk?QfBJKVe>dfC?4gAx`Zx{2|rfTAEoNP>?
zo`w8fikaN*hxrLO{@*PQGvN#>4{0T1H8fN!Us(OKLLb{j3@=?u-}-N^({?x-)}Q$*
z$AV%%mWO@l?5s0ylNbE5of0IW$)_ZjJ26-k_|)k<CUBJZOc&bU_YsTZ*hVuI&P$2|
z`<S01S8lf2MxvSHx*#QX`S>M|>YuHVqC6UDrzpYBHoxn(*Pu$yt-O5hmx>J6UG3O7
zAgE?)o$F)HQsQ9r^4Ij{6=8SYJZ!DRrO-#c+Jk4^Z&1*mx*uA|Ch=>0`0&AOvaULg
z%i79O%j-q8ong84l&<rXXC>vMM+H{s!aQ8(ONoTq?w45wt0Tops1x%f6)!*34CZ9C
z&*(=TzWMU-YCHQ(&Ck!9>@WAZ!6&?>*yeZchovX-Iny@Pc8o5bHL;NLMy?L}Q+Q=J
zo-NSTUtYYPu<sF2twc>p*uC#3IF6U>Og?S%%izAI>-1+iV`Bzhs`;>knP+c!j|T%7
zlDL{?dSi_#YEz;uTfJcvncfAYArfIvt>J<K4(A+d{DOZcjmnju{Cp?myxgRZa$=yP
zccN1I&dk!(3}gPF5kIeo%*pE^Q*3g6bvimaVjjD6m@E>lKPQ>=5K**r*kHXb8Dh|L
z#ww)4W)C-prx#_JG`2pcsqgWU5e}hplbWd<&<pghYSvfU22q)|7SXj3Q(Ux?ryyqe
z0b*-t7Qajt4`F#)Hnr8<v%9cr)*lo?Qj-?`)NC+3FeIeW;|}ELcdJ&#5}6R_H*r6u
z(DvEVNtbF_dT*0Y-5f7|+LIL|wNLn0cWf7^94pwF(;Kklf>xH5*TnaKoB4ckc9|Y5
z#J9f~Z*6k2Po!FTBZahW$7R38_0)S(bIA-9ecnGrD<oNAH&#Za{)c>Cca?VFS?i_F
ze(By6L=R85S>bz|Qry`<MV+}bXSUY7oBH);5U#GYofA*w4*fDlQCt(=Pfex1IzCw-
z7DUj_g2ltMt4`~~)*1Qe5hHAn?jt)<ORkpva(12greJ5{+MwcVn_{XQ>p8TOo=wlN
zbk|$%2TMbY%sihu+;=vl$9=Vh)a<=%RynX-z19{I8Xc>i1v9j0Uy$%5&qRjW&9<8r
zD!(iy#b=1)lOPhkiu5Vt(lRm;RIvKdS#hb<QI7P1tbvHAsLa5#4B}#Z>si5S!Skj;
z9)2N9M%4-Q_}#T`@4nt*^Y|syl*{qr;}RqjTYQ29ss3K`xRn;IQN5iQai<Eq#~&G>
zTG0#nT6)hOiO2FvX#8!=zIGxkeXqi)+4}7}yUW<{a0AOnkoMq@q|pSM=hRzu(i5Yu
zt(Akhq0V(e9I}lGaw_%9L6<g}Rn~7B6PUdwbACp?+P6666r><oeDEi~3FmyBk48c5
zHkUbB{$gzuovu-PxO{v_I8|2-K89<<?XWn~y4#HnCF0?6+w&$l2%+!}qYBojmrENh
zSNt3f4~j-TqrXm|DT6U(snaBr0)7vphvVUn4bEH^s^2Dfq&MOEc_N76zpyj-hM!Es
z#_q|_-ue7bw)9baNX;NEL@uB@e=2X32>Ya`B(qMd*YbH9d1=HdL~l{7ZRaLGV);h+
z3^Q0xJ}fwUHd-xlau9(>OvzD5Gcl-nBEm;UIqp4o-|y)@v!a>iMfpg${nRoJBx8cB
z_TJv7HPw&(-4cE}`_B$5lo|(nXzbqgFwZ)#26tD+OvcYSGl+QB@8T`~i9L7&AHw@Z
z(lQ7qqoRr;Kaoe<X)9{_m9gr&S;sxywl!_JJY)YvMJ5%dC<nEYzZi}pYdH2oYO>hM
zpNeok0|V^4IeY9fwzl_4X!I^~h`tx;%4a;%&zBW-V5(N^zjIxKgOvM<0~!98Q6#EK
zN<88<)xm@5UKboiYHn+fOpkf(TX~E}Mlp-djaL>LDImq8sbvlu0*<^}b$WAtzLR5l
z`t@~lGSON>)YS2#V?5s3I_z6V2aap!Y&24zB&HEFGL1pxx4%oZV&Syus*Jh<Ym^LL
zU7WIiEFHJaY<fO4zE>#}rO6RZFLvscU|*+e`fc~jK8qbb8b#J0+xbP{IW9ys2OmGD
zv`0WU-QA%1JCmFBUVCI9|IcGHaaCHyiML3eg)SEAe^OClZz`4P(&X8m^92(uUv)bv
zfW>tEUjHNSXdH?TY(By#(S8T;bMB?}kz|U@ih53G>Xpho`<`LzNZu*sb!&~4tO~jx
zL`r8!kO<!HvPT#tHMg2@uSaScqu`Y8vdS|LQ>4Q_L2hN{TYu0!`<_M(RhFuK828c>
z-AHc_W8<9XE?yAc)3Gqr6p|<hm6l(#!zi?#^oE0dj;ofdq**;lk;AyB_Q=tHR-T&4
zQp4Gt+y;~9*72!}QDT=)DiQJQm@QS?MpTqnvbUEfJKc6&;|?`>gj|LcD?YOTF65{`
z&)go~_B8#uo{2K~ZDg&W=m&_gZo=oadvaMNwVttce%=so%;qWdiON#?mnXk&SfMqO
zudmdzH`UBayi2FS(Mc!}kB!8_L@H=jcT#!PJ@dq0or;~uLgjDuu9G&nnfJq<$U+ld
zZ|u)SKSMDZ&a#b{&gM4B_Cl7Js=q@aDerlQbEu)3haJy?_`ZG`W+~fk(l0Mig{`Xo
zMHp4jm$qQ8XY(0bU7hqUUKdhwdcon5jOWEXmvt&ZXLIK9^Ar?(p{J;ji~c7SVOH6?
zvU9_h3Q=R0j(kxuUp=D_lXwgC8RNAkq^eu4(}#_399IOVMTdZH?Tea&pWnA8&ZRGG
z)wA)OChvW2AU2+>cninzkg+2uNq2yS^b6EuBxkW46ne@<zM?FwyDgl9?>0zHCdJ{I
z^g)L&XYb<(@vdxw;F@DCz2A*^L1WXeFXg|GUs`3HaG7S0d1+x|UOu|c+YAo&B&MU|
z<8xbFi-o!QPB33Mk>AsWspBQjnppPkx-E<(_xiBP&y!1xp=bvYq-sZ<p7jJh%ei?>
z{iHr@9V9hMg(?&=g7RalDjfAaWG4e}w;VQF^-Nqk2x?NrIaF?9)XWM6N%2(<=V3$J
zatZFF^4xMSquTMQf#ns@IXl&Qu*!&B$1AmLRqPxM5wO(9_ZsV%>MuSjXBk}xXpHp#
zkWWXEu<Iei$#Eu}3JH%2)ziZ>CF*~ZrZ^E7-AD-W$dFG%8@0K7SzqB)|JZW;3p-OZ
zj4`%QY(d@Arq!}$YjhE#SgT*_=Wlw5PQ6?CW4Q$N<t132+VFf^ICGWP>L6WLEW5<o
zP=S)TxF0cxS-RCkO^#BQoE<N7gKTc2nUo@;3~1svAdlYkR&DPNl}n1cID4yHnmbtb
zCc;j0w5D`g`jfc8gBIoebM7^-cA-M2$o-XgLcIx|S0BqRZ8mpGwldJsBiB#SeHdUY
zk2RIt9o0UyrZuk+wL6Sh%PWZZeM&dZ);TcC#Un6t<U0y3K<*Xs$7!3}!8<u_5d`UQ
zP)S~Zdamv=Eg&RB&BLRPfLS+xSkGJM&6_t-gAwUCpo2cuufC+?>P04bEAb|Uxo=3?
zZ4Muk;;eiSf=*bo8Yhg{9=D=wJvr+;^qR={ki1CFto=}cMfjBfDum$%`i*R_eK9Y}
z*Y}2XRYq0`w)rZgy1kAU2vKj`Fxechw!_mm{o1k6AH!<!`o7-@e5(imN&2Txm1*RY
z28Awma4n}Abd)k>`U37eI@47O;{_dpT)fzzUm3X7|I@{E;nI#mOvsiEPAwL^Dn+I#
zU0D(Rg+7M8dne-Uvz0DO%;YrBytBFWgz;#44ojPiDOL!?Fa!*P=jmaNZoQirh7h%B
zR~(0wyu4bi)5cSa-^F!nU%!4O=5xwN!zNCzs^Z2b;gp$~nei&1gOVXyzr1B5#ibj4
zy0B%N`$NiBDagrM+4sv7Cd4Q4_u?D$qT#%V2s{oGg?$QrCfz#qu%|ZDO_+k1*w`xm
zIAl~%KBx6<cUkBv;#T0AFOh|~Ep`$?6*or9gB%+5?~m5I!>P>ydS-%3NlI>;|9P)=
zb(%zoRM^Lkaj<II$;bGSy?5uKP(@ME_XJGZ*@hiaE1)9HRm=#qYB)AJ*qqP`4GndL
z8^3-n8HS9Ej7RLYQ)!ooNgf^^ey*t$Rx?E+<W;xW74Q7b5p>xPo5XBuvfg(HQ7137
z<ZF~tK}}Z_MPpjJBG{HDsn_O%p%5RocFs#wKCiPXs+9@nDVPL>XZnvHq6_o$h+;U7
z$3Anj9<JNc+PXC1eYmaxRdjH0NHs_rQ)6dj>?-1*&N&rmW|=>5V?IK(^4a?AXeTVa
zbcBFQ5hIbC4iyzucX#VCzzs$f8c1&QzwSU4qnUIZdIaFvb_?xx+@!p=v+o-&PRtoK
zN<^9Kw%=yT#%C@sn-Ga_HZvk2Aw`q%v9Q4P!Vx<Q?%vneS97wO$6S>Z5kVE*;W+XY
z1b(006fp4^I2eO9vR1G>Lr1O*e|^czG13$qGAsK2DOm`FHrV2x&*8?H8hB&<v|A*-
z%2T^}Nkc<J(}Pu*X36Uhh-LBl^KGc2nVDItfsv7sRJ-R|4OJs7B7GB1?z#U$;AA<C
zP<yK{&3~v!PtOFgXcUr^^VMI7Y>{ModmqGF)wZRrs^o8GJE%9vjh6SM4L|-^QiBhH
z1a@(TH0gDdIgdYtj+PpSq@T66mX?-AhKJkt#o33zhI3Uq5aWob*%kP$8;&#STU24o
zmB^8wO&qy~*sx6dq*f++5{S89`kWlB89PctAb9bXsy@2i)wfyc>7~N(@bJ<>C|X|b
zOOuk5YcL6@G97pfRUF8aMJ$S|-PLlBgd|TTZzx+nx%R2qI8M36qVk@fuCKoojz4T7
zcFUsTDO_u1C_eq@y)c=a-P)u(x_Ns<$O0|yC79L<!yw3hXDoXSe9~Pd^*O3+Jm<#S
z%6r6qsYQ(!nqh-r&g#|#+z$sM=#=&khV|A~vXYq;!jqB;%gu-Bsi;K3DDVB<A9(VB
zI4QTCdkpR(tLMd|BR$;0@aWM%mi>L*XsTwecRUGp+#Y+wHS6o+VvlU|03CC(8Y*#-
z&!kgx^x6Uf$$uRLc}hw79Gm!Q@8(49!H|~qN~%A(qKwQ(QXK#OfdLNB69+_y!Dmn%
z4Cb1QQ5(*XCIn0e;5!NwLF2GeV|i3Gv{g{k6LVQ5P_2AJr+894Qu#M}ryM<h{(OJC
z8Fy{QpMulyERx-vmXD#0#z1Fc*-OWMuT_u;69TEVyT7E%rTY^wC{iH;hcTdG<`Y*F
zlk7>?r9{lhfTYi#Q;d^655I7b@i`5$_wd_)n<Kx-rGri^fbv&l%Ima%%5<v1Ggq(i
z0zc#mEVdtWLdfJG7%9K&Zt*Rg?4#t3`QY-b4XO3s)YXQ&SEwV=Ls?l_UGu1;8MQJ&
zNNHbj$n`ou0Z)=v<Ft`mTwIK~8u!%n3&==jL-`sfp#KqI7NQuna*cbE%*HA<#G1vt
z!J<aw0C4#PtNP5g7b7F6mtr&qVWbFxd`LM3SX&8uyOmdOfk1>uuB#vL-~SO(?H}&s
ze<xG-|KId_QvSa=rm88|oUjj>p^=f<Xql;Il}#!@FHfL>pFXLInq*=<YND~faid$+
zW|<zU=;BgUSi8xK;AsTh_CGhzY`2j@6o~JK4UzMHSs8O)DYfL=EgUYwvuyW3!pSm|
zP*s4tm2>19+Kc(Wk72_-dw0#?d8`A%sW;`>JBQ`ohmPRS)`<@i%6#`Ib<yxbxa(Vy
zB1M<io<x4BWo1tbbZS)xvgFl<3v~`)YI->!i3!G{L-Hd+=lUN_F%2i^xGU$`8BPp-
ziM^=w%o^pg&UBI`j%Um2nCmNi=f7WhSvVjxWZh5W?4$yBbCik2T)YO@5Q@<f!=M_6
zl`kNgn(fTCnDu=Td(#8A_SUje^OIi;47ALkFLzr?B=Jji-CHz0+MZb*EzcV&)X577
zxS8CE3Q@IjuU{3i4jrPQUQ5eL9xNWKrE6MgBqM{<(5!u{e7w5JgIWGk*xIpBsl(d7
z+4&FS>hWXk!0dOiey&p1C)Gl2*Jnqjoc_(JwOb8Z($P$?Ph@=i?F>anj$V-Ma$HW!
zfz@~Z9H$wqZYixBUf4p^IaaV-&8J=oqa5a6(C4BUIyMg1&bG0yna=Idg^p<Ru?kDT
z_BrF@;~Q#FSdLY|euX~baNT_kCRuX`9TM^Wp=rOo7DtMb#Dtd!Nyo{}TNwTGG!(<3
zWe#1kzs)r$Xa)b?4uo`CPEJ*w8i%hn_y}sS{nxMFpTQ&rR;Z9b>K)}=oQzD<mrHW0
zm0!$PVJds0<>jv~ALeO2W4(PviTZ?)3VNu6bQC_&6Jnw(Q}1>_1qG}=C;=g~Nsm^l
zPpP@Gbth)!>|l8>WTD30>Jw;z-%c*+5!u;C7LeOMK9Ibz&r|@|K*RXSLs}9c!aMib
zsM@wpy$BxDK2RWUF0R4G6InSTyYVM}KK@JeKwfISQRTY&DJ|`adM<%yr4C5m%Y(%a
zHxNofd!Y!8oXJowEG#{xNm!t-D5Qn#+teSXBI5W&DE?rwQ572Y{rlTn!bmAMGJkqz
zsWrqrga+QkriL*3yWAj0+*~U39vx8}zhODd65_*1%I7Lo)OFOp%K4J@dK8O4>)*Gl
z=0D+}7O`&_$!Btw4QpL1uVIAHhGR7=d>Sj5SIRmV3Cl2Jv*7m5e0xn~G)Z-h)zepw
z890(aUSiHTABJ{pRMcQcu;u?ZRhmB^dd|o7k<KP5$yT+1uR;F>7et}Lu%(OHCdqB_
z<EvAg;%pK`!a>M9K8in9Y}6-^A;0|wp6#RlLzHuh|IY|I|Ep;NqpSvzkRiei0Rgu+
zM$W>nhEmkezpdWkE6lR#)&K2NOBi*==sRLUUXpysQlf^oWy$ANud0t@x3gXI@;=xA
zT}y_&ef##w<_$>T(#H-)C~sn@zAa6)-MAGh&xihYN-r3q{U!;7Vxt0Jrd8$Gs%n`#
z+zQ$c!j7da4Zw=d9Qia*JNG&Q=P+}4sz|Oz>RXs)%KhsF(<D*jyj3_jsLZHd+_b8?
zf4r2Or<fsSHjo(%D}V@p9W1<`;Rndb7Uzz9D>0$X@r-(j@WFb)klARl8!3p!ryo_0
z+XeADET;fLhmMD5q(l$^iA;IHV^Mjb^D=;uo~)rjq`p5~iFGO{cR5G%toFz-^h`H4
zTIsn*lk*}YrcNb?R_kwf05$`WEwi&zX2N5yOY`e1U##Zsu`?b(YF?hh9fx(C5p}fN
z;fV1$Cl@3}U%g|orq=Y)gwTv-{l!9KWNhNb@G;lp^&yHRH}O*(iaz7|8V2qMO+|an
zTrQYb<m8b;vmbW3Q5P{HC4d<5;vE|H;1@AWUq3&ou<rE7zgRp<B}^2y*?iyk^n!9o
z8DLpdbo4b)+Y|FTWD6cmqk#XOz9xqh2a5EU=}sJ+1#WR1mNGw`$<r_!&d;G26K|hk
zG47@&@lw<@%8VPT(X7QGQ++4Zr^$O{Ri2~Tl5k6(_fN3;iU{O$KZ60;{{~UbfvPGa
zJzX+P(R+frd2P3i|C_!F9(io-cLcOP*-O66|Mm_&&v9)?kca{z_3lBI5+F#@B{X&N
zikzrSI(F{y+Rav(?qwJnNrjDys`To)PZWB|64SnqOg1FGmQ7<GJa`}sY8(}SD>1e8
zcuM_BO2a(ZT4BL%t8wu{hi&Q*g;9<V!F|$XV_Fkv?GFzBo4fa~u=PKQGXCzK;z!La
zWp)cvjM`NS_h{rIs2taSzu+OFSvWW{`qUe+g-0}?6oyAmUbt3RD~E!PqYMVilO2$c
zl>P(kPtBLCWP_$s-$y6~-o#LB)#Q+kWoTMa*vOW2EK-x0SQ~!f)2!dlzl}SFo+5O4
zR{HZJQ5xVG%(kb$5!NpO0W*%>_|s5<R@O%#mYEGoPSiMp^_4j9pz9Hu{}<{v5ud`h
zhI>;*4$k3Ic~X(|uERP^(a#^Pwr$sVvdO=+iKx>p0(xlZBZ)E~71bZj2D$C`NgMzX
zn1PDw2hASDCz9VURju*zh4uMSI9TH!rwWs8t6%^48OLTEHI{aaj<fnsc5JX@Rplh#
z*qNh9yDg&nrJxpi%F4=$+wns5!j21~dUlh4_3BHK+bSKAU#K&qprG!~&W>iC%U50R
zb4O6CXe$0G<c7Fiwo?#f08nAO;<@BO5x`Vzqknl=x!`1Y^tv$nw`#6pp7W$Dpr<C?
zw*(M9mI9Q}+40`-DY6e+*XMo7al5YH4bt93y?Q;k<k>Njb{Zu>m%8!-VBnvzSL&NA
zL>0L0CvWBKMVFR)*m&#MNAkR3=rkN#c&A&WX3+Xc#mUJj6<(#y6imb}?x&yaeR+Xk
z3L1_VNf31wqh?uep-$~n!*)FBzl<2MpZ@a2=~ndc8mz`~ZLrc>uTiiZQA`Gt3uu-a
z_o!-XCro<6bsD>{Ax1IP29=qYbxN1;ACmPuy(nIoLey<DHp>;N!{!sGm&&mANDhk`
zXFI}@CY+3TOO#%q>^^sgZ>j)ThFNd?AKVX%udgroECM!wo<}y%D1esO*xAqjGEYP5
zjb8Pt`Rb`3?$P8rZHz`8^kvCYV3YAG-+L_cxnrRu9~VNKb4$odIa+miZARBwb4VoP
zm!#*>;Js0=SvAJV%Bd1T1&<0W;kmgv4&!c8Yzjf`iCSmo{f3qxLg_Fx98#BW#;K{P
ztG%i3fE@UV%&X`R?PUI=rG<`-P4WB`URi@Lu-$~h3y2s_%hA`43=oAUU>lUpi3VOc
zNX(3UYILYOB!=aws!ToJ1aHQBw}IRX!h8MMZim@KO}QI<Q)k)JPu~Q-IT5D3EN>M5
z?_`q~Gb5vHYY0W-bjRFIG2v}QcsFP~+YKV-w8&0Ld61{|0?-B2P;?y9;wvb0j>ueH
zX2T^_;p43{@k50L?@U)zE8o5<RzY^Mn(6^$$@xK%HyBRZ{mG9X$7Nel5PTG|Tc3M3
zYDimK8-iM!*a8afX&j&PrkGHkR%PKKAh1jrLEAnIb2~10q@)&(Bw-P7G#6l=P*8RC
za)0PxL#NS%e-|%Z<`BXdF4emAfa3UUuO~7lCL0?57@wFp$^IV_3jW^TU%FjKY^5hI
zZ15^wLm<D5;~@@!ods<clDjQGgcWGPOW-7p=XehvJeV$lC%G$+>mf+1dawvr{Yb2;
zrcXeOh>VQ9=8_P7E9oRaGI4XO0kAR){4P(eNLMVE2hJ^t;D+D{?zL_f7WsPS2tG>0
z53BF<LlLH`$utREG9VQ9Ua>78apPFr@!6be)C1{lw9d73v*C0TIoZ{FEM9-HGd51Z
zy@JXVFr}^`A;v(&rlDE-a=K3{<fR>^oTs{JREwYq5EPS73G;t9P6Q1m8Qk~dCznr6
zPQKJoQ21;%n4JpvMS32d<1Y_qK{ldVXcXu}JpL8ebm)1W+~u~Vmq!KF*Z&4`z5=q+
zT*Cl^Ugu{jvG;y=eWFC<@<_IQ&5P<)_m+AZmxsr!8yTU*EH8y8%k8cGw?;~{s8-gB
z8XI1A?(aJwDx-%FALfGLnsofjF$j-1Qc{bAXXOK|2r2e8ppjoGC(_fHKzm{YC?&-N
zIpa6pNkyQ}z`(*9%vH(TPrf{pV2YOZ)2F4QgZ2nr)?KqR+mw|5x+Sm?X)}d*$+q<-
zSaare=!DN7TQV?1nj%@TM0l(!!{if(jHRP;+lpt-VlmtWki9d$H#c*5ow*2{u9u$d
z56B+?ev06z4upf%$@*IAzXRd~i~_mio8K#Uks)|GM~R;EcC`goC>b)<IKtF9VEx7M
zI_4U6#br=<9X$i#+iavH!|iZAr^0e<QGm%>>|X?M#5+jYvv+85Lf#kDTOWdhVSwp=
z`szmjWyo65zZUC4;s(l<af!Ce$?Cq+G{3-HzNWKB^A{^fibY2tL)A^DgHEEBL$a4I
zxM&$IF7Hx-B>|z8POz{?g$T|$-1dC)Po@t3$7}Mx?1uUu0MEz;R){V^J~LBNin?<)
zzP_psLTB{tn03J3F1`;arK>kUeiK~3Y?2^&qMiZ@>AW!rM2kBykrCS7-uAWXUWDTx
zOie*2JCilG5k&>m1oiP-5WJo!0~$cF<B*nBg=<@dNFd>SDJu!k<3QYaJ2WfFP3pyq
z7i?I~tEvvuO-NA1ax>T!t~Ry_dK{J>^9uvu<27Nsl3elkVBrP+88eS`^Wj1%Y3YGg
zATL>oAu^D<dW==o3|gKRlFuvC`}Yb~L-C)UHkt!?e9Iq);UfC{M5F+`7qvzBx+W3q
zuXlImupCuR6AxB2GRidcJOi-`QR*J_ir~^wQhqNHgrrabXArq1Oi_<XF9$X@%edCO
zv9r7bzmQiGd%qSUOQJD^9{5MU^`K&5Q2-zXQFIT2dv8oNPHhQ1(Wy~JfLQ=w4iJ^N
zC#b|TXsoMrEzOmck_msBb+2xnH?;iv8lE-iz0O@QS|WQ{lNLmO1O3?1>j=&I+L%UR
zHq}r+J3IT-W&7n79IACz3nE|yA)6k`>p6?=D$wAXW-EClCnjnMJXv!NryG@PKsAB^
zte+03bMwRXkzDn!U*s2&AbCInGP$7G>hC?4erhtlW~N#jWmSK{s}lGzqFKK?As)Ut
zP`SpHt9D9K{<>Klm4_^@FUx<^LSgvd;qjPNt?aE^w*avoQf4#ryeo!P1nm5<`F0*^
zBHiT9Ijoy6gkNrUJd1jw@ny2!eJ$2FsSofnE~k8uu>hP~eP>AZ_?6a;Mix^U)KlI?
z`xaAY4_02uCA3X^pI^Q>cgxga5b!uJ&%JxcYSe>;;x9tx<<$i0Lqd}KW;OD#T6ccx
zdOlThr>$e!!<xm#5KlG&pjF^Nf#;=zUIq`-{g0Ch^1pLip=wTC?TX(pfcWtV3A2`x
zy>)YPauCqc(Zf5A-)ib31<_vGt0jI$C<ZRBT1OP4<5Hq??sH#^hFC;}|D;l@15VLj
z_r5+3$iy{61qF+d46$r#J_9+UFG5#Uiz<kJipH~{pZWixxSnuus358~!P9kiBEN^t
zEL2dC@Rej7KED<{eobyHT!M^z3Gh0CE(TH;FbSHDSJ}Fq?X)h~aC@-fLew`pHfr~W
z0e!Xek==DTYC|ZyUc*>6wrAsTdATiom)P%w!!#Q@VFgo}(a-Ee-MbD*KDdd3o|=`D
zZh2t+r0vYLngil0FgUpD)Z@Qk)=H{27A$His%1dg_lC<0=DdG@0|6i2r$E~4b}q+K
zqKD5?Cdo$ET2e3NhdHvhqJ_r#+1NP@MlkBR9pkKGFUnQ1SBh5gDP@ug8~(VhrbQk}
zd>RC+=V0pcLJiEj_M7bi861BRAU?nVJD&-VvP3wI{Ln`t_A@q6b{a0c5pi?v!gVv6
zrr=o0V6AD;M`=8NNkh-C{m2d8Ujv^~iF$Z$+NVh56-(6chnx^Bv0sw&1Yq86E|_bu
z!cwc_<UWU4#AekTPB0mts&3IRA`qnO>elKO5OemLH($OU2|N_%Fkh~x5npXY$1y$l
z)*Q9(6L#&9gHRj;+ZNcFxu;u=Q~iRVsu2G~ngi;`a}Zym4v`@(j+E@uLZpJZ5}T_z
zkgW1Ms*Rtx@-!XF7Pp{fWw>~_*Ngq=7tlvwz+fpRs@^D}4df3hsQckMh_-<25}=WX
zhG}wGytJhW8<<h(@WU>H51RfvRR)}Tv`_LK>ZrN-`L)9w4)b}eCD9<I)m~%YA3fpz
z4M;Ly|M$!Jt$-7vMf-=b`F9Jh%q?Lq>&bjD*s;0@%z?O~@0$Ef;hJ<w(`JgsYNoZ5
z1@z#;fm=~F!oCby|L-Tpz|D)4?d(eG8Z9k=9s?xKiQ&tUeE3B<ki-ZGWI&LnX5rZ6
ztKV(G0+Ol;Es_b(d^n^`wLsIZcEWWD5p0yp%~W_2qrKWE*A7;XcILv;foOX2<w_H#
zEtN}uZL%8yQ>{dTRM@P~)75j%F#t)X#pABeV}RLwla}}hEeK#hNAm2uQdA;{{Pcfq
z9MW?zN?)FPh3I;*%aJF2M)7$CkO@<WL{>jLUZvN&B#%p!3mIQ)<le6at?fX!=N5wZ
zbQC}$)7v8%@<3WL8?XAxJc`JiK5S{OwQ37a?o8f~Ak!t^x7HsGyUNp{goPmYHeQ}@
zp?2$e9;j4nGu7|+OE=uV;{U+M0$U=Y9|RL~<TeHVbw(Gzr9)bNf9qc7eFDT-Sr{+*
zv7&Yluy^M=wsJjLMT4|g<X0V|b}RKHys(%;B5-`yKf6ofaAmPR`}yzOT)3*wg)YFy
zB_$=V@BcL10`}ot3f`+X0tJ&ihQKE?BVz^dO~W9p6?S%ZQV!!&Q~mLB`+Q-chIF;J
z_m>ar=DkD5jbpuHZa|OlK7bAa^*Msj3kT3Q2tEMtv<nU;TYqj1xn!k=oUi5;Q9-xr
z;YsA~>#XdVh-U$pPY0VJ)ChPTuHfQR1DER`Lul_RA@^oy(?S&`B?E|>|Exjb(3rbE
zlTKrnVd((4vil({8{qPx1Pzc0dH^N7gSqj)+@w2^Z|xJ`W-s8b51Mg>Vp#t?5JWqQ
z&ic(%qj#ft2zdeEr<XJ1t@@kZpuj+htCFKA_Uzd+AanZ;Ij#Th`xtfI5+p(-^G1qn
z+|$sGg?7GV4)Z7gD^=1pn^`(dZe71cnCG?BHO!z*`oAC{!7Tux!DgWT=cZMZ;3s$P
z+%dws+720`h}5l9>jY2_<oEN6)<@7?mwE!{w<X;Xl}(7D5O!6L`!3&O{q0+Z?sH5?
z#>nouKNXb3dFv}cTI5A)I6KB>8n4|TX7~L7#?1uvF+wP)IX@hOHF%!Z^?(KF4W!%V
zbxuSNnjC;UT$Iq=|FxS6fcG$EU=j}EL-Ls=a+76;Di-7wf?`q|N=KJV3XCftWGLyq
zEpuE`Mc5k9amk<TC!KEA%7;C*Uz9mm8}8|Jy1J>d^Mqq`rhIZC5Ds#HAtLIq#AW9d
zz=ql22T_MekQNvfr+8m#s^<Qjo1^7Gd%td&T+(2a=jqdBTrRCq^uX3K3R&RNQ2}7a
zQ}>97lr$gArJju*Fj;^Q1S&Ht&x_+Fqw=f8ky#7KDCK%LM?|f#t662Nqo^46SSBU~
zP)opvAr*Ci1fjk&@31OA%sJ2+G+VaT`>^AL1xq9?%Om%*4+Pt^?+1bn05dT%GO{|D
z6Hb)_%A3)0^JuVN%3if#0ObI`3JzioF}u-6sujSAsDaHy#CWycSH6G${uD?fo&ZGF
zY&D!d8QC=NNfJOPJ;*312&wz{3jXzxjpk^IR=M%Roh?Udd$#tBYqi}Lo9W^jEPN;H
z4hUNOe2?qVBNFIBn!v}kOifK|_PY7XUw(Tlz4_)I@C&Hvx{A@rCG;Y!);)r!G9f}2
zRYaE8eK<MX$V0HakHkOr_Gd_Qy#94t$G(dL1G{U2YxBp9?5<n)$?L*m4Aa0s+2K0Z
zy}=SgX<&4bwqNXg0n1f605*~wu%54UF-2A?bz79v?4EU0Gx+konzhWsljZ|Pd=h-q
zQOpmY&f>l#J8iwF2c9td7J9N14zZ$()OX}{7sN`7?~Y0R^zS#2mw}2>v&<wEfY2;p
zfFL}2!p*G?f_GywCs<|x%PIg$rQGenOt13&)CzEDG@Ku5L2H(a?tum569$!(>;-WT
z0WryhyvCXBIZgXTLF5?%E~yL<8X7xqrL6BhUXNHCz|FTxlPnpxIybkkGqRBH?V<6i
zf1t)DBBP#N11l)V8ZB4N%gvY*lZl?YwgziVi02rzWAxf1837wWpte$hCI_@1u;hQX
z@mVgY8c*iK<*A@VL_~Jg+F|J&dvSC$3?nr5gC;gbaTDB74Gj%|hg3j(lK+Y7)BpGh
zi$a=}NyZDVGDTqfVrQx}wI^Z&J}P=&-`KLBco#-e5aPRczwGbtzslH&jg5Wma3wSi
z<tU^+pR|4bUTH+ojYLHGb-=4vukh~On~O(6O2N8`b5p@a*ayf1SgZ@wk_bFTWeO6n
zSRoHwQWCuA`j5EdHsuM2h5=KI@;rVbbN#-e?%%UIM<9)ugpdmiX1{o}F;S}qsI~!6
zx6kjmFu84kbhh7wg1ZdjqcSi;_~BAmi37LF)z$t`=@5bqs(g^+VEG#CWc)4)!1y0E
zu%Zn3nmx`7z?L*}T8(Fesab>p(ZYud|4-oGCQ?%Mkx~Rrfink!jw}6QVSq9P7LuQJ
zJxkq*@q}JKfBrNB$pW!WG&cS;Nb4^H97o+MYC0JpY?y)*J16kQOGJb=0dTs#yIU1Z
z*<h*&IzmQV9V&=_`|e$r#h<7SqOTdmO%4=DPp!tegGso$DFzUm16)2-_bC3J^?~Fa
z8=@|^B=Fda1HKl_FXD{ED+7Z9v%l$1Hns-&DOjd}hfxK0!_d|Pr0$n5UmONxxe$6s
zu6A_^f~W$DP|V0*SBBlsGW-;TSp_^D@H&NseMu5<*R?sH3lSOufw#=#*bdOYrK5jy
z?x&a%<b5DwAlMA59H7>?fdsiL7|&yGwE06W3&CxI`qm6QcHoB(eP99*yM<lz@5gas
z@A`P{w$Xt8pUhEQW`mTm^0k$`<J4mI$A_?Vz<uZ1FLh(wy443%nk*uABgrs8CK>(v
z)FSr+<C1vTeu(#(h@9~SUY30F5J273P>>R=-^9r!@<PLaA;sW7Kf|b1AsJTKaIzYt
z8=Vez4Z+b9;OhyCh>0=z&o}z`_>6x`MD^DLRxPR=5cn~$u^G2xhu0$0F$@1DbzATl
z6;>0PVGirRzt))prW$w*L#cpGX!&hidO8(UvDB#RKl3Xu2yM{*y&j>WgalF`@DJ;{
z%Yq{xVd=nBqoc3~s%^<ItHyJu{~RkEKn(*Y+R1DXGeW)S?Nvaq>|u8B3k%>Pn5?(8
zwFRU?wBX6|BWOR+6|DdDFsZ`qXxP};Fn}@#8$8{C9`tK&M=}h-PkCRQA;S+>1`zaD
z*#1uv>umjIKcligDsl=b8#Mrk;B%`0foN7(fKyVV>1ME8(<UDz;ErIj(GwQ-0i<Ro
zBC=bKma*ntfmpC$7e1RJWmEaSf4_WL>O3{`BGQ$lwUK@0*896sKEjY&w_k_~i?ZM$
zMN-~X&y~8PrjP2WvL5v1_t0C~^+3Zrm>drHRD;-x*!Ov*?tD>1zVZCIBwH-c7Debw
zN#r*+i|;$ns|%|qd6xJa_r`0tTGr>emJ+f6h;;xdp_diVWWOyq*$o2zx%}A{h*F|O
zy5s>;xxQF~a@2f~@B4iNAVa;ug9nfTq&#--K^Y3XcDPNTG+Q1jATex@VC4(^=Qb|)
zl}J*wDlAd~?N9aiF#)g9bkiG1X97<+a1XN2{qw#6f7T6xBe2qX)1l$xi&;fQ%o}5s
zJzPw*v^V{mpMmUnyxYOt75Iqq??wJA0?ow8=rI^Xq!i$+U7xIv2ivCv<ZcfSkKK>5
zQc|yggA0O1#P%K#@a&Wy{%H)IEdz-coS2eJ7L)}x$AIAA3_t_^-awUKn44?PD{5#j
z)^E0{)|QE7YxyYow?WF;H!`L^;4@@Cx9&a08$GJ)m1eHLAq)&{M4_U8uOBF}{gB|G
zAe0CE{B^%LfSv%MqN0kq1O593QU1?#*a|{ILJ+Lxs`XKjAo9p0{`JcUI_zGcZXgu{
zg^_ZYikcd!x6al6|I~Y%X<-Ne09x}5ysaOJ!nOW2E_NI-{g93EYIZPd>l3xnzkW%A
zQp$H=<%!KYh!HN^-_T>9n%=#*xWFRg9sG~StCQu)spMi%E2Ls$3j=lRXr*<?gjJU5
zKqfaS4#e$XRm{VK9~?2OcAWCq!~XW|n;O5hswx2pSOy@huTM2nG_%~q#7qOHj2t${
zV}a=gU)>ovWdWDo44@TMS{iioEr>FGATT{w`|7MFYrFcdM^Egk3X2R3Tn2^O{?=47
z@TBNDXiG}o+*|B=L{GmOxFjVjYxpDJP6EGcNz)u>d@x8YwiAxS3nPu*mtILOvmcNN
zEdca=XkE91HaR&NHHeqBsmSZR*$JQW#%%&2Ia{l8(=b`&s!67)_E=aOzxz>c0>hJ6
zv|MEu>$wTiqWmFiOFG<-y<a<GRR^w@ug$WbUFLs!F&pD{{Bof^Qlr9xXsptD2@v2k
zprWE=VY$uB%nb5U2Z$g&l~Z0FsUrS?u(KN=)G%|q?oxs6Ww1Nn)|o7njEYIH4eG6d
zLLEM==4oKS>`|Jt;l3~6_S8?mH&2b(s5>E6YXD?yr>)7ZtYj}U=)VsSmA}?zvejVr
z9l(bhf+P#Fncw=l1%O+vYI>exKvLT|IAA;^Wl+tBOz*T(WC0lbd&9|slLXNH{1-=a
zM<*wCXGgYdY;1Psszti8W@hw20elH!;f)(N=7;KlLIq*~@Np#=KcsPn;bqDFGc|T#
zNg<>%z)I+Yt!R|cAPiEC;mmhm7gyH>KBrBdSf^St3X0_MbC5leQ{<BckIl}`&rM;u
z_z<8gIk^AllYliY9!&hA(aUpnxJYlhlwqQU1xry$39&E%&=Qr9SmKn#Zk~?T^Fm4i
zs~2@h&-?rUg5Yhf3h3k}e7!Euok0?b+lov~3~p@nPIdfemu9@7%WIMl6I)>NhzJdR
z2Er}T0ABloT_3QMYBQ;AHh^`&_Oi)md;`$ikh;1$49goy=KSOWZi59JEjhKlpwYJR
zYI{t60s?}cAUk3f;QtMQ)kCp%sLQ}}6@Mzo#MJ$fj%lT+OWudaesOz1UTDE8*`2`T
zcy|F)K;q@g2M~~5oy`CJ!RPsv3L%#L>9;TmnbJW(M67ZU@)SyqWZ5(6KTlX0gAoOC
zZCH4?MvVg<Ft8bS#;|&Tod4d>Z*ekCua{kfd43~tPcOXIEKtcg6=+}KiHV8U-WC>5
z8bQ5B-NX2I9PAj`Gmt={Q3XlM6_hEii?PPpDtQE;T1WwzIof0Mb*0rrp(al41URx2
zxjfY<1mwUc8YS<6@BU<e<y~?z1q38n-=SnJ9RVt;8%>so(NO>&!Z_P<04z6%P57EL
zf7??e25ZD}vW^GG|4~l^s6p9mXMac!#g&`%Z&Xfohfz!auC%7Gxj5Vb))G)-Z66$T
zso57kbJ<)~`On4H%?>ipW`*ae)0ea~=Fyjqyorz-=;$cZb8{jf%)I$dcn4^8>=1A`
zz~R3RFEK)|=O+Ec;HL8dB>V(98j$6W350kNBMTV^j+W|EV8O)zL0hI_2XMK-UV;W<
zF{l9N4&!DA|H%=OI7-ARAcm-&)!YR>^e+Huyng+9yPI#z1_U&$<|Y;_xg`F!A0O_G
zfWCmhFc+m|)mE_9n~DNHm9JSIc)Yg+b~g*wyJmmfz4g)_<ZYlfENBG$(gEXm@9&BV
zCFN|+_qVDUX$z+z67+D{{4Pw0ar5R&!0IBfd(TC(_eCs-1vO1^pa=q+#<$C;G+oBn
zF!!$)Ad}lMy1&}qxZ(1`9Ri|(1ehA;Q&KQ5c9yJI?|}QD6adjA&qIxMCq7Uwro+C<
z7?h4=D`sE-K?c|*W&<18Kg8%SbgsTm67W#z#wTEnAkp-8Qt75%1Z1h6I#`Ne)j+0m
zvXbaw$xr~z(*aa3GaV2oT#v9ROm?1$ISYlc&eN9+7Y6lgUzYNUe=u-E*$0F3<Hrx+
z!y%oWo8xg@#S;`11V?8B0EE0tMkWi+d|+tLFWDJ_aRtHi9wA{XsNRdmHU<l{q5;O}
z2X)GLBn-2i+kEH=fEe;1U;&7q2|&wJ>&d#_%NKTt{J|uStpG-|2SM)oU7SB0D$<EI
zZj^@!2i>iHgVY@!TSY2oWi=GhmV6!!&a0~@N{M1C;MyH+>-!0v&|)rjH5*kLoE`t-
zOwLtFE#nl%Mp?>Hs$yf$A-VI-n8I5IR3rs*)V>NRW&1Bs%3fx22!p^>51(Y%Uu|i`
zz`%Iboxn2=5Vw(NC~`IuJtgHV52%lRk&*Fz))*2ye8-Vuh`P;)Z?3PaS%E|%KYK7G
zv@IAVyB?+voU;aw^gA|YK5l-Xoa(2i4@}FFpCdhC8Sg`S+pbR5A)w9ld-tAhvgccK
zI8&y4yf`onfzF`*?Tp%CZx<yZfdGTQ+vZ^x*OZL1G9EbYPsPU<4-$A3lTJH`U?4tW
zfivJ-z$o#no>~|Y)&qglAlTjg@WG=&zdgNS5u$;Hi%6MCP7{?oOI-`o$9e&q_+3fK
zaD}h0#PY&FNYWYINAo;|)2vvab%zN0n!yS_&FebxjgMn3q2K!&OxE1=?49XQJ`vcU
zrM$5LF0@KsCsSVyB&_!K-k_5|;Ngh{3vmxP_*u*bX@RK2a{-)G#l}74UZ)#eo7GF?
z5Wt<c0iC5QfC?<#*-*jreWg$2f^YYZYm$b~&hjA42|dFL%2F7>qEyBk91vj_M96H^
zT;up!=%8$gL+G-741I2d&;|`{8-Q~4L~&967?5Uwl!3B6{SJ2*EM&#h=f2Nd{d|4D
zzWjy^Vhhp{_(kWR_KbQG-xZj;-oz~C4Lry_nK9Eyim5IwkfWr`7zeJLokCSItNN~0
z+96D=?;|Tri=FZQYdyv(fdN(Vpc%jc;rM{Eq(E}cl2dDwc?XaztdGJFh*M(pe9ubz
zKx*f?M0B3+19$j(Ij=A1ey!7`j8nZ^hI49K2mXuXfh@Vugaix6yLWaJGM@?$^!Gz<
zg6hdm_F$E#5)}wM6h@+V!8+fO1PO9_=^-XLzHf43ey2>HQh+2~bZF$Q1~1iCTAnle
z&F8;<{VL`T2ymfSYQU~EgJJk|b{yHoCMIGF1E<~f17C-21cqJ)Bo=y8AU;E&_B{#;
zTw!n&>3R<S^%j#xuk&oBEF9osG6LHnn$&y!H!E<mfXpidgqx?=4|)Z!T^?Ao%f&@T
zB8qJjko|#z@CXGB>m{&z4d$tZ0PEJw#)d3_?Epr40PV7Fx1Hg&q2UhDeF&RbxPkiL
zV6U31-1adW(Y&2p3CiY__WsTm$J<|QOO^1f<^W5ct&Y>CvP{(B;NMZ~1k)R9Gkzt<
zt@oP>l>-FOw3Hsw-D`Jw_KvNW^TrqYw;@TE^pqJ9`ohUcHmmc;o>N}9UKgk#Q}vkQ
zI6{pAws)VJ#CU6WVjaC!>^@~RNOc^9#p;QszX!=v6S$s8aU?bmL8OZTWgPXI9}b!3
zGHWwSp%f<~epd&qcN}%A^ukN+4+?`4T3tW5%JEDm*Um(<yTWN|Urw|o45RCAi;I{j
z3-2k5&TB@y-2Oz-QENT$*aTLy9VavlgG_e~Pwz2*bNYrW#x!?U734Xt*LFrpNgd<G
zNeG$<4Vz0xL8F<s^itjR9b;J3K6mZbpyB_Da+uWSjdLf|d;%V{w-wKxw|sz(+qBBP
zwn%t~PG6=Pd4K$^_|*tv2U)I69%D@}BtM*pE!CN%S)%W8d$!b?;P_q3-M$dnH-SMc
zivzW`t6oD)v~$+(+ZoSN<;G!xXFS(4O=56T$Yn#Xeve>$h}gp;A2BpHSTm4Dfmp>Z
z$AD-cW29U8ccPQm=@i#<!>Kw0<hqv*p39ZD<cyd2EBB+>d$)Hl9th1yR%5&oEBv_j
z1x~J)BoiHU@6)<8E?#9jQ}?5BuSskoBAzX+ag4~kCg-Vniy}QtvaLFo`bKX#Pfg+<
zo$rz0wa#95Lhpj*nvGayzYMh!94^jvBKNq2dUh(ly(gJ@t=jFBJ!L36$A(2;P;-hQ
z5{Nf7*m+O6PHO4oWSRA6(TGiC_hU8LQ6gHc&33a!4Vw#3IJrXR`Z3yk=SsAd@Fa`~
zhvIE8$#kiF%9y}y4<d~~zQzE_Tj|$L;^lXU#nDfQ=&QTTw1$)U4ii>D`hO}dO-+&8
z?O4Fh>~z6cKr0???Sb#!h=LOw*1JF8A)?n!_uX7TO-t(1XqmA2)=1N|cj+*NnBdRE
zDTv`yEDkCd^!lnY8KkJ!do0J(ZLLN(z1ows^LeJT(OW@wS*w28{zE{7@~vAuveG`8
zT@P4TmOPcU)Xy6e;I?@w5<Nl5S$6o#&J>+77+<HFJ#mE#g3e4F%a@pcgyfi#%J6uK
zl}F#5ofBzpf+%$}QVt7p2#bLP@WZVTv%(WYeUNrc#YU-|yJz&?9XjaZx#3Rgra*V&
zR{wzqirvuu7mf$$*lj#|Mrd^hyl1Bd@NPYyq%r4jefQLq-Ra<}xxdaxC}nvn1GG;E
zAqbot`WD|9|7cf`B~Rm9tDmLs`HCytb*yO^N5k9cxxAss??`!OO1mx|cY>h{8?V}9
z#-%APE4z#HnP#4zqkN|i3{&UlGi@)n7tM>!jl1bLHxm$R?gff6B5M^2o{D`M!KkXa
zm-0g6Xr%`>QkmpO(!MCrP~371k5`;N)=|UjLc=PY+z2!iEA(qhC!4m@c6)he5clRw
zNt4JtJ0|&(EekZ;#}k2`Eu8s1wG;K2zcG(lbiQJ3lDMqcJJsCFl<78AZ#dv}Sf8If
z788}ac|Bg~ua^}UJWw2dQrsrPK+-*KC2FEUFLEJ{@A^3}&a7)&FKJsaeqqf<mZP?y
z{bJ!A(Im@kZbC=6hv&X}o4HSSQmpgXLc1&e>mix@+rl$Kw220#%|U7UKWw<Yu#CEG
zwh)WrM~hcHV%y)bO;*LXvr0t&W%JEoYzF?Z&VWsa#Y^ZcE~!x$1d_-wWWEYQ&$AUg
z?rrnFpPV&!;x1iL>t9HU40Iy;4ADUV5lTwbwUU|H)-k`P(u2ZD4?!oF`iJ~AjK-(a
z8ZInI-;q<E-6rjM%jm89)p0X6bp<sa158Z;Pi50>;c_z;N7Zj?YB*#!PT>o6mtJ?U
zC|;E937wPf*Z=x1X4Jz>`HJ;#D;=Cn7j|*QCvKv*&rZc}UUY2jHzxZowDxcboo{U0
z8kp!_tRg<(8#~b5*!a6(!F$dA;{5Hg@6;EsRC<*Zc_BpH`Wmx>t<cQ!yT*|Ld<l{u
zNL91o+|U=pjy;CJ?ikj0pj<1xBB*|qQ&Qf{<nmoweHOHPdpXj+W$D0BU!A+t1dRhF
z#VQ2`6$AKYh4$>?^mv2Q5|d(V^>A;Co&6=;@Ch2%{KM^oax1SpcQ9k!FwX!MU7oDx
z1KG&;6A?Fy@j_KByD>ngmPVwAd8m#F6udpQ+dqe)#8qr^Kr!L6!Yb%gm8RI8$AV%u
zUk8??p-S%ec8WmG3hnB*t(+h0_Ib|E_1iG)5s`RCrWV-&&?8Sx`>86ASJu{~LHgVa
z#>Bw*K4a4h3etwNUFp)<tIT=z7lN6Ji;acGb6P@{K`l}1ytTMKQkt*+6&Y05{08A+
zVF<Vf0glFp92~G~_J*8}ks2SzP}?mfAV6Dy&$+3mMnGFMJ)Px$l;uh?Ut!I=7eR{l
z8$)YcM0DQlWX+yKAaQ5u74+l`|4tC7@PJK(7Xol3ZP|`%YP7D~EuWZ}``7H0OCb)S
z>!-{tsFOm&D^vt5F;#=}X6v-B;!(GLxW{u@X`x|m9}v(Ao{^Z45M9{C#U<4Ixrhjo
zq@*OvAFxydtRy>=p*T#<0BW%}owg2?H55d`prYwz=HBRMJACO$6}-^4VtoN^HQci^
z1U@5g2U!78t@_y>>y-@*3=}zC&0WX=I2#)7Nkv9LUJfwbtB_F72f_LwSf#umo;o=@
zp`jdYWvSA9os^G%d9JxU*}O_cIr|Mzd==)yI6ehhl?2s~>F93xfP>ocSM=SFN5F5x
z3b<&CaGdXOD3AANcVZNEThdd1^yP^xF2uJi!9IYF<|(533Ese5{5CKDZKaGil!LRS
zS(^EI8cFJ!fNIvAJGhIMKM@w*Utnfe#wwWsJ^GM^Wjl<+;O9rA6zi#mKS)nK{IM+M
zBldSIbD47o6SfHj9nVyRilb*+Xi3rXW&QjNw>6{R+Dl--_p73rGd)tpK8yYkIKlB@
zU)hEc7TfWR_uVlYLbg`Az+xs!U^<$@s7hu1ad}h7>X)fk!v%qht45+60wpKc0fpx=
zVCKNt;M+t*ZG?Ir&p;WWi|R2A%(CwZbzR;7izlhKw>JPlX5v(cjS;f2w$?FNl+|A;
zulLiNC5fWp<c)<y1?=niod0W0@rmxaQt~-#@6-6WWXf4VP7<thzA}qj?3v&4i9Uzw
zJJQ3|$<VQhzbo|!FBO`31qLjDad&ffF9qiiz&BMvWn|L9`3OK?#c^5&HqSs(fVlXh
zwKY5@CPoVm%ADK4l>Db7Mx`TFK)$F3e{6XZsc2~Qc3n5NC=$B0&>QTOzIU*f7+GIv
zIR6rBP0z|wN)yKh?C^MmlxR96*y{aP>$fv+2bFEA^W|O_mJ%uC++8kJ6P3C>P*`8|
zA`K1f)-><Sizj*wq|qIK{Pft|skUEQY{0#BOMkM?Ra8tY7GFa2haIqN8w1`af<f&Q
zkQB7YTJZo0Fqp3qUSimxNhT)B^u41a2nafQcG?P>ED7|n-c4wJ5rB4~S8ct0hmJR%
z>}`ie;&+dQ0-OhwVaYkT6gk?}YzSTzXr_PydIxIm&!0aF1KbS04x!8p76|x=z8obQ
z4^PjokCc?JZbf&r0A-1Y%ZeYA)y2G@v;dW7We$rY)N@G!C$iQ5Qq24bpjrVqesNL}
zWb0?Z5_zA`=?Q3t!+;k6tQBy^Wq<X6E9%K{Qe1GY&FSYXZbuWEHw(FU6l<$tJ2Erv
z-INQztPTrL--4q9yT1)WSP_n0dwE2978QMBAcax}lq~@1k$BXER&e$1?1+kxdwgDC
zQ0trNa~-Uy_J`SyXPz98eXzKs9gC`K_NNe(HZh?^n22FtukZ3hWTcCGMK5_Fc7`r*
z{6oN53n)(lVmX4ePUZKr7mb<XgGsi${C?p)ay?#PCFOPawJM58cc5-e1?vcyxNEDo
z=4NL*W1fqqi^lW2vMuE4O8nR$)Jv@@Gq>xJuEKf;R6WGZTwVMK(j9OE#Xx+9fG@)6
zv2Hv!`UwuNQxBbB3n`ZCOIUC!2^tj<T^Q28psS+;BBH7UVi$l5*RY*k7?5I8E-u`l
z3S{cx-@^jdR^TGOQ)W3v;O%`GG~~24)P)GuR*v9{Nxl;+mjHrfnnd}-d^k<Pr-#=)
zM?^6bz;xSkle!+tCUA!W3Qf1<aC~)#l5&`z5fa^Ogzi}ube$^1_)jNguG<HH)4>4X
ze<~m_@p0f^!pr}`-CIUg)qd~78|(rRDJeyekXArR1EfP5B&4OgQ&B+amKN!j?ohhB
zTe_P~^UjSvzxcoJJI48P&WE#yV?57b@3q#w?woU8*Sw|@?pXjUF;bQ{URp@}sYuJC
zNj{SCnZYMC{f1$sj*A^)xMf}B*7hCoE$GITF(<q-IbG(_wYJD)9BJ<&E}fG==bKwf
zQNJkj1}~@HeD%)NYQ$tNFeqy*xdA+dvrl@WSOQqc6K3r6T=U}lF~J76Q^sa2j<iPh
z&GGq?#kmS%dNd9Px2iRVxH-K`=nn9nX8k?ug14_j)TWsB^iUH)kW!!iPd7-CC#s^M
z{6}CkgmGBY6S5ji0FnlNh#LeufeT3p`Ep{Q4$ul$n6=_>hFJ<H4lk=*KOyI6=dPY1
z-!G(dIGi;s%E?7M`Jsz(WB5ic+4nyWZO~%MM}z2}BHW@4x_5J#$MpAk;IvFf7*|XJ
z+57=SU_eFj%li6vo^vpxGr1$^PQB4sOY!baio$%ww4x=(tV(&wF+wc*rw69LeC<Oh
z=W5@8$iMxnm5B)>F}uax$B%;^>9oOut%Ae^H2HEn765Z*Ad2@L5DRU-TUh?;>zljn
zj;e{0E6u5*B5ue~rI}}B>LvNCcTdCh&L%|vZ5TgL14wvNXoLWeF4&j*z$2VRf0l60
z<cT~y5vU+f;JX9}=0dV8K1?(RAP{oa;`#33+R<+M@auLW9oZ7ms!jaOY}h#DL>flr
za!m`H+4&VO8XO+$QDS4|65qJk^d{;B>RsY(VCB@+)R@2&h4kiMzdl{d9=@ic-&kMo
z4tP{HXjMSh2|=J4LK~B?#2(!b@n~R~RDYOLswCDuccOo;k5LI9uVpZP_+8T{Rz?>0
zzrEh8E_BWUBFYFaQl5&AS)J&cV`C0Lp;ZFP5X1*jF|+=lryD;^oyQz?a?=~lSd`My
z2s`Z<*^oB9-n!5>p_mZmel`5TP!jp`W86RGW;mgX62f570t4wBDkM$T-;=3K=mMc@
zlr8dgc3a^|NVs)O8veE7h{CH58||e!t+zkCB<v{KGOzUo6zX>@degfNIaoc`V2&(W
z%#P6O=HeWu?9XN178Mjk#cl*rW@ZAB#f|iGtX2#$Wv<uH)em-m>WkFKwP)H3tW@$>
z$A7}yJ+L{tWbo#Q0m0g88%?<CKyykI+Nj-Z`&pnd!8r~!lz$@Qo!>7shce7T4%^31
zpKwxUAvlVFS3?1V@jaUiQcv5h4jOqjGrFQ}9}rZV%otsgrP|9K)Xr5cI!Ah%^LX{h
z!06oVyHggo-s9-qeu#7l>6koFhp`jc9da7TYhS#5Hf?In7cN<-yayrz`ueYH3~%xI
zN@q5Yk>;{2t6d46D4bG=F2?2Uhm)?35yq(v&i0MRN@7lt$0Dl*g{n8ve#$VI`<U4l
zL=A??eU(Mc3-A=nQSk(b;!S*}jGh2=6m^)vM1Bq7H1v9qwuFUQ5BVh+RiH$J3LTlX
z?k&+QUq~?*RvqHpmWg&O+_nvsAw{0M%cOR%n*=i!cK__b@~3&`1PTk3=Jl*=``RZD
zxfp)deM0^W--(p`v$Of9HN%!Tn4`77TDmshlY$LA0kx(e>Jbp`KrIMB6sP9q_$4K8
z@f@w61AvGHV*dzE<cN961r=V-v07O*_edQFmE<-$M?uk{`b1j>^oa(6nTc83{W_M#
z?eqTDUPVU|tb|s^s!D?j&e1$8o9%yhV|6rGeF#tl(S)N1P-IuYVnCZk*ucObc1jhJ
zeO2sNSWt$0kezpF#k0{6<c9eOCDB;17j`9ui<(4*-ThR4t8dST-LGUSo_=Xp&3JBq
zGLU`V<)g96jE{3mv0~@nJyb+P@!V~cG7i8gOd569xli^qVAlwpvjKx*|8V@YEzXx4
z4ws^^*D!YJRc5{Tn}fZ@31uPzPg=jLy3hA}Dq&0COBdK0->*T|eg-=Z^-;e9xI4nh
z$?5u?#~uJ~tKnY+cG;6CG0-Dg-`aX)Z}+d`W8-bDo{JMC^aAok;<PtOwo*7~5Cpi?
z6R9sAO4vDV(M^sjiGTNP`MJ#yw2CT}JQRuTG!`&}O)5P+lFYPpvR4iZGkjh%<0~?$
z{FylEQt21p0aA&!2YHx+qN25vCKDFyro~E#{X4KSqT}^yR%H39u!zERZ4NIg692(v
zc{!!C?|cK|2Et-(ENWunE!a34ySsG@7PFBfM)lgc%6-DT@sE>Dt4HyjPw=N3VeH6H
zL!RyH`TzS+mPx-j@5xXBjzvRn6D&c!p5g>d!NWc?gHyY-IC;JH@b*R_9=?@X=Un!`
zXQoZEI^73Usxk0JY5;y1sC4E5ZXn`1J|n^f;_G)x%jLObA>`I&Nw=oMzx?C`*@$50
z$^G)}OMbe~v=!Vv+q6%r$7n3VZ(Uoq{%`+cLQ|+*fX)CN5vP#5<vfCM=P^(kenQw)
zZlv4OLn&}sg?nVfx0u!`iTIjIaacr&ccN9U%6XZ~u98PZnAIkx2lFJmnPI<n6Q+{!
z`@Z(Ma_6&${233MAL?w+t}z2?b>x#DFE6KzN$=%V0^;^Mg46ms1@RoyS?qyj75FF1
z|LD_0j)%zd&_YuGmDA_l^ZogVv-q@!hdh`b^SSNZ`gOW|yqQnI*y~*N56hcB<mPw&
zeE+!T-u3Hg8ewEJ>-1veC%l11aaRmGFvhDkW%*G7Y4|WDp&iJ%sr@h<W({Fb6?^dl
zo!^J(9tFjxoh5VJi(yO}7Xg!s12%;=SWVd?Z}%5_bUWJ@@&8O!GLrh9i}U~Oh@0F+
zzzIsu&OX%g<nOOw&sGz8^WE`(yE}dMyfdyxTQpY$Ec|ZJB?7jXZpSw?EBl{P`+Ixq
zKyPL?7+_SQ$AeAjZ=nBdL~QQD#sSw{di(3+l~1hP3_$p1K)Z>Jt?|cQIqR>DXAt2B
z`xJ2*6O+MA6s%takZJySeP=(w!qgNx*VEpPiKu3zFJ<N8I@bGVKEd~xT3}=Y`wL+J
z38f7zf`Wnw0K`WAi^>6qIR?>_n=Sx$<L-ZHgPmo%@9z?Z|0vyn1^_NE!pnOeaA*pk
zdw}ZFqy_smB6|@3B5_!M4(9Mj2=4>2hULG*`%+?yE3M$meD?ApXJSIpJv9o4b%*k(
zG2G9&fUFx9DJdjLs~e57jXS(FG5PV|+!(V9RL_9%j>3Y$QbJVCn8EL3>>VJ>vbYtM
zt_rzNKmdtTF#WTKpw6FjDA4_B@IyImI1GR^<_g@lA3*cDb?a6`OA9$CXCzPrI<r)H
z*ew@7W@TlecFMmWxOND^l>{D%7coZ&0E#8x=KSkZRr6-z?=}p-PuB*_62!Sg!1xA^
z%4<NrfILxAQ30JB33td=tG;->nvuipU;Suvx7+Bwb+J+zh@GK;jC~*A`vFo=1N0Uf
z;PxfUWs!go6hwb&(*I)`zZcPW;oj+y0(*$tx)i9VBc}Dl(pf6Ruqyxp28oDTrzgiK
z?he?cVL+zo1b8?1Kln@T5P~lOyKiK6ZsMn_Ex;(?E}Fg9INh7VjJBO6fXBZAUXy^3
z5RFCpzgpWc!wA4Sz-Q+tKyYdn=ya9uUgiR%EE2^Y3<P6*7fha`j{iA}=oNC#yFka`
z7SBS)58^H42mQ~9it02(sO3X@;OZcNBJ%%Q4{W0r*a%?O@?|!*wOs^a)c=^P3g=U-
z*Tz8q?>&gScK`>ItQVa9a~+W1jWXwya+xIq2MV2NF8$UC%rmA#^xj`%)M1W@axv>o
z?)yfZZ@l2@t4d%#3!5`w!}5;k+N2k&WP34F_am5tCpBEd(M;_5-bv#cB#PW8liVcc
zL^Ox~iX8NFx_ES)tvucklybckuO}Ie&e-+Mn-8r)ONqRh4JT66Ed%@40t*f1kUftP
zL!bH|Ytb`Ac(0pjEl5*F8tp7jX!KWp(np!0(0gBjHWCJjKrs+RkpO0pr;oUpSR%X@
z1ToHXAW$!2f8A9%?c}9HOE>r@$r8nOtFhs+KNCe3%-EkE@T2pjAAV`wQi<Q0IK5@1
zfhRS4%U{1`!N6c@S0i1tuF~P|-0D_}Qsyn3hjL6>LiK|p7#Tv-50cvI{^0^Fk5mtY
zht}K7vQw>TG%0}tD=CXBf{!^)Uqi`wGqn0W_T{TrziTH5C5nVa1vhq*viX&~CE2}r
z!}NGd>+++dpX8F#gY<ZDQ+*pheR(G5rd+ec+mP=OJgQAVYi3TmKirt=`dMTw$JH&E
zW|Oawqc}yVlXQoQLKu#0owi77lgd$}IXFsCQ4<u54>7S&VAe=L{!UrBd16aL1InAD
zRm}6yiMDkEy51&`kd<;2Xvlt37dRl<Bj$V)Un1{~@NC|o`S`KKaCO{yv^JiNRBO;3
z+2OQj@GIg~si6@T_4_vVY@YIKYg1w+Wn3TYxQwb-XQI>YDv9B7;MQ)fG^<Y8nBiq^
z^uPQ;T{-FL5kyOmSJ27uc9=-qeJ(nw_S?O^y$JK#W6(Bvdy1K+u>Q+D-W%U%X*}?o
z+f&g8{=!Tyv7RBfa`iD8M*4k|>`uLN-lMIC4KvD@YEV{>o%&&P&*r2%i-$|gU~el*
zryFPD&`5uS1MWfvgby&MAG7m&;Xjv`#|Op(1Y%$t0gXh^x^PD&&cMMoVO9*_Xs^sZ
zCItlCId+#`u`<|vo_{X>$gD-Pkd+7R*COv`6My!|$Rd58h^?Pj4|!o50~tk&&gTIP
zEhf#y?#BzgLx!#5h|<s~!ZqEWFLR@-m*TOx+2;$2`FLqr>~ZVrcD9e_H57T6M^L{5
zn8Nb4p^eShiw_^rvn%)S;JXda;_auU3|ljq&oMZ#FP`E*nnx>S;kh?A{$h#n&PKKS
zR5F2L@Xztu-zLQ)td&Y!b}sM1+s|{{s7#hD$3XHNlxPtEe~MtEOXmsk!m<iWV-69S
zdUlY3foFinoxU(V@ANn6IcDXHd8C8HvP7<NVIea&u}rUpUOQ~jeY}b#>gp3k>-joS
zlA@&I=xQeMu-X!Cvzf;nvx_yW!EEMiE51dN<tc8{g$qmzpX&!4Im3*RCwdAP#b&i7
z&hsO`_y8c@!FTiIv8ub}czQZHW4MyLN+X*o)-7voZc=|Md}J#iL4mi1><Tr$m}|uM
zj7q7b#&JdLL)q&`LFk2wz!9k7wMR|}iOnXefFXG02D{ZSh9aUc!v1Pe(R82tB(sIx
zDMt=uVV%C5nwGZ}SxTZyVq<Goc?2wum*D%*6!{w{Su52HA~Legq?SjGPnf$dr(x?_
z-bSRl_v_3Rw4&mBX+*6x&HWLZYWHL*@BMf|aMW}mAS_7B@-`NhIq}o$&t~j)Xfe=P
zOQXJ6>Y4A&zfOzuxWxX|?2b!=Vm_@{fH{YQt0K$(%GVP*B~OcMWiSRBMUjXQ)ni*9
zUtd!Cq-wCF)Nw-Jj}H3)If(rVLZw2*FJrC)SQ~Rq`;ob!erv8{6;2Z^1?+A%cl%@7
zqjxqobDjud@Q{*no!h>X@CV^MJ1Bw_^<4fHL=s5(@2_eV&!5p)<EbPuD08QVNzr~B
z+H{(k@OxsTh*L>>D+5g08DGv-)elG1e_zXg;$IYjf;kf!+K2mc4NjZ~ce*jjiP>FL
zpWVz~VoTIId^coO9K-w*C*6mNjASuMf#=6!^(VyM#e)=x3aAH+I}f$YhCLGW8gED$
zcdln+^h>>QU)QEz`SEIJpo`fIT=v^4hQVj1H!BhS=JZ<D#pFwp02RjHHqqv=*-Cl$
zTh2LD$5~eP;;@raV*#n$=EAV=hOe{JuPCllqt`-M8`8S<k{%%IaTbT|jT{_oahSg6
zq>xS4ezM#m;-p*{qS~LKwedByc&@eZU7xWFA2r{|n!u)@N(}qxBEEXIWIqDp1!_M(
zTtJ;)aCQO9<N`#kg+Sl~#qod70&=>Vz!6GUE)4+)3#?0&FAJ>YpGFLOOew3W)M8N{
zHa?gtB#~MB3Vx?$yNp`RGSdYZ<s*JVyqv*m_(eHUksxx`o@z+zq;-crsd-Tz`gG{%
zwfL;Q6Yg}xT+ID%O17D6BegeDct_e9uM>$;Qc88pmu>Kq#Oz^-g;u!Rlt}g>MssyA
zA>jLq(QCvJ#MHDvy^dC;>?_2jfLoT#$pRWXbKv4PNacbd4<1n6cRo)WNR?gR-7Sa*
zni(*gFqQ{}Mt<uS4`Z?AO%z+tNoJc_wk0Mp>ug4}>>p)}Okvk?He{^Wzb&*^z>?$T
ztkS>XKG|Z0^6hFgh(^mj4X@448TA*XTydmyd@-0R)5vb_)_WS(fJu5xiQN03^Qv>b
zt@MZe!r`|d?)S>8wf<3tme;A(gjxYnB_QBDJ=y{SI;59_Xz|5sxU_FUP6$|(q%M#k
z><)PvGkw5Rh73P%HT><xUQfXCa)9f@eR^aL%BwYyX1fo%5?d`(nP*%-oxw6~)94kQ
zF`s+!LN{QSe*gUE2+!{uFH+rRNW^)}!>2XeM~J0RwxmdoPH!sbu;a-V*3Z~j*wc@i
zf?}(?+iwvm(jjfM74M`7@{dMJ^472?vH?NtEhNj70hEREzDrk2Ji)t}C5eGHgA-mG
zP07t2ot{q1ZnZ1`%$m#Buea<^@i6+G93OU<Sn4BywH5v*7ji#<5vvPvBS1pQ?%sth
z{1q5j43$UBKz#<u8W3(>{li(**|%LP89|~w*GeH$eAsRZoBfT}p_#!@TiF;wlsl79
zyB5JVQ-Qoj^Rjr+mh==bljPCO0?tDso%Rxn2oC|lw~OW7|35VSdG*!{9RPcv`&us#
z_yf7@2T(JC{d1!V<_BUPoSljdV3VUz3DlQB!vg{3)xJMHOmI|Fx=ngQgBTm*-uEGN
z``H#FJ2xmtotO9~=d*{-o#U-~+k!`u?VGZJ{<sMc%RXLp*oet<-lRq>e{vudIWm12
z?KqPvV=c^zWzA7)qZYja2k7O6&f)KmB_+RFT|haG@1RzP(N=eVe<*k(bA))eZ-1<w
zL<LVBfWIim38cA01n?c3`0Do$4$z;To?C~+QG^7f?x5F$a?^hji@}rr30&39a*!UP
zMs<iKnb}IRY%WRu{X0?|-<i}b$;^=)QG=13rXSnKoV+yLF=toTp6)3e9i7YYpaC#a
zyHJ7nImF!4@KqvUw@Mje3~1U}oLRMgnhOHdUbPx3vpuBF)2o$^111jW=;<Cbzpbp^
zq}U$B-FBsxa{*D~E{Jj?K+vrUQiqLU^nelx-0j~cPt58cK#<64S-#GPxETa(JbZjA
zzSjX&bc){2;p8at(u;Asr3^%I;i$c>?c~~;>f6@hx~a}M{!vA#$dYTQY%Lf^T$^-A
zNQqBO#8rGq)}PtwJMt5D#@XTV;|fZfw>5;g<Y#WZ^?~{`9@qd`$4`dcxOv@^V)F0$
zdw5ui=(++0{C~wSE>2Yd2N^P-aZnuzq6jEJq(R~Fm2VFFo1gkFUYu3BXL+!#TiA!<
znop~jm6e+o=P~D^4x*$Sd)q0qViYd$TQ*S%jUB{-5s=_1@osAC1*E#&LSds19=Q46
z&uV)F?A#qWl#JFbZ0sx7uECZyR#i_qXG`U#hIlF`ukxj^u<(sOiZBfvS{j+uAmdZW
zGHpeXCL7H?1s0gsJgo=Hwdjl;aNRRJV%kwlM$#^QU-MR_p^r{O%isYw98c8{jZ{)1
zLI48CmwDyJjSF}0-HRRAqNJqU>f|Ryd51u-1%G)4=#7otUOQV`=k8KaBy?;=o)M^P
z&e-J=0IUM-c7L}$g3T8)_g^gWP7m!a#?KT*4n}fYMxN4zHwA5o&yfrV|8QezRxdhk
ztH6B#mI4;mqZ^Uz#G&1Sc4o}OxGnd0z5C65DeCHTH9D<FI#x)oVFMS@mzoC|sXrD~
z`enl`Qh!nTN;TtG9%pV~muiFN-uCifAS82`f@?ZkdTVE7WeO4#0jk5Xser^Q1ZWX2
z{8_dL*x<^_9Frwd?r>f7QSX>Co_kuM*;fBa8f6;@==yCrc0}aJq;@Q>`F1Sq<82a+
zKQ>I|YvhIF2YLx0NCeP0^!@eQAdiyG*on{lj@wVj?)(Z+9MHC>AvA|#1QvVOT!Uze
z1my94Tj%@8YCLxR88j0$TiLNTXnoQ}`e?ly9ajc7G~Q1cEp<F<-#r=SDZUSx)*c>T
zX*nKx5_?uOJ053S3csHsTU(6OIb8QyS@obKlOzMK0`U_FBSB=9ZWAX8p!LR#;#)k&
z+t?65ngZd5*^I)m?=UGt(-(&FA0og{-}gyKOw0jY!lR=jpgZ~u!<3;Ui>?k^nTXOk
zoUTYL<4LVD#Z&QD&3vI{U313a4lK?guHj_oMB*GC=zsERIdo)9R-g2h1TvK&#9*u-
z4F+^yK~YOsQ<EGvoL>vHHocQTnnaOOnSmApSrP=W)^S>GGX<Z36cial87kYcth`(y
zUk?Wao$4n|-r`?$$0VH{)QPz;0TS9xxi|e${C`y2cz#^}In&NyQSfN;5_4DgruT4>
z-t*{{+1H~Y?PsCj*{vJb$$q@39a*W_BZZh}pz+A+zyRm)4u=IJ7i#BMYa$Z>v^rg>
zH~><GU<HBNHJ4Rs6a*1q%R$Wdrr`nJ3AtOdeU4kLY{0}<-E|O43WVS|VlIoD4#hYI
zS~<DjRwMxP4uL3h0?6!-Y&WMJG}g+t+BgqKWc~~&@v5Vrt?D%ARJTd7jin}Dw4FKo
zJsG`;h9=FG9DMG|iKY&EN78(<)U)xOnB{g<*xKNK)`t9<;(DobsrvJIcejA3;Mz96
z@mq(l#vZ;jwtee|^BNo;H6aM1Lnvc;xX>4lQ554HwjOAAaXl*a^pKmJ{3FQm;uI;`
zzCN-{*#4FIm~Pny4%R(LDG6iI$6e~rqGVwSMq!Nk`f@O5!IW;c#;`t1gW=?F|JlAi
zx{cY`)P#6DC!R)NXDuH)Wsm9T-|=H<1m_I=7U?_iS2Os^90qnH0^_8}-?T8F%EBQ6
z>wXG8$!ewz8Cl4q4795fkZuo07VO|0C>sW{bzmEL-_jxic^as68Gtt9z>t9rUuA<6
zV1G?WT#!>Dy>;sspev{hfzN4Y2BTZf%!{v2tnu10jA|RF+!@#lkR5mg>%Q4WLTBT?
zdQ(0Y{eDQsbY;1n-@zPu^(Wm&>f|diw0p&XC?^D`+~_#7i6GgEfJ4^V!qekxtkBMo
zL%|dOkCxf#%B&G95hwN8)t5GQwvxnLA3Qc34YN`v{bu%#E-miNL%7o9{%NCTXlL2e
zQ)BKoAJM>%=pdexs^NCuj-{+5mEqlyo?ObA)czmrFrLH32`Eb1?AGRH;^2z%J1cuU
zaC1DSC-J7+zdFV!fPom_lWT8_-FbpRHZ-ngdE5V1oWWbfpfDbMM+>33$3STZo`QdH
zuuS~Mq#ee;QZ+o3D1elYNy^;<N1}X=W}{Xw=zB<#K{Uv9vgP7L!`dM&(1g&J#j5>P
z@<?;|c&BNe+MP$p6;%JeOV-Ek#Hth9|A>G9m(pK9YSb$*8dAg|fm6k53oO68981>`
zZ`kMy!CgvvL>MvJjB>Vu^C>gL{M?%tm=_4s!5_=fAVh#-;2R=x_eaBMa9ZFFSRS26
z*q=PnGz}oa&|y_!jerD+t6w3IX}?xpztEj@<?7Yns;38dkkL>#{Kl6M2aEa(bMO+n
z$rK&M)S#%KJH}1<%<x_&-WnQl)k`2KvLb&~izYPpv5@wV0<(C+O932VAX9zKTtfi4
z!%91`R}gYdq$*D_b{5|d^KnX4E)lK|g?k5Gx%`9uD&i6b>0~1FbJ=>ohi`@C%pi(~
zK;`-X(iFk;37}4`LVKyW*OX2kQG8kV+SPK`@#?l8B@O0jGxd+2{Lmuctx;#YaPEhO
zDQ6+Z@>|^?Wnos0BCin<m`bv|jt0-4FYm1q>aiFb=?0Cn)&@E)<3HBb)5X{jv1|^F
zdu5`aeuBr3l!aOpIa_bZI4~Y;+RXWc%=$glXu81*R4Bv*XeO{sZNu%xc8Fspsu-@&
z`t18^5kR8cmZ>-lx?r_cHC^*PGTgM^?sPOCgGp)c&Pe2FF?IPQ{Me3q3SlpIMT#k!
zn^;)gdk5+kYd?LZDrY+iGD2z_gv94cL-{qb6s{TN7wnTqHyY(Keb?2{PCp`j=p^)m
zmS5J2^*(o`z`H@_qYSl=T2H5&Acqr#@>pFG<me!5*Zj+9S>$+_bh_KwTeEG>*nh=j
z4V_#^QhZ}A+gWF!ti5Q)cuiureRz|v<`IE!--G7nM(4wFzTb)EMlq)y1?P?m*WN^0
zYQE)^tzX&l^pIEYxS5R8hKC%lUmc(@Qe}LyKY>=^Fu@(LyIQ+5;yIC^<MM^&)nnrq
zX+IWueVsm~G|NqxZB>36_aUAdbFOF!U*Xu_Ud1?nFpuTiPbPIz*#|OYsH{Pd%qZTd
zBWy(xUwTRnC+ckugD2CjwC9>E+JxsV3)D9CqM17>39)ZK)DNjM(Jv(-H;{)?YOW#2
zJmfjm0#dWs{`vxZq)KbA0s!?6JS^kOC#=3wnnGWJ=A`XNzA@GRkXovk2o2}M-u+3r
zWtMZM<=9tZon}R-$RxY=_NVNuITf-hYP|SQ#2PGjl1IN^IMH(b8Sm-sq<nbJz0f#m
z*D&MB4=3V+O&rFP9TFu7zV7pW`BiVEqmpIBxi>0#v3!R<%VC4X&{$7Iz~qO*>Ft4{
zL%fqht%ug5U0yg5*i=(X)#)V4^U_HnPqDu&z#;2H^fUnY7ZIrFD=eqqb4d)TWMrYu
zM#$N|6H<w+-n&i;$J87JS(NuzoD*9^LykBQ_DIriPNn?116KH2G`*~b+*rkRJ?r%>
z*;whEZT_TUW){5{4Z`TV6U`HK0vs5xyy?BVuBog<{Lr;<mHrkMrc?!Io&A*BsG=C{
zo2JHm{fT9TTqpZqVL|7s-g}97srjl~tXzDrxXzc3x4a}QVj(9>wC9~=y{ulH*c$(}
zNML|8NeWGX$1bQd<IQ@9D4PUR)p*{1jRW8^)xUmyg%io~U^^CYjWS3ieQY=s1WYoO
z&4N8RpipGQ>2RD|V?o2ecyid?JOsqf<w$9Hi3OYXx9Y5xH;MvbI%hDpcco*|s^xHy
zn!#~bNKo?}drzjI`w&Zp4_W`+Zc-(`pr9I1Do~l;Sar($_huTTl1Teu8@}j5Mc>6x
z`GlwKV~+0pGBeIZry_#J+qf^*sJas5E?h7uQdhuPRv(smeS7DUes!;m9G`{ZTfXxO
zUq0L|xAv2`#r_^4=z>@XGjn!sVK)lF2rGURh+MLrH8px_&i#`zjAQ&MB>ups%66;q
zGD>DiC;$wQ9e^iN6>_z|czJ2c{38+MG!uwE4@x=Yf*{ve0p9$yCq3#?_)J7PK;X5g
zNIszF+w*g(W&cPOMdHQWeg^OsoggUtRM^=5`8DDR12?q`cnjqL@q@WKhregg{(Lb}
zwik~Ws)Q;rP=N;G2}!V^R{rxLIhVVDbFkBJ-}l|g(SE9GWrb1k4DJnl#%DkuzIyE%
z$CBPUz;Zi$X9AFLQ!PL7tKWcINpE`E4loa@q~u(0ALDgZ-hVV5F79hy{inI-(PO+2
zt!-_)watIOuDY-xi1zrwgE~BhGpQiBCEiEs^}p}&0{6l{s!{ir0urmhqi8SHI|L;S
z%b{bu#Y)w`dPxZSS_+(JS>?NmfKCb0yk0+AgRj-ECH=k6dgWJi%s>BsS5vv`a>w%W
z^0lGAe?gpB2wvy_+U0-zWjc_09e7zCpz`84b|wb~SGUauymE<d43<-&cCR*le(g@a
z;jpYxVFaJw-^T=nyU3rXzJLEEXrwRW|NY|&krXckWa<ZWFKioeK=UL)nMNx=y1Z`D
zsx&w(j9O3+@{@II{tgjBFhLs8+R(6$%XK=u_NjG?#caBv%j2)K1R_8v1;Gv#gNgwo
z8Lm7<D%}S`P)LI=SxilXY|6O*{^bGiV}b^ppgsg>=Rsa#a@W76ND&pc!iAG8=VSfP
zSh0Vb`t~BsCdhjiRV}0a`zIwo^YL#e`5M&c!|i_$hbQ{q^2jcJd}x0Q=00oq{zoVZ
zz#7fh{|I{D%%S>6y702X?SEuAKhj?Pc6L();y*|E|K8g!Yr|xXd?VIJPb{SHO-q&g
z+X3;abOO7fW0!BOx{UM-78%0<nb%i(9l9>L+?{(IZ}_w;$3VB~-oh9;^{?58ieI#n
zl0iM3VmWUdx-TKlpR1-zNEAww6Pn@<T_qu_rQAzoh!9Oxd0Fy3;WoiS1h<Q)rhrfz
z=l5d!?#Jqr%v_FI`ZB)WJA$gLo_zVgr+O%;UthtR<X$*;@)8CRePH9|aNc5Ry@koJ
z!&-&%!^+Op-tql-p6Zp5Y3h%C`pywUqX$0Uf2SS9Ga#PV`8}qT>f2Z_Bc(ggIK?~E
z%05VZqFZ!BFd$#4XvF$~P`^xYvMV~3zASyH<uZOTGU+&bEo~!5-tkdvw>=j=4jp6@
zCzP&+EUR%ZTr0XMyqmO?Vk;aJ88$PD+)drdQFJgjonYnczvpjNN}^|3{K{(K+dN~_
zyPsc9%@H@hVI<M_j35uw9BNbT9D4Bdx{G6e-=b;I%WeK0$7Wl-BHK6lnC{Rr;DO^w
ziIbgWcBZYi|8JW9EkS3*2dMJVLZS;r^LBI-e2_{Rcjy-3!lv13O!PJ~cML<p?$oSV
z-L0>^Iq|bi__8r%R4zGF?rRCYbyV5P_Lhgk&8fW=<;#;;IiL13?*gzHoX;h*L@Cx2
zZd?%%;2g+wKu1mXOUDs?Zc}EHm2111Gha_aHIUR+fGe};MW;X3)xyJ7_L>>1hnA;k
zj>k(%<;W>%IxZ85N0&w$-gVrq<$8v87l9M`B>e*EX!?Pu;zA5ITPkt~nXFWsnlo7K
zAbS)<?%A74!?HM^__KX}^vV(rmUicdY_`S~v%Pe2#PheZiI4BU>@TxPGdi4q-D@;1
zpyofxP3I&Oz(pn^n6b1Ze(E66Uv5?CrJ5U0in*UBj<}4!)F(vh-5Kn<)>%G&`pu2Y
zo|^1IUsA2OV9wx5YF~Qs!SRaLvF(VB0xPAk2G6=~Ox0jIPh<+ec6<K{m;Jl_OKXdD
zZBf$ntaWD>DP*9;SYLr2>TKn)bpjDbW^854mTt^CF)aL>E&)1~6xA~6lg6Tw79wTT
z1gQ`DdVh3cTUeYvSW%gYphhA-xODsNG@sKO%gSIMG0)d!8XD;n%`jRZYc5=neBYc0
z7BxekIFIZOu3Yk(@abIc!s=Ytw>Q>1M@v#&Ub<aBEu@$c_e0aAqo4|MW4E*a=xp4f
zPs~eqZPA_G`<S{yi((lZk-9<`mF}FiXgA~dIoH9l!puV%HWvQwkf`2ow)QK^9F09h
z2$8&n_>sZdMGI%u+`!~1M@P-+rwNvnEk5Xo3d?!Fkre7hi#5lbaOw4Z<-^iT_RO>@
z@ANQ&L<c|dg%`4k`i|L@lU27y**O7WQ<|Ig6;I+aBg-Xv4A@UEGx~f8Y~r$d=1pJO
zwM-Qd|A3BtetObSyB_g2q<lX|J|oph!4RHm=DK**6C<?i>e*BCbqdL;v$Y-ux(Z-r
zXQvMQh|M#O8GJ2q(U2#Eu25~hDN(e==etR%GU$9tH+351#!Zk(ATGEdrYvl@2l!vr
zhDX-iXJ(i6`OcZEfnXHQezPl3n$Eh#KuCw{`Rn`r_!otPuMt&+Upy%UCz+TYT-xyP
z`W$?9fAh+T+`NTQfcpLWqGG?X@Q4#QDVOe_zrU5GiH1OH{bkmSK)>P~fT%ikD7s8}
zQ)JKJcG>^;{$DzXI`VEFh!L?T((z>YlPgpkEHX~h1Y)Cs+_<{4Y>1OQ7(BPN3zt4e
z7%@J@I`W1yN56d?F8fm>;G<v9bZ|Jwj=ACR59zQ=NE$MlSC)73CETOgQtT@^a0pe$
zmOT(n$v1dtdw(I5gadxX-G{b7>79WeDzIRshlYCDohDp<#l^|w3revp49n3O5hW*o
zJfC|X-HFYpkEr};rx2ywEO7Ok)=ku4q0uM3jEkud)DJyUQg=2rjSc9hmDmiwaIoD<
zx!vFyliXf)Nkr(%xAdk~_Sjm|6HTk@h|9Y4NmAIPC91=7?F`xD3exdh4@30Zzs!U~
z)y_=9W4ECZD|bO^v{McT+BaE_96y)w83~XuRyWtj8tx_C<`LIfPQ%ozEh9blyZlT+
z8$`!3lFWP~d)ykEx1F-E%t}YJ>(O2OSS<&n&VF6D)vmjg9_A!){%bYkbsUbUrUvfJ
zja+A$sh`NknOXMubdB>=i1Wqe9IUz<>1$=fHyHG@=1!RSaA^7^o;&$C9jWIz5h2}>
zdza!?dxN=*gL?HI-KTe9p6Gpm**s$rMtly@<mDA)C=|`wTN-7<Cp>zvLHQ+3QKr_`
zR*~7tE5zw>pS{CSZgo$*xT=Do5+>O;FS7JGMpE}O{$)llc(uEQPJ`*UhTAKi3gV=Z
zH>qb+u|H6;=H4x#k<D1(ijJ|Ctk1zX;CKKt#6h$E*C&dCOWwx<8@;KG@KZ8Y+42~I
z%ssu{@_J7*^Wea+$+;lJTN2bVo|a_GSm*Zr@>yy|FJFkLP@c9VUDWs1;g_qM$5<%s
zI46_0$Lwoc9$)1tdIV@qN7S>E#oT=;cAs9odK7R2fsQ6Mfp)jyhV3S4iQ{~80C_3b
za?li1Ggxw?pxBj34Vex#Kr|(WN}kH7mog?jGtP4uiGK{!_@He4yB+sV9(kjk+tSZG
zmNcPd;bfI+hCzh$!>=smHM^h7IAzK<a09#UN}v|vT{*6lpSnn3mPc_q@donPiz>M9
z+9ENDn3b^nJ5Rn`LgdKPY4amo{A6RTqPkbDOCy_jLNaPKxH~qsE!IxxF)?N2rGsL+
z&}MTKzaCvjy?lU*U2-VQ$R}E-JI+p$)?B+KS3bi#Pl&6InMUlU%R4ol_#j8K#@es5
z-NQnBb!UEp2)f5ws%*se@DZ4pK{B!t%72X42{|y;zFpr(h8i!_Ub1hR_>PA_d=pXJ
zZLB}o+<mTuiB2U2Yw6i#yZtHK15?ZFy4I)3_naNmYO)OpLv;s71Rqeq9bI@vj#ti?
z=V=s1tK`c%zABUWEk87rKg7JH6BvNzatc<W%T3K}dCgq+6o-Vxl8sfhJ4c~x!bCC0
zQ<ZHUx1jN{Rpwmp<dlMG7PqTjmK^M3QTSvXuD&_DJp1x3#CaxuQ-4{8@YT^w$vB@o
ztj5+&w#4?BJ)vt{OQT~^IX@ZpQC;w+?YDYD+i`EHa;CN@MM2T;psn!Ta*!)N(O_zM
z`G)n7f)sHc9o369N@DI~ceoo$N6ReP%6A99OSL_^9#>=aV!w&I9pwy+n{yC{Z_i7t
zS*`5KEk=J%Djj{@xb*YuRUF62>j?Bqs2SwJ_o|+zLWyE+x6fNx=4*_7lWiKF*u(76
zf+a+Z4ytF?KJnL6l{j2jvl<#=p=oR9m?;Z-YA$5Jwm3RGH;DK<OZpGp?91D46ux2L
z858G!)J<WV*uW6J?_F2Uk<J_8RLH&9?*&a-kCJsY+Tms!F>>V4{2@ldP22d0&#AkI
zI(+LOjbbRTRR4$SBXpNAlpplmXu?ltXX_(UMoEcEaHvG>^J;10EJ1taOa-2FhPI<~
zEvk9GmQM3(8Fx-UMe^|l#5?NCE=BpQwA(lp)|1G68p=P$!9<AEUxqEw%=8Kli*^%h
zaF+CO<lLCq__dNLf4L*StI({ITgnw>BoU?0Ol49dw?>S7xiX{0Z*a`nB5n%dqg_SZ
z`zGYgeqbOQ&YJl(ZpNHaf2)MAi9Jtz63ahzxF|u<x<Z!gU_7CzS1*hhDwra;C1CKl
zpt_S4ol1H}uD0G<YB0NhG`&6V$5a;m#EYZ$i(lTK3zoRwxtxrcu!bpomv_LxDu1C%
z+^}T(5yw?T&`m7tl-H%)k2o$OK60kDW`d!UqSmgW&5wruO35N9va;<l$3>C4+y8jw
z_fRg$`Q-1v9CGg?Tr9e%`JJ){si7`{>AE$ZA9<mgK*04t^SgVfnT(JMvgZZe*x&ep
zfi^%tTE2{EdA#p8ctwbD#Nj-;#Koh>F9R&dj$Im`b6#ibcV(<^=RgVs$QlVVJ^kz+
z(f;cB*~7jT2|+Qq(Zl1xX>t-uU((mS7PhEEy4jChFd(k#s8hfw^iA#4Pm+<X%TzG4
z3rFt>UiJLK{M3VgEIUrpAaamN;^Js$87LG!_g;KJxzWJx^kU{dS&Pqm&u1uG-;0h@
z2uDvPu~i>0jF``DIKw+_Hs!?%D|Eu!5Y{UJ@uSU!VMM$_Z%SWgswed@_MJVZHC$LW
z&5BMz-Rf4<ldqY}T=wyGU}St?PS#Dh`UT@!_lPAS*c{6vN*UwXP@q7DiQ_6gI_lJX
zc7uI1uGceBw|=%aarP^lMj3Orb(&hJy=W_16|c9tRt90*g*EX&h&I!?4bSwIS)b79
z?@*;&i1R~yEc$p}f!(SvNBlJ512(0IH}h#!q1NXwAtlJiESH)RsY;by!&$rdyLIL#
zI^z<x3658gX2m08uTw75-&00JF=@CNl`K5K!NrZ<DFZbzM@R!em5*fsOS>ErNKlm>
zM|6tYKUWSZ32`8PrU|z5OUKU6*sU}Y=2Ke6Q4_%4lIg4J%xgJ(LvrUqxkw|JAM)hL
zxO_{SC8y)P=_HyqhE4nPI{q)a-F&-GEIDSJ(hD^6_qfg>_98ac<C-|a;?`Cx`yX8x
zZ6Bcz#Vy+^a5^;2@w!AB@_TRd$}IlijL;{jX7c;g;_i~i_#xZ%musS7sn!p~JI#rQ
zQ?!X5z5d`ox{Rp!ZC1PuGD6KW?1uuMK&))3kuoMo4fq4c)6<$1l$83QLT7ip)2~%H
zEUIfxy@EK;_C`M5j%Op+$+6JaZM0WqG^_k2ECi{2p<b=*y;;}F=7pZjpq($m$;n!k
zVqGN(JZZC9j=eL{rrZmd&u;2sF>t+69UfE3IwrJA9bMs)p4u$A>WhI?2(9nR(aasT
z;5<(5bE*`e#2^VgkrdaHYXi6E*|QV{sXBe*`KnXzj>B?0xUeuyw)7zm!CL7H&7T`n
zO^q!<xyb@cl%mqJ1%_Qwt_g%jn+h>i#nQgsop}*i-8XpJI%beu$@ZBGX(V)65}%6X
zV;otV=%{+yTrX+mD(>_~7Ar;tq+U`~&cuou0xxw>ML?zviUvA4RR9v1hSC6qf;T!`
z<4~(9Ayg$m4Zj4W>``OR_v#g@-tIv?&q*v~fXsucSudn49godQ4xwGX!8D_<SLLan
zZ@bHrXSYEa&NC%kxueXT9mYO-%)qi}7DFqGONe$iMUH}CM<eKP2aRs;ZOL_aXYKFu
z2Dl4Nn_V8w@fVq#7O)ZMuhROMxle`oRGnhAR2@?ifD#eoM2D-8`Hs5$IJzg?{%5rV
ztdh<H`_+!6X1S#$Rqdr?XYs8!<15%{`kPf~i{sxizq;Nkdy$Gb7}C#z646_gs$;<2
z2nr0010IMz0L){@fS~mN<J58~qY&V2g*%)!KS25Cc<AhXh<DfD<(I9V_Lx-(H#KvA
z`=vxGw$C`y#1)>HdV}L3vk#&DxhLL3ubEX0sBlUXyvR}--1#~T+;1C)zGhRrin&nA
zidHo!lS0O|73oFkE=b^>aO5z)BO_0RgC{H+(Kzww0lLV{JXN{9jCFZ$TH0<*OXuhB
z>M=Vq%VrDmYO(AO`XpX_T+=SAa)!6Civ<BdYQZ~?-*r&eLlSf^<oCiDh(MB`3`qa}
zZS!+nPjwn{>ckfT;-x7KdTWR_{vsKnFU%Yl5J7ZSr9txD%eP$?mSxixLxdSQE`AUZ
zaEv+uEkZrWzzgSgEJ+mQI5c?&72klw18U!j`C|5BAh=2{2x4eXhNyM%=W}#VliwwJ
z=~Yp2ad7~7L~A053Ux<s*-tWol6aOOeZuIeVueF-q2Iyy=GK->g@dh@j!sNA-Jh#(
zF25*1*wI5xuYh>7nYOZk`88xOr|o9E>adbf>1jSB+KPrT&*at9$fgVY6Ct}t7oS+A
zdKrpd*uIjF_8c`r>PHCc{*7WzU}_bZ;bLLk6NFOaH3aRAr|pY&FuL2Q-ykb5UdR6i
zq@JUoV}P7}fW>24cP2~@>giNexfeRdlYH@Ta8e*h4D^kp$UszRQAkK=^Uu|wTOEr_
zrx=+7!baITQS6w1-s0m2=vHo8S`kpM_OqDf`v@vz41#)H@t;wz!ST1JoZ<cwf{UPw
zOpc21CJ{9^Hg-a|9!MLpAi#G}-kOjHHUsK2@jCI^H^{yY88zyM1WOjnMQMo1f}$<Z
zMo^_<1L6V0sH}Oxd!qO6k+2CGw09ATJ~Z%w5^RuO>{YS%`<mlMEgprB^~#V8-~>zw
z|Kkw>P#Q)9pb#<0jB|x*u_>UNoByu!#ql9h4yKY;w!Db`zaQ|?1`U;;IcKQ?$!;tb
z^Ab>0ZM#f$SQSS3uR5(nU+}`eYOGu~{68P4yjum?X^yacFOH@a(&7p?@EEX01n80L
zHu&8GHKMq|(<21pZp}dJnK<(VCWvH}EVxDF8PqacrjDujs#wKvj~o?~7+4kFp?J*5
z5iMSHJ#=?AUUw)+0kSu;ApT2*D$bB2Ppi!jqjk^aEQEXWF2wbFk&6{WrH;PZN{3N)
z6js6H<%4QL#Xpax^LH5*gDp={#RedmK6bL&J%SkO0(^wh;4s+FTx_5iOdhpu;4H@1
z@4GUJ5ybUtWCoLV`Nx-Wo?P4R@?=;LcRmON2~1-v2r1v3F1myY8{b5X7!}s}5Trfd
zF`^BuZHEFZR@FGVo*hsKMcWFZ$v4se#0=2h0<%qsCn$4iahrwX%Yu5fdT+TRC7nWE
z4VWZ|yXV1^g8QLAgCc|E)J%ybje_Ve)kQ3|w6w#)N&Qzl5!s4GyrJX~wGRr14Z$Uc
z2Nae3g&_VbK5dnIHjj%KBYccr7wVoX8;(KRsIv8!t_E#ASP9R(gnP^vpy|(7dQv8a
zL)K*POUT^w?uES;4$;OchYeDeZ#b@P09leg-eX_qO|k3PF|$^Z*7;4U&2ZE45^-Dm
zx{=)H)by3YEN<VJo8VAwGoF<^zsyWW!z!0mV`Nqo1^?`<W~hq21CY%Yght--$6t_<
zm3cNryc<#hlD=y_Kg+7e%!!PWa7mQx+D=Xn9o&ag1`p?A%-R}%6Rot1GMGkpuaxDH
z62WRj#eATHK6KbesUo>~B4bka7leg~RIO}+F=mNF$UiBp+_9>?pHx=IHcT0nvsT@z
z8fopun5Oe$Fg>rBSu^NK0SiZd%~QCCcI;Vj|9oux054a^HvQV}qlf&#R4;twcT$o2
z%`2~{SZND*?L#NKDBo0W8x7O43U>W+6}J4oH*KpmA2py1#Q<p}UpdKs)@a7a3LxJ}
z?wl!80o12T?$2O9Uo>8c4($D196#>lkR+><8A_f&G#<Od&ng(8zT9b-JUAgWl-WKM
z&#6;EzOWe1B{WjK6`E%|*XvxqH>4+>`JQ~CFfKxv=+JgND<EFw^C<l^KMXRJWeYzV
zOx6xlV}^kH`yU6Hgo4fOS1%L_UiR%DAHwJt)_)}&5UslZWmc6(CM(0f-o9-(O6B!O
zx}%^1)p&;`M<?g*(x96pSsN+X<btOnC!t1$Cf!dD*)-ohTj92w+Y_S`P))56EGM^J
z#VVmOk0MWz%D2KhGKZ%IJvdt_m)SE%Yt0Orm^fagm}$?dN|*98l&TsV9X)&Mqm<F4
zlmxeatko_a5a0AV*uviReNrB^6~}g~56|l9VB9OpPYS100Y>a4d;}L1Qt!y@M;=>e
zjR~$DvC4MuwsjE_pNI$Ku~CW!+&1);bloe#+c|qSZsJ*Wjpww#a5IsmWs{>$7LB=M
za~dT_%DdgPFY>lKgw|3wxb!Su8_1`_x-E%q%O5jOTn)2TIaZa8O(CbIL%Ns|+Xlqv
zD-e<kItJP`8s#fbh=UZkM9JSlTNU5Ql!_{fzje4+-~YN&Q7&rt$hXjEi@UONA!=;S
znysSu@aTBjIZAF^Chp<GF5G*9LR=M60grV~UkG-|u$0F1>%;8Yy9JH{tmp;>I$6h&
zm6cnE+{*`v*6-s_LzWu0=q1Pn%TM+Fzti_-ZgY=E<f_`AJPJNGpv_4ZmlPL_4j#Wj
z5FW-V7hs_k%!;?`in<}<&kfJ#WjgSEUXcx^=UK5e$DoB6{?<~n%&KPz4U$i}_ozfO
zqOmj)T36}29SOJKzv@HNKdI1>qT&#>r3c64gfQ&~cwv`s8f&-W8Y0rdiQ}X{Wo6wJ
zMLP0m1n@k}USU=;BS_*w9#4z2q|a#NG??}EA34U)1-yKiw?1JeD7kf2^tRfy&}wCe
zg;<yp^k`=-J5MZ}F4K2MO}m~-rtk5%YB(t!XQE8lQrqE?2H)UNNV#%r6C3xE?arxw
z0y4C(j9&1LL}`2NCuj?m7Rs}&(ppu_T=C@T^_fu>a?|k;lUAaL*@lx3Cs;yMh7c*;
z9qmh1p?5!K`y67uyw+XzU00^=!DMA(cD-ADlitqAg=W2M8W2v-GZe?2YtSkJpEF4{
ziON0*`L)@^R#oF@cWtM7RX0a5{mWHP>GyFGhvwZBf}^R{+l}oPf}p@*-n;j-TTa2*
z1}WBs`ASd6w{>zGlT}{6au@5`VG}Yntys8J^vo?ji~~78Oh+9aP|{smDQS@>evqq3
zmr6{U#BHg!5Gxy+Cgf#&@ZR|>JUwIeG;ed5Ke%L1N+vx#o?Z^KtLrJYOr<Vi{>soe
zzhg;|?Ypc#`h6klYuwUH5i0#Nj_!gXx#EuJ5$AC|Yrnsl9Z5UrU4+u>gpaibA1jZh
zh8p1flAfRqB4-guPFn1g5gR@+c^+a#G)^IyqmaHRpI?s5xLfa^<jo&IA;_Yhn&RX@
z{8U+tZoFLaA!|lZ_d*aj59u$Y02Fza%wZ)G;?K}95SdQX8vd|ChhasYQ8iT6*_Eaz
z#eQ`c@28AXyw^&~5X*_VZWq$5@`<G0@ED~d+T6p^9|G3xUo%vZ%TnEnf+ex${Gx9P
zUa(oLCgWN<Zk0}$JPi)8dS;xZ5{K+J?5WIL$awEfxEIf`M*_3uGOt@~!qjSSbT78W
z_z&yewZ3}I+(DLQ@3rGDM>VI+ijxW3W6li)1zKF?KI^Zl2Nw`Z?4Gq`X}zhKr29#3
zN|s616nlZSPw(XJ`yM)T52`1BU-nOO^hVctKQ_9ZZEbHdH5`O5r5s>E#qe@UvNDi7
zKVzh;6#Xa@+n&W~%NkWw7c`q$=Bi$dc1~!=9C<wp^NV3vk`kvPn~OIhISsyJu{PPi
zDo!NQACwd$J>)UTe6k^)Qj0TZbbjwyc5}iKrMR5XuBoVl?#GK(E3}@Zm)>b=lJ>t`
zD@)=LKjjNFUd%eV;UH!5$@nai60|#feMNEKo5Xeldxv1pW&F_B%cu8THILu0iT<wd
zed)BB8(*@9I@aK5nq#OV$zW1G$*d2-=#QVP>~z^Z79#S+^zHWO6OZoK=NpV`voYHF
znwrEV%M<IPvcUPmd1*Uxb?@*dSN;*7GMr7paWvFxC&zdv=g_o9Lsc8h_WUlM+J~pI
z7|tSBvGvFo+e}p#JfAyD)<tABbDg|t6uW|Uf*cF?R<x#ixixlDTB70<Hh!3;SP}Wa
z*Y5RkwDW7s_UJ6b{8d>4n|Gt`WwbxS!O%2GHQXALI!-jS+z~}aglnv0JIiTGdr9It
zY3gAhs`mFymgW}w3)ZM>#hqc<9n8HnPU(<|&ni59tB7$o83(74S)*iV+_I3mwtf5)
z->cN|<jTgVX!;B#HARddWuxkjfg2Iqb7^*Al1E2TquSkHDM_ScRVp4NEi=+xqCcBT
zj7Ei01RNo$TfNSkOFcf_pS^r!BHH??G!Dt6%z@>w@@|sJkfT8P9LM-utJ^XURb`?g
zEIR_KAGZ3SpHCtl*1A2NHkd!_7`9Owo}u_KQAlQ?@P;=3bAr4_hea4O(~Gl4$`A&6
z#e!{VJIbz@cj_|u+5`(ui)Sv>Z5$6mbLD09dZ2&;d5YQ25X}?Q2B1pZ+ytrf#r?JH
z9#1{zCp=OC(xpNxlFb&>dNL^q)fd9F>k%KiJMV+1#S$oj@w+(7q_~E6Vg7Y)+_&rO
zkr51nsZ#l=je?wdav7=Bw)@)kReayTT0yI0ye=3}dkf$ByCG}o^=g#qci9%it^>&V
z2FU1B%kQ-xy7%g>F25fiG_)$V+*l+1lu7o<U3H}L!p&{Y%C;Q&b5nL5dWFJDn7_`C
zAz1@;>W<rtavP6h+VwBfkNiZDJd$=o6$=-mYtXdopJY6FlQ)PZq#Q~gz2HdsMf__d
zIimqrbgG@<A__G5#~y{>+2{RUj+J|NT^iCVCFXphSe)vSC54SKS~vW{5ZVI{1`)C(
zDrwHjrwqAqaQNIr#qAJ?%QsKi7Ee(%tII1+l*<QZb)}ezo&;o)h8$4|Ud?r1*jREq
z$UlXJZuhn^F!fDo=4#oNSmo{w#AVUL+??&|U$p`8jHSxrBbp~4-mf!&ZF2L4WT2eh
zYEOH<VYQ?d$w-<<GJdaa&kJI^vG~>FRcQkUTR_7QZQnuh-VVxBvDj_DTWJ*R_B4!^
z9W7FHYGqvQ&)RPK$VM*6<lpJ^(52)?;Vg-K{AAi`;`aAy#q}+8KK@7O1Vs{;SkwCt
zK4S+LSmpbWKKZsj|4zoROIq|`SH_d@5)IA0pUWOD<K7W0{0z>=8RXOtm;=z?5%9u$
zzi??@=$KD3?i!-t2^~;XQc9j?xfWEW{hr%#>k8NAZ_33->u7<J(8ziy5u6A_e=g**
z`GM2#+r44}lC=$2bLN+?J-hwn8o6L@*Q{<5*UEJrg0Wf|u{q5)!E)9O(Y1o^m9C<o
zwYHH7T2Dy>o@`-zvBZ>NucbEJH%rW{K-nV0w)npOQBk>5sJSxyJIayHCw2)?5=(l(
zZBbd!9;2#W2@x%OS=#<+oMsmN9sy;xgB$t9D;UtM6YD+s0+mCn4}(l^nNdBCK1TR-
zI_US@f|ujY^I8gNiTW|N!+SW033Yi|TvWMlp!$_2ibbV@rqV*qXLAOxps1*f44#Bk
zC3RbwOx2lf%9)ZG?$u4BJ8t(vhe5rW({Lc?d5q}E3nS~s;OK|1WS?lnfN9Z1U?g*g
z058R3;4;gS`M!cVo}q~zC~(<rr||cdS{v`ocfRaHDD2<4{6NZ`%Tj1#%R-y;H1hkc
zrj)__V+i4|^HeiVn-m8{1*R#a_0KbCa0e_oj}1%go*`I;*7XnQqIc@&Bi%yt`^`_k
ze!m*5pL~~xpT~(wk;lir+WyG(HFB{jnypDMKMV^GVG<3h3Q6}~y1fIRcFyty2>6h!
z#*+d$?NDE0ShwADjXp$QIM$kDHc`l|h)Bmlw6r&IX-UbvZc44cMo=w@^a49nWz*8u
z28XtSJ?jWTukxyil|npXaa64AdxFvN9(JuvM(Xb}PC5dPSHe7ZmbkcA2G#VslCsiS
zGcRG;eLoz%>)x$vb^Sbkh2sAq9FhXB`N#hj9MYvK1i5yfqCk8)MtWx0#uxEHGzL#a
zx;-g6gj}KBgzIa8w#;Y8E4})A6oeDT7jhFr2e|!1lfJNxsY~8De-$+sr&B3eo~$^9
zHm3yE`@Y-M$)aBB;9fg<|I+Ot8S{|p1_JTCMo;Yn{!K5Mo=o$hW1~>1{ra4Io0d!L
zkeoL<J~J5$(8Hz_dzDgI;d2!K>)t|wyB)P{>j$Y_)zdjj%YKcl2*1m)d7W&}5EtqK
z?wz!}W;1@gc-&%weGARS7LJB0imQdf8wq<CO3DYe#u+1SoBBH^WF)Ej$xW(R4Q9{}
zuFRd%CAZQj&?X!{5()@rPzO8us2Fz)sA+$=0Bu11S+Q!atHpO={s@5$rSqY06f4CQ
ze~1b_l`hFrmYCnqs;0ns%CO2ZQn!E<Cpir%Fi^_u=XpchY^$<eG7_LBKq<V>y+5Ae
zCy61%6(J;=8gu9X%~mSWig8FkUg`&=FR=a4?k3^`F_>gUi2BWVKI*RD(hIzx+`hkA
zzcqc*&h-qh^+<N=bl(d(P~hF1XRu0cxnQ35`3l#kUKV>!-^H1D64Cg!8wv9!QX23M
zA3v5>M%^uJPB_x258_-I2e?juOhZ?{U$40(=Km?~JHwi6x^-j0F6b+upa>$pD<Cz1
zB1%9&LX)l{T|iodDAf)K7^;-e0!T-C2gODWC4iJrM5^>I2<-LXyU+QqbDbai-_9>D
zGEU~1XJ*ZsweEG_uHD|hZrnJ4+N$kBqaM&cKgY=$+`p}Z$0drHncY4q&k4!!Zx(b&
zG~acYS}V3D><}&~xAE2MZRpezElUJfW+*X=?X@BMG}>tfQAU&$B`?PtLoV)J`yRT9
z$+7F9l2tS14r(g<P+LeYPx49373(^j^&LMKj;~W<sbf4sZIepvmA!CcYu1iVt4rmW
z&HAh_R9aT{mTO&F;~aR#hr@3*Mcj{=)Lys|RGmj0lgmJUC@jC^R+_WdvVBB`zSb+j
zY{`*|6px&aoTTkrM;2Scn>_3GqkO!BLQ!+_xO;9h)1|nepDxj{qf`H%+-m?qf<}$U
zw0o>usH6Wpu{HjL9wwF<UXfc$J)Eh7boQx=A!k`8soB&m6{@IXn{~5*JNlGXLSj(J
zE`>qF5g;65X<LXTeze3Fbz3)^-gC+P(V){-H%ycohRO#W@Sxd~%v58F@k0gw^xBov
z&KZ7D4Ko#JqW*9fOPgV0=qW3*bj8CtoF^nJeXVtbpGSTpl7rW7nwep2j8|4V<&!(F
z^JwUIdrEonz{TC(-D_(H;&1eN-B`DIyYzGKMD)l9oj-q!XHU=U<Y+7ZxXH6L-H7=4
zmc80)RDyYRCifq^b?W*%Q_=A7P{p?Ei^-|g5h<Brfum8euytx4XYJd3Uy?k8`_+5y
ziR;4L9So6oAv#69HKxl{ur-Bh{A4rgII8gzyhHQA+7W&C#~&_P>!oyC3sGy%ktfFv
z&SsR9+$0pZ*8W3%sMaJ(^lI=BerM3MzU4P|g8MKj&NIZbpQBJ2d50A#_ufawLX50t
zTUvf8SNHYymF7ftOk0(B#<Kd?6-(LbUbCQ#E7Ls)Mj<5$B7u$HWO)UZ--O#mIQciK
zYvBQjXVVyJ$1Fz*$qPCo{j0RNZ`F6k?I(jg=OVf`*i&^;mhUqL&8%vCTGjE51vk9J
zxw!F#k>&4wH0ysxPltuLuQ;rgqgP7sa^h2&V<?4p5PM=Hl~UWDi;-1k4BNL?J7*J;
z{^z;l4$Huw${}4UT3xDkllW`TE~Q)G_+XuG^XTQL5RIy^<&46eo%*7M4?9$1l{F&J
z>{mLX#I>Hxpj|)L_^7;lV};X3&B=hlyn@A56-X)>#7w%IF#cz1*)J`|-nv7GeExj}
zCFzwV!|!(@`N;C~_xID8UpfIE?1ML%m&%z}rjjA${Fa60z$wR8O{Xnsz{H3!fk8P}
z8*CXm=4;pXWs}$%$J5)EpX{IqgH~ple54zkqtiO24I@UOyuiq^jgK&3REr<Y%)WJR
zn-V*4RY~1iGkr`mColC_Pd1U>UQ~t&g%VNBdU=3)3n)t%)PjOKm@!9_*u##onB&eU
z8D^;*3A$O1+L9>0DAx&j>fNVc2$ya3QY;I`qR_-Js^YVJ9tbK~FUz`CtuuzSKP6K2
ziC<SUz<8mxer9k+#vVcm8G~Rb2d|WrKCoYjN*z7+y)4XMuh?lGT;U{cIi?J80{&??
z<(&#={_uG1oJgVa%XYd5amWkTN7`Kv#1LrDNF`dh+C<0f`w##j-+1iLVBM(&0h>@9
zWlF5-PNBQK?t}|{bj+vPA)W<`L!9nPZ81%4Inl`};<8Cs3_A#!x-iK#pAQ~h{F5al
zL>sUgl$glwz=b;b#eVg%HeKkz*vF4oJmc;h+VB!;_K#JEygWVQN&W!3BbQ%o+oF80
z;ER@;oF>u^w&^uoDPH5nK>m+UmQNj!P$F<enW!Psf7U)ibuc`NF{D#rZ4P;^wv@>p
z7i$U2r{n?fI|YBfJI|=z7!J~iJ%x&^n|tN})mo-XaG4vt91@}o5_&SYslipIl#&!r
z77SM1)2H+@2c$zn3&cA7f&h`QlOf;LFhS6rAza($s1#F*s#fEB5E)xfC&0*gsruSs
z&#f;@n(@~~X{~Dgy3^hDyKm`lxmAgIJXvbIYTamLy|_!mhaTm7`TqS(;IB*vJ&*>w
z1r=`~*lq(hO-hbwx7#*K#Smnx5II@MnKV!F>?lwUe+^pPIHE-f@Yl(&XAs|tkg~Ym
z`a`X*C;CmA?PSM)r|)-h^IwlDxQ1VUG!gQwE$~SGiQfYwU9?@T0h`R*WXpS}MneZ<
z0`FEzES!C&^d(Jmb-!BOZN}nrSp(daqK0@+!jfuUsNp@BG0EcN=dW2o{U0Jvexo5c
zIy&RgOzHJP+pbPfP_l<BZe;iUOK9LXT;JN56$km*b|;`T{sl1g4CK+E_9r@gk9lvt
zUQtKI(bLOy8lq|q4+OZc(^YR&Md@XeJhz#`{08ryhHMs3zGBTbBEUM`SE@x=im3jh
z1afN<QpfA1i_dKqnMUPe*X^IBj_aQ}RqTF2$)%Jz@A@aJp-+0QB6j6~B1(<CwVeg!
zeXdFUk8#08c1G!E={Dp2FoK(%5990j2}LB;YWp{|al!v2k`4)6JjA8|*<}StCE@{Y
z54ip`pqJpgv=St5aIHx7n(jp9haTpu%9s4d3XduRO=cHVl4gQvpIV%pF(}IL7gl-a
zr9f!~lIdAJ{2Hp(Kzrh&utz!7zA5s8B8X5!B#HvnA)hS295X+IW?y|?2)is}G>o5b
z`d;8QjkvTwiR^{;6a`Jsbz_hLUX-{`X*zGJ+xgxsYCZLrwjkFRVGmOQ)zXE$!e6&>
z0p0+6Vm8>TW&CKma2=#$I51Gt{No3Gij*dspn4c%vhvrGS)eQu{%vH-CC7VRGn+7J
zoC5k44R_zKk#b@UyZFXFY3iiw8eXTJ)w=(T!?n~hCS78BI%LaNA4@}?oM$<${3dfh
zimm3<^8~oj7(z`uR0_U8GLdi%fNs9=JWwr7mO%Go>$2S!$Lc0r(t(E+B>Cc)CG0by
zcHID^<@gIdz4Ou`tBK?VfeRxkH&>+aem55A<mAfF#-JY+aijsubbVu^HfT^U!L+^u
zjb(r@j5?><hYb4%H3mAW3)Qw{sOn8*>HGq$870xgSm$NXYWM8Nn8IHi7;(PPt`BD2
zHkyTbUsh+6EW-HKNY!okva~erC~n)pOeyXv2mqs-ZznUZU+1wYS7$zBSHIiD(X$Ab
zvt^=Ru9q${_L96V6;)(H!%drUXY7%7((jIR>Lge9M|O`+Q=XScLZm6HnQU?NYh71R
z^^pMqN7N!kF2l7)2Zt--sFVCb*tj`Cz1iC-aOH4>02WXT2=%0~YljZA$Zb@H(w^k|
zpnex6>6nO!Z9=LuZdnw{w*2NpH008t8aM;U!{zuz&WY?*-L>>^zmwB)ZPE0);CWSU
zzsJQUABD?TRS0LTJ)UJ}8>)Itc7zKmd!FR+CZ45fx#+i>r!m-6K(P0m;P=tDDo36Y
z_UR-O)krv3&hn-yG1LQmb7?6b(XE1FB0L!<e(%*C{avU7x*v2=pG!4SXfy)jeOiUq
zZ}X+MUKUFGZjNKFk-%BFojsw8F=oc6eA9a+A%kGIP)NQpjt8g0|Ij_9bat7VX~#DB
zlM%1`Ixbc~JZFOUX;CJp{u3YwMH)Xn1Bb0jN5U|XccuZSLQ_@FdHC4e5ll3l+u0E_
z(#&gg&{He0{m@jQ`^ql*8cyfZWY<UNXJksPMe45I8aSeuYo(Y+5+yy{vXzt~&av+q
zlJN7Mh4A8o*1%B}e=O;9Z*m9?(mHNK-jp%3mtRLUxJ+WcPgsoW<J3K57{`z_M+qC^
zdd~^i6j)bVp00T&^Y}_siAq$Su8k>Kl&`?VJYQ>P9OK}n3|7Pl1c%>W4Oo*U*y5Pk
zTqKR!@>q!1r~R-&DcBe5*uAqYNXIw%*4$_YGmBVP`g)zICUp~+WR4!|Sz!sQGUD6J
zHJGH-K+o5t{_XhzxLJerG@pq&eJ7KZC%R$G^+eyh*8RHVBu()&%M)p8#&!1M$f1uy
z?PZ0jt5{)K4p=cNRONGuirwe^lj&LIW-fiT%4D1JWb4&*dRs8Zm-&0xRY$OY4%D0!
zg?kh@k2LiqcOwccip9L53%DEL?Kl6t=OeZ#Vcp3!u@FVnOb{e+_<x3}@qJaP6)+Wh
zw=oc&9KnkET%4Rc2p0#-p|A?}{C1S2>GN}TJ+toG0bykl;br3CKav{NcPgoTUhU?&
z`MiYMw(ddvI9Bw8a0kbRiuh@jEoL4oX7_{wu68SltD{LBNnE0Qo<Zw<_r@@XwUR%l
zwY_=_ee>7}6&F|pjcHcKSmGa#J;!R7bk`w`YaYWuj+ra>I4Kwevu}oTl`_p73_bWF
zP<0?Hlq)AJJ9CVQ!622;own0j7HJ(~$RkJ{dOJ!)94S;Ib}AI!p6sOVO6FNFa}VA=
z3*1GdF%m0Nc=`Pe>=LdlD!-I5yn<uOEBD;bp3rko`Lb&D<1HT@($E~TJ8Ul*q)7}i
z=p^(itEkLTZ-od`fz7hdi&j{wD7wnkuFm4$SG5Ke0Bi@Q<wu#OOYcT#{}h47w?m!2
zEh@hxlwAskndAEs1&$VcR9K;@vZQg_B%0Vtlo4%I-@~RI!xQY;!_)Lf&F0NdCI-BI
z;=istG~W-+Ys`sD^ymot9lso!mmak?iyG79ZFY`p$$m{LQU1GuYLa2$?u{QfeEe_`
z#;~2Uc1l4c=njuZ!~JBTt*aZI+r8zzHis864;cZibg)A^%TYsodbxzJ(|$U{^Vs(t
z*IHSt3uE^Saji3Lc3kUFbM<1wkfP}p2}7)ygG1CqugIp+kgfZS{x+PSGBjd2(49re
zlN_HmNUUqfSiR4XBN$U*S0+!%=C{|}&;GGnkV)d|qs~9^Bz#kcLrYTB@7fnO1Dkkp
zhjuErN13sZIZIC}!{HR0^#x&9rRw809>e^NFNBi74<GsKhv&bdyicGXiC#y5EDsFc
zy8kXSpS+Q!B2NlzikkG<>Y8kivz+*_Ym*{!p_SW1+DP4Qz)N93z&J(2t2dD@dnTb=
z*kLlHYVZ-{+d`NKMJ!7JD}BLSrBn;<D+(e<1lN~nlex*16G>mv#$2C3>TPMz+3Zi-
zCR(F1i&f4;DMZdojWOCdE1PvB!t;0u;?nCXhjgldOYfIZ$oKcvCI}4-X>ViNU324a
z=5KFB;fw7#G3(<wK^w{2#V_REeH~!U*l*dSqaA5nEM`3(;i-ujB`)wm!e@fvoPMDZ
z8N1KNg4~1YDZZ0V^xTwjtsKQwud*_pO^}ygPmfW5k>a`RH(6AvG-4yR`iOQfk^RNk
zY-b{UnkNH!X$|L^EHcn23)WGA32LugdvHtd!YjA4d&CyJ<}aM%x;7klIwj=yZ~Tw0
zXOmjRMgOpW+$8Ra_+o=Y#b<WMa(YzV!rS<YIWm0Kh%u@6y<pRq7*e)W*+e;0EHA6_
zex4yTZ_Z51nN4K6K2*+d>FYp#K!^o5@^d_nOwZ+h+xhQXE$xAL=xipVf4|4lITm4o
zHoD#-aP>|G9gp{BrI$($lS!uG#rA0Cx3jgB8q-4NGr{&OtDt|~YX6a-m)Kv;9)(iw
zuhOKBeP8e+bw?*l{l^vIaT9DHZ`Iu4TSa*b3od36>K4~EsWu%hN|j{>W%9;!1r?1`
zk|hf8j#*X6EA(i+so+CrEK)D$@-&mfdq=dBZ7>$s-*D)pN0B~`EH(xDu-#Vf!M7~(
zoD_6*5OB3+4-2}qV|slxubEVF@TNJL!##tNX{7)51s&i}PWpfitA}MtJ*zR~ym1E}
z;uO>KbC|qayUo^OI&4Dj<Vv3uYMg&;-Lud7ivzZSgZ!?=;uWFPEDEgav4Bi2a=->;
zb3!)c<GPL+!OKtnshqvEJx%qx8RnY!b)M|A=Zt2kWFBvZ$qhBNRNE5~x@AR%ew>4a
zVs`cqr(bAEeF7jsl;p|6X#28u@k)1=hX#w+Z4LN;ScqOWD~eKiuj+I1>PbPb)BG=a
zUYO#&$F>|hMI%tI2b-OnYd6Ff17yjI%d)H=n8{<Z17qde!fgq|Sl7yKYgCQAn1m{g
zrEG5yURD0|xy?l?HV3)*!iF0c;!*Hntq~z@qq(PLxL8UDyZ56CvMLOE<}E=s<B@Kl
zmmA5F(;cyH<0|y=3j4LYRQIbSlU0-JnZe8}+;)_}m_8ov-gnL2ujRcWDfie;$2ES<
zI@(rWxAqW7$I6YJiII1m*w>h8c;+6islFYvC;N(GPXt|k%E~9$+}=+oF?nm~T$tBG
zgF9$E-^$FR<&YO~EgB)7*(Q~<A~NhPA4?nBW}VOvp9iu`@+$71>q7D@u~yE_X9au9
z)CiRJp?K8vb9fTNcjZDO#5vi4{3*X~kkAnqFDqP#&=}8n!dfipxru+`x44=n=@oaW
z;=*-d;jJt#b&pn3sKN$wACWF{g(c)Ub%OWG@LkSFxjTo|v2R=WQk5KsM*}<ee=F3x
z^mMGQZa4yc>!47qzUSUixp9M{NQWlHvu)qv?D11KU8p#|-Hw^SC&BMWKl;@%W)K!<
zr@x&MLEp-)V)j*I@iJe$nYz)^{a%)~l5+H6Zl&wUH62b-5OXb!j?x}U*KCek6n{u{
zpimm=n`Z)dQ^>Kh?pPq<iJ0Hk?cx;@(sguvpO>Fs2<*e%(|=CMG%RyXXbj-}$n6nu
z<n$5Sx&2(&EXyByeIIudG$>TnY`u`4<IHQ@u3~PtRk-?R`_GsFdTwqQ$Ui!z*NUU|
z@un)Tla!g!HO_kDm3xxeqIVvDFy~`P$~Wpa;~Q45MjL}lsrG{Uc~_y%1|&nH0J9Aj
z1tHU_OyF!zBayB`E%r?toAhqYJto}@Wct;<6GYxw1vcJqP1f(L=U(5C@>(*(VBN_5
z>%WC{t-MwqrB}F_5VZ9Svl<u%(q?1#cn}%J4=7^i-N}p{TI&-JCK86QaS#5k(*t`O
z<5tW-BqOw81pD}L1tKJiWJWGUtEs1Ifm%j70P&ohoV4}yEO8WX3PoTfj?n)t{9+zS
zLzPGVc%+Jd)siPQ$?|F9#!(x`ev5qTrA>KT$0-B!_JG3eG9&)Y1cmjH+2m!f&_EAM
z96Pe$w*qApHg6%`zcpNHvYG)T$;rXNBO##&yoFGh5tB1alUdI^v51|xo#s;5KP6T}
zx7zndXT6JJjua)4I^1hTTngrH`<r=eIIeQ<HgwKk>fj}J<^+LH^zmkzW#rC0-%eSN
z2{M>6&B_W}oS)!I_Nk6XEo(i$Grl%IOdtVmexcMTuz6|P+h?n&s`4s0d=Md>KPgLI
z5s>B<KV7_qR)`;$e_$2;H9Cq}$76(mt6D7Mt8aD-((n*6zC$}oTPwYV*TKtc)YnIB
zO$bfAdDAv%wd1wcu<q?f>XL3&>u716#~h-i4X+)ta&CNmcj?Dx|LH1qb@f16R=K{(
z5}0-umjXUMKIwB=+fL$cGfRD@e8wQ>xo<9(GoMw{ajQ!A>-NaNsO_k*Ze;L*aeP;_
zEZM0UeQxtfJFF^==gJ9xY!MXgc&D4{U@9{+GY96)n>W*e2Vu@W=olMQXCfgE1h>hI
zCTBpCzQ4cAz*v}tC#*_ZQ2Y`m;RE)@#(+vq4rOb)>udt{YHyrZANGL^tJBJXe16#i
z3~poE>kzZeYGzrJm*N4A87^z`@rjm$WXc+1l{z(EB0t^#@#DSn{J3u39&_hSyJ_FK
z2Wgf$v{m0$`!~K*=DYHGlY?(=ZxN|OVluv_iO5SBdYi;@=FH=U_c$5(FSEa>iYiRQ
z-QEh$^NC9X5DWS)8diLM#>(>f#;n9<M012^0z?6Lc*a*$TI5%jW%cIu+NHx0C)wNU
zkYA;0!uA%c#aF8cj{&&}iPH>AL96_3ly|h1*K&V7YQOx%mq8Y1^unZpaYXy(&0JG2
za?7PCQF#W&QT@+Dpy)ImSo(gwyuI7Ad>c}7{1+~~@0bT(?g=2imKWhfc`XAqSK9mc
znn1eU29b@Wdj1?Y^2q2YpQxyo26{R4)j`DMqvnntx+(8!hTnjgj?3n{>$B(2i<q@c
zOaQzjnY-B$V`gT?aQbxbq??u1yUAbryIFGi0C)|g=*zPMroeU?eCHHWkg@uw*P&Vs
z3WVS@&UoukgT}L`PkFewu3<O)&oI!_D-*5-oEpkELj7rg9}0c@c7h5J?cOsIAV0*o
z0hsKnmph<p?+G5*Wlm0c9K52U`5jUb4at3dM)18gK<BFh_tj{W?8H}U%YivCzR6L7
z$p2Uo)}ru_ou|1i4Nr3k)StSianl3UrmUW834mOMIs<+Ew{7hJh7zur`)DT!lqga7
z<;i4VVZ^V}u}G7mqM~$%fn0-nS04EB8uE8~8>B-913DJ~SLVfloeP2dX9}-YiWP7b
zNqcQo0P{&Ii1tLq{Nvqi(G)eZSyo!A58U@<4~Bd>$n?LVY9*&5HUS7W<(iO$8p>P$
z*4{euQC2ylXbFc-;M)p~h~S221m+JtkUIDC_YY~Z5S3v^g$Npz+(g0+ii%G-CV){L
z!U_uI!=e{2ih<ahDN^tSw<m4-_+2#<E_~wRIslc2VxYknG3u$%ZOYT9d&9!Qs4n*b
zXAAlVGf6v_V^m}ipjQRhE?v?ECdv*_WGMh@Ob-AFK$129EI1ObL0CbcE(I|E!QGmF
z*9h;So2c9a<<xA*)N^ujLM$Bl`9C1i^Aus9hw3YKW#ljL5%Ax^$HKq7%{9Gud&hyc
z27v@ZLEnCIAm866!oJ3|0UMi|y05}@STUYlva&`53S_sz2NM$$;}H|1{3YWkV@hr3
z1^lCr{9vW#U?Kox<=cS^1qwodw(BEOdQBkwLq&A1zam|?Gyx-=3IWX04Rv;==N}#=
zp|0(+xoi(Si!hN+P78Df6Ze4x6)8DTkCUUi525-9O%__P+5pUwzP^sv%8!AwHO_NA
z4WLs{>I9v}Zvc@>D+ptx0!d(G6=4G=&1GduKbb^|jM~GZ2tNRd^ieVovyn3xnEhb?
zs;Rw(wN?mI`})*{E1y4qj*a=JrU@9vl}AQKz-VhZJLmH<O-@dVL24Xfb`h>xOXS@N
z1-Lk<MhJBP_8zE9`PwgrhJ{(fT7z}bZWe^LFralgL-URXz1S$)p989bI);WD539dG
za85^8mkWb5gY7RP$m(r)8jUZ2m}DwQT-z@EtmEV5o$oF@TH6A#r?BAQcqCTG`gqO<
zYC1e{9?0yK41u0=SB#W%0Te<Yuyg5$Gcn|B>i6&OfEV`j*RL7b+3FyIS>@~du5I_5
zCXtP^cU`1wMV5dxa1zQKLOBqB0Ei52FIYgy!TQe-4GzqJ=Z2lX<*gWLR~6tFq`=k(
z6kKRrTn^L}3_SEgp-e#Wfdc<K)6ehF(KC`af$}h~`@XSLRz-y@27B)wDch(t5s5)V
z$M`{?{bfHPsRvg>nIp9=U<3<YCa@hH9oC>r=_DO%yw(2>*zMD^vpa!*bO|JSY`|(5
zJE2|Bhi>}UlgWDEEO698g%Q*f`MrL<Osv-5g$$IBpI?<A29oY}eI>0AKkO>N>glCI
zy%xW)um;%A$-G}@q?|xmPYL1q0mjdG*!1g7HvY&^z_O%47&t9HUI9j8jEjX@iu8&K
zBh_$$)}$c0^rugs3h^5bou-dxTmfP~TX+xx3CPxLK}N6Nd72>}1pV622g$NyfWZ#Y
z2JZmXh=G0omY7H%q!Y@`g_;u3Z7?B_fRhgxBnMN5PoF+5?$9qgy9FmRul&v~MICG5
zQ4q`omiu~OqnRPc%IK-7sZ9rZ0ped<Tbl#JeDU7P(dO9R5vITGy|b_GvT+1-$)dW0
z{NuG;h*VNWOVZQ%1WhWwmsbPXFWBAR9Osdzgx#eBy*I<$LBMut#`501ahU(CFGw{I
zSQ-Keq}|nsvfY;dr}HbjdX579yb>fWA~<xkwXHxrNU+%p_!<q0?PEFK-1CDt&F+1e
ztCzKTB_(g!c4ppjg{T(txwV_dO5%1tP`x!V)6lqk_Bm88huhfNR*r$BlB@V70Qdk8
zF_JUTazbK+@QH1LGW$EY18dg|45D^wV*COFb>WryF9NqI)aN`+OiXP2d&r<rF4Biu
z>`8ffSfKXQ($Ue09_xa=0#w+nssHv@MyoYgemu234N*{Uat4m-Hju{jh}kjYaK;(*
z?doZQv*<OG=rwQv?HnJ2NO}J=!r#%LG6rc|?728NuEJwkJKhK3mC(4qU*&Slpr%u>
zSnLw)w+5I`;80|idtB1}=I{63>eqMmsC9I8*@CI^3kvE2O2)>(9eWSsJ_wfh?<f4T
zsKVd(h`Pl{!}IT78xSkG^SP~0AOG(k{~w$D_piAE|F<ipzXvy*D0w{p8=Y?N-+QG`
UeN|2AJ2EzmlKR!`E4O|B3;I)ta{vGU

literal 0
HcmV?d00001

diff --git a/docs/articles/vignette-04-nonspatial-models_files/figure-html/non-spatial_graph-1.png b/docs/articles/vignette-04-nonspatial-models_files/figure-html/non-spatial_graph-1.png
index ba0a3a4ac1884cd47dd3928d0b6cfc46d555d604..a653eab7d26c380879dfca581192571a3d7e5961 100644
GIT binary patch
literal 23044
zcmeFZ1yohvx;MNL0YySox|D8IIyVYPNq37#cf+O)kQNY-Mgi&W+=PJA-Q69+rZ;?Z
zd;aI18|U8dz3&+B7~l7f>lnCk?YY*PbFMj`=NHd1NJ&8o7n=+l0)gPlNK2?dAUFC^
zKesT!ks3{@B=CXxT24v=e1bn`;&r~@<F>uDwi5($&l&ZDHfu`l4uL#`$Vfa>b5Giy
z@z5G^nY}(36JYCfKvUzMR&vU_$zz^qZLRFI+wJ&WEoyksw!q?AJ~8%0zbcGNRo*l)
zJwW58DQ`mhi$ra49vdc0%iD|hSXS>nywLXX{q*CRn5aG8nqOasi^J`;Ly?Y?_`IAf
zL+IK?MYSGB9I9W#AtPhxs9$<+qv3)Nl80$%-~;_8H~9ONpC4*8ICsIwAbw~!$iT;A
zF)>to%v<Z2lO@K1=SLgqQq<nVW9i^v+HHb<Tr~}i@7vqflmd=F{^j65Io%u_Yl8eY
z@A=<4>F)X#YAEyC@$vDpI+ZUYpl2C{6s|ci=bg<Vk9>{1bNt}ZWnEikP3peI+o%DH
z-TB?VRkY<x4U?`)9W$)L6IC2@#?$<mIi1BdL8)MK#iSnBxL;2MKdyZdO0hBIK^B6V
zAoEA1rKLwue39%hUy}4&x*v-=9Jq4bKKsm7RTU;b)!vwLV`5|L9;4+U@kGt8&PP52
zZ*{Id@_DxdFO|9d$6*^Is(bZ^;YCsw4^Wew_F+6RCI@ZdYa>;<0&=QipmE#UFk*n|
zPaowL9DiAsoWg<yF|^nm1(HD<<y(K77;2iE5D0%O1qa*U1Se+<T?1@q<Q(bOXuD+w
zuU8Aa6JqJLn*yE94S)2g_LW7jE0xG<<tPJ_bexpSSQ%{qYpD(S9?l*XKRQ?m_5mIj
z7cXOkl3CmY5!;#@S@~1G+Ibs!g~mP{>?51hr4Q=3Cw9hT1GrF&!|ZNc2HZEM5bsfe
zi-{V3!E}urxZQ9~*v@BTr@i>E{Tj_KjkErBcB+A+1@tl>R#t*_6TWSy9v-!pED08=
zO4-$8q%PrrXV64ScR1cRnQE98FP`o)_32quhC1}kh<knoRjL$D@nu4^zeb)8)=lZC
z+HnQk-?4B1-2bYuT&>VpO40L6UWNpA?!}i38zDLwzLX)u260vnkCX1WIm`C`b=7c@
ztr|&NO~mQN$Ay>yZ<R3&!VZ2t4gud9?4M4woSIpJg_1jh4$lkUVJTerun`cb%v;QU
z!Ur3`kITTrqek>{UtB(cTW09HylR#0jF(7MB8Psh`O*3SHa`9df@z;%ZC9VguyXyV
z!{q8=@=(#JOJG+fabu5?R-Ojgc*#4qsP?0sf{OTwPQjF{vOIn9N$@xq9FgZvxEBd|
z)Jx7a3KwY4JH}G;tgj3~7vAo9YsLO7%n~WgAwYFvvNu`C1?iGLoG#}dPJOK{G?Y5p
zcY)jztW9057(n;M9==?=`bus~vbW{XADT(sscz&^Lzjab;*?cc2}I7G+6J!T)Q#`B
zM~n1x$QZ6t46oSb50|@8uV;ChxnJR)-UbWMzMX}cd7#3$aPSC9(Le8$aqYJCi%?Tv
z@zDSNdF#kv>h852Vh^hH<jIpps#`t7-{hxQ;6*+$p~Q{9`qSeKmMcwK-{6-^4t~(@
zi=rd^9P?|}nWca7Db!X91dm=VtI_-cKLZRhuka#9frfCS!5>z?d9EisWc3mjLAP;Y
z#wlS-b;;PpTF%>7QzM&k_!lyRl3lT^OZ$qGpD&_c_iawk&!t;>T})Ani{zMGm4NUB
zJGGrMO`d6J8Z#cA@|)$bx#nmcGfs=2thR$Eam3xjOkN`Di{d`K4936P7*$g6J&SB(
zqA{6{dF!+<%!l2#jno`0zOIGhJRPai3O~rN1^ZJN4!3Ih#gWRdcWxuqu5#TTwe+j?
zov9~R4wDv!c0=VdJfc^v-8r}kMFI<`#VdB_U?9a?(EVHgm~pv#l0POS4X-V8FjsA3
z=~_B4Tv}gHwv{vL>ak9)PrPolG5v;nBlp==@E&ynGj7P?)y1=}W8>!AJ3WVcHZpCl
zb#DuL){A^Bn?TPXmdZ*@Gv{eX=a$Tjnwt+eC@*DX4QS;(#BwK@89Gvq_KN81(ig2(
z>f}!v`yCsE&t)lDISvpcE|bM>@LszgScFJ45N*ZSudafRoEo`d)#b3S<wtgdo4(S<
zQqqjrnQ#v4@sg-zM;SXfYlR&v^7#42#SEcYhJX7P`(B96c%5OMRSi!SXANCLQllg^
z?(=4F_HH9pmhyND8bl0NRyOn6I*N^GBm5@wdffGEObpp;O3s37t;uqI7M9rj(Q|4~
zN7fF1y^Ve4jqfEY-C`R<MHzjKk?nfgCwgrkE)c1RIM`lTE9VC)Ty<)M$c{s0npi=N
zL%yy4iGRx3Z(YRR-Cb7;0Yj$BsS3w64&31N>{+9)Fso@E(bSw0R1CD-KjbufXN-A8
z_Ja{aEAtRb#?VAsDnjnrGX(VbjhOGLc3Tm60KLK4r&k4tEPV5<bS)_LRsDh;PLE1c
zkb#QG4V~QT&}pm#xaR0VY6T};jDy<n)5M13#@nA9=dDdXo3n)lc7X?X!7lM5O9~C2
zRQRmk(46oKOq{c99`w`Xcm`D@VPBoQd7Jh&dSGT@kEbWorJp1!%p-8|4-rgJ8IcX1
z=dha*G$0pbKLZVmNQGELmx;AjzM<h|9rcU`?+LxvOfo6lD~T$%Qfwn4kRkKbZ7Vu9
z<k-6$FdOVyvP|u=v?=IU7{K*0G*6k-Lejoh_6M0L4YNR3h8Q=pahi?YfI~i2sYIl8
z_mhFQZs}C!qcvF@pNS|V$L-n(U!?c;$ZkmKIXd&g`DCAIVEImtT)WrG86OUOb-F};
zWsdzfg5(tA&wANSWsvvO(m6D0!$PZ*l7&*`-g>S3@u(l?(73CU^%$yKK&ZKqU2Aq8
zaZS53%V@G2bB?eatio=B#Fd~!(uj7ZWQ#=U<tb=gHXrWvL<+u_<-3YInjBKm@;o%=
za#h=or*hcHq2{kY&J8vkv$Kw>!b<Z;r64g%>qmp1p=w?pi{p1*XSl_U*7KfM5gI}+
z$w9~XjVU`t*yO$zWkLF<PleN<XUq3{QlWD@C8wwEVboBAs5ezpF`qwYoXmPE8W|ZS
za9czL;F4>2<X4TMLD<IzGmETVQE%X<dyB{*`6*Py#F{qnZ_cZW;;#erB67AIEqF5R
zWTj_V{K~(jrR6iY(%_&<PXdozfo_$Co}ONGJ~~9FaC!L+J$sG3ihq$*PMp4HhEwm#
zJy*Y?vrF;!+S=N|p`rN>>`<s7IiIcC^XG4G+_>?kY=YZmQ;nXHk?)rvB+Z0e$Ti<}
z+cwW;?gADyB$UF)Zb+T$x$BJrPiDfYss2|Bfz+RqlJYqlPoINZkB0HhW-F(^`1QKP
zd~db^+(FIW-hSm^rp`^%)bu+AUf|f+nblK>`NZ?*&yRW&@SQ=naCSX;jtK3J8`)gC
zpwcWmfDSe_@_LuCODjSTqXKY97q=q)xKb3+YxDD8wzaj9@mhb`tzN=gSzFus#h5z$
zRXmW}r1Qbb>T0f;;)yP_(qiaEl$wstCc-7LXSMQp4SKCD6;2VU=5SnF$+ZUYTV1L@
z>>YBODvL=-$kng&`P90=kBrVeTbPofOBSv7Ji4~~P`+h+G<ki-Avrd^FU9U*oFwS7
z`9_!=(syVfhJ}S?1*SR%8W<3;Sy;C_NPT@OMNda39d!eJM#{Nn`Wj<p2a(%&`CD#f
zd0Fks>uDt#q(|4@zPLS{a%z4piz9EYGWd>^qFP{C;bYhi7PVLQBU-e5npC@m>_X#?
z@F<pEh?Il`6CdB0O+$9JG*O7;qeqVn42yOv#tia=xejLB<~D*^yRYbPfz-0vmqkr;
zxteHq^pkei*AnWy{^_~6IN`l}-?ForOiWB_kS>#r-WLz<-=_^}TSOo>!M7t{#j#x0
z`=9MhRoV3$9s<~x<^e{APsS}>=ejev+LO5ASmV6-<2JiqwbJd|w@bCyu<zU%>i;TU
z@wyGWMvr4<WyNCPn`F$CP<;a)fLLl~4kHyCRkxQ{S1l(>P0oLJaS{>`d?mN5SJu>b
zQC}gQFD-pa&%j{W8i2bwUXuP@K5^AG%+#)EjDeO`Ts#10Ae5A=+<xIX{uSp9h`Fsw
zoVI!@K7j#0?A-yCxmZ+^P-TwS1P#EM%YT86MyjO;U^}b^gV2&BA=Wx|%lvxrkUR%;
z=D$xD{)SNhPvZT*6O;eB(_akje<tF8XCme`(U2pWh8tHGdv2cHwpCVOvjs;)jO|p-
z6_%HmGw5T1<)4dvgn%#-5)y9YWTmJ8CNNF)UnU9&ME$IXbbT$=*Weu%zN;tOe~}X)
z5T#_po6av@1h5f>v`0|y9iPrNcoLG5=FHE}cXrge?G<)+zZ9UrD=I=U@sTsHJ5n+-
zK`EZQ0u85&A?F7j)H9=wPEI3bW{OoEljRn)GA!)ui1XSWQzzZJy~c~riHWQ-EF2sf
z5mX|fbDZEB34EzoN6@PS{GKEMRS*by3b8P2Bi0>j5}X@ykryXhP5$UH9qJ-ZfEfC5
zy;DSot@S2%c6AMXAU4oif=Nh7tdPz{2L}hgDbadx?;a`yY82{epLv5pfb{Z>fc6DR
zgBiFHx1A{NdQT$X)|4H5q}pMLPDUY1l_O4WsXc7yEykS(goK&-`D|cp7&igKJKmYf
z`tpTDn~mu2Lp>6If1godSZJ?4T|H03VkB1`c06G=lBda8pi?0$0y|RD()tpBN102=
zs_}Ehk(rZIF@%U+1*Bd&4{u7)f$ZKj&$PT!oq|%GnM9w_d~Mm6FNfdXzCT=IEG3^R
zs*jF|V_xb48m6&?<P|FReAQq)5$pT%MEnjh#}R@b0-=`GH;r*VzfM90PQfIX4<}g9
zLzYmWsf*|g7Ro^EwbxoI;fg?zWbKD+{aJuLC?NIUkyid^<Noh#-2c1c|B|0U>6;*_
zaFm3VkPgErZ(Yc~Ob~GTzU^8!WvmVlr-l|z+1B+;IiZ4pD1lb$I{^693=9m;oRpN5
zk~<>Ecyb+7QK$Rl8lW{eZ>R_)E65Bk|B9}#o_MNVz>`|PJD3>~fKR<!a~stRs#4^a
z!fm0la$?tb_Wk?I_{9o;a2VoidikiO!fFg=jDjI)+Sy~I>zks;Bh?1v@l<Db_h-<v
zgD=Cw!!K}AYT+p@=z619&EnG1NR}csE(O0*b93|TmpkCQeYtit<mpnxR>R5ck}Nqm
z!cR-O=RaFzm--@<WFbr&9D%^iDEb{799Wo`_S@OOzy0R8@YW~F%(h~IR;R5U&$o)|
z?PbCN)jA3j12I9_-Mt!|{}qo?aM)cnM$^EfjTZd|Xms_1KW)zU+Pkg2T{a4Qn;{;E
z=i$W!zG*Tf#X2%e^<E7VK;6Q$*&QFHc-k{qsfiI26L0;Dq)kb{07v#cak19{h-2m8
zP~c6JCm<k*iNFVA+Su5jB1cWgK3_Xpa&&Y$2vmrzF~vNM0>`==9U!(nAScfQsNB<8
z8nkLMB=P-dzdK!niZ|dH(958qp~YAouJ))pJC~dGr4)Sr{Gd<%HfVZ!dx@)`RkKjd
z%gd`^YpEk*>>RO0&hr3#HD45pF8KJd5=UI#F>$$$N!8Rpj>8sz1nJbfI}~g&Gc%KM
znZ8U9Kuy7A445YYr&T5v7Fi1miv(Vq1i%pnhlVPg))W8>v6`uM>Fn)2ZqEme{rX6;
zoWW$e?o4JftA00M{uvpi=3%u5z^!FhEc@n2UYt}Q&=?_?9~{KwRN^@FKLdbYUoX5f
z2R8&j7{x~3yvaDQ0#soWVD`XD+nL;tA8#Rd-QYuCo*0AlKquo`r!ONT1KVpjc|b(;
zT_&2rV!YTWrtjof5L{SE{oRf8jQ)P54*>z4D~=X@DI%S}ekp?tyNSH~{o8TIWlTRo
z#IushbNdM@{eont31EL=<GH2GAS#1^OJ-!y#l*&1g5YjBQ(LjPxR~(fPw3Bm=0>Ae
z`bK7D!ZfZc6L6>;#Z=MF*@imQ+2D-PEtlinnY=>%I?Mg}W&p@7hrWjaLXig&mR!{D
z!@~-Pr5AvOFg<)I29|<qmV&yYqa%t3f-DS{6PO;vek;D!@%BVihI}ILMk_va_?u+N
z=lJ*>kjl2=%~Ep4tPnd>nulwBMVB=h84}m$WALNIxyu(SlCPv*ykOzyum1gIcczY&
zi%S(Ruytz-Tw2VfU6=prcO0CFon0Os3-2>s17cKnbD}KEwr+=4W{|b0{u^LusPVoi
z@>~kz%PlWg)30+K!64RqnDg4#8O-5KxD^-+T~2xV@YWV@Vvh~`qesuM@F+gffn1B&
zotat}0gEN=j7t=~!DUMWu{Q_*y~a1^r3!wWR{FsD9|IU93`@J@^@mta!@~7-bCL7)
z^q2^M6Zm`(iPi3&4Y^>|Ftf3}1Z3NCs<I$S!K|PIby1V2qUbXqLxabaV>4AL2-s6;
zSs4%PXuT8h_U&6t{jk>8B|lBeECw@V5`;ZUWo2a_0Ae#etyi`5Zy<$%&)G`n9@o>S
z+;4u~T{;1K#c#Fe6z8Zro{Pt*<z~ZSuTbszW*#90MSciJt#p+0<*CDr=UzkYLIC+D
zLMKnNXcRn+X(}hdWE9)N##T{N`wXP}>0{cxsVYUFW`e^8hJWxK2suK#)#AxQZlj1X
z!y?dc6xaagM5#G=uSY<yw47GE!PC_DI%qqee~14$I+`|TQT)AJEZe|%i80~byJ^M6
zk8vpl)mSz1KXt9B7#O4wvFl{Ke}7Zv_xYcdp4_HB_I*weC<{L=D{A$dUxOV+#%=xy
zC}nOQ065LMFX5V)nzGQ)JU4C+9pD+$Q#~UCn+z2}fRgbCu_W^L?2z@$8WdUWJ`xWw
zZx1EaskS$20H7{yMdm*TD#=$(L=301fhIxQn_Ev52Il=@GZq(L6A}^Sf>oW8l%tKZ
zffxz;_jzwj!nxO9umBdTsi&7@&6`L=L&HK%&2OIuVu%PY1z4c-b|5M-GD?97+#D-P
z)hN(esnj=~u676x3mXB8SwxTy)rvEe)0od~R}aNee=q-v%JB2^>#XoNTxAv%)G0IT
z(~y-7Ui<zhy1YLiU5{baWDyl@97P-~byy+K5n+G4wtdZZG)fW@t+knR;NWb#dj37i
zIfyAH@hgWAa~8UY(9=tT8M7Ex0=M;pKb=JvG3%v}l$2x!8we=((X9(gFj%eY?sN{2
zZ-AzvFflPP`-KpaKndakz}7NRJ*I~T+=M``8Ln|MOSd{kuK+81IGj@G0V(OmmNLtS
zz`%>1>$(1?|1n->jE`%{sLVBbSD17WAvQ<4(#3CxIW0NX9JRC%BTnnDhX#;=^fHv+
zkjpyXUE`9VaT|?wy94R?Jq#HG4}mtI($XD&ejcs3?8ra-Nbp6emX}EEvSZHWxH4`~
z_-GzyLk8H27ciP~v34|e_32U}9v*cM506Q>FTAs}6V*XY-}9Rg!-R_}=vIz8XU_Zv
zm5MvH%-|F1zU9^T`ic$J(pZl8mJ640ULv>&@zlPT%5kA2ImTSOpSA`%=tJ7zS7)v@
zpN}_Q=zFcGPacoN*`DM>gqu&xTYLEC^gIZ|L43vqLI$8Sq6at-_)owDhWf7cG0JLX
z7M=k~*whByW%1f&=r@lYDuOqrVD@xx*EqRc1{^pdS5>b0xAc<RXvX$gowpO7z%Ds`
z5eRN`|4{VW;$o$)I8M!H&-_7__cPtwDuQb?r71Df%^zW1sH91*5Lc?|=p+;|KI1wZ
zXb*F|I(9USaL_sK3fMx1k$a0^>O2JKqt@ACrAH-6A&bxShM#Ex(09V`LWNm1(LJTx
z>ivQS7w_fBbDe%lDF_P=Ia?!7F{{TqN9kNZd_i0u_H6<BH}a8y_5lTjSEs{t&|R>M
z*drap?B*KvLGscx^sdMVW~cn{;R6*l4n!wks7kiZN`UoP$<V^RBY3aFmo>Rq?z&#g
zQh$XLC(Am@hb4M8IC}E3R6b*vxj=xf;X2y6SS{F^UVliEY)BH@n#!zo;07g>Aq*gu
zlZdMb?n~?wgq4&WJ^f&QM+R%*IpP4`6D{hlsB@smTrZ9^#5bpDM{981(Xs1BgP7q3
z(s{DGIlE|hOz-BNRd4q7Rmlx;Wu$h|hh%URWFYAW<vcTnTL<q9NN{@@FLez5s=Lpr
zD8S>sa~xchb<Kh_{W6@Z-nN8yQPjufA(9rvz5djdkx~SvCKW-YuB7xy&}}!L*Jje)
zk^^FX28QyS98t7@t+<=}aqW7OZDv-^(xl+lHrzy(JVJJlBI>FsCaV939*2cT>9DAp
zAXH<r`s*T&;|NzLV9YW_a2EM%yS{p=c@x}oz7vpO(oauMJL{~Hoi{S$Se`sl288qE
zr#+<qo5$i6y~oPncVXsnmYES<?9-K(PHB?0*VfFuYS?*56&PNeDmCTB3(1S;B9FV{
ztzL;lcwS_svy@ZsZ?egMtrjIF_<5!*n8Yfd!h$ORr*VfOC2DP_b2=DaMn)OL6jC#~
z<gO(2Fce|A$AiZ^Q>#v_XYUVyYsIVP8;JD*aX?P%-1bHXGv#<*H9@?uktY(|kfh$Q
zxfHU2>)QRXTcR81MKck3wcbl&7Hvf(U*V#A((&9LLq#3sxSutCjPk#2UFi>=jwoNe
zD8Pl?$u2hF7%_|P4iUhLap0cXI&@ENbvQRxnJYL=)fZ5PK>k(>faoMpS>|+Gc<GV8
zI^W<CeAHK!k#)_0T*+yNgK29=L-mA}nHgXwc}GuS28I`7hQ19%oCf(FRGu$U=)EK0
zrXSI-ji^m0p8nc!r^|ao1-UWRGwO#YVygCnnN^z7HdC@9=`6ODzY~oH*DMIlqUv1S
zprZU*uz3Oisxo%wc&-q+{2I1n?|0_L1a|u?ZI0~wCJoWEd->C#*OyKW7rS*8g^G#1
z@@}&pW*`dtaXt4w5raGVU&w3Q$`<i=v8j8#Qpz1~Ur=cayd{Jow`a%}Bam5Q=n1o&
zKWd!Qc5uO!AwzUwAr*X1_mFBEu^U;UW~VDEsb!fp5t<T?#)f;uk_uTG6rw6?)n~4O
zRJ+=1kNn7&{6yE1oKRH;O-%`@oA&(G1$>ZUDR6I1xO!<tXCFbcO2l0Y-;&Qq=?~@#
zUSF5Z&?AkDF4RRV@4mquzVB@}CLcFr%_}r*ZAg_93~e5^Ijcc`vqiCn^e0;g*mK?7
zdV4?2Gm<CI7uk3|!mZIrFW;bEs&)bOYoqB&p_Nahq0FyTtbYVn4G0QzZgbwXz+*rR
z%5f&&8vt6y^@;Cw7AcphyR{g^yu@-6U8LT)ts+gno04fJ_EE!C`q7K<2M9SOxxqBo
z!IVXvRLe+7%D0nwW?LMlaK>rtqBV!)ckhP3y08}Y4r&$Thv5vAKHX?g`(+uU0*!i`
z>-7;)shiQQKZp;NMz=g&Hvtot_d=T5dp&*(y&f>02uh*vKYmCwoUJA%JT`jUE3_v(
zVa>ZUQ6>dq>ljGLh$eLW^^L4lSI^nht0Tq;TwDnIb9v81O5E`xiV0a+QTQUxQI=F@
z(!C8n%8V(P>>Qyq=k~#maxtX!xrU=a=D=8WM<Jd5^+m9iXpz3E=l-k{yxd5<&B}cd
z0pWa%w;o9Cqot9j4!i`NO6v*ct!K}kB|v?zm?Q4@L-s@B(zo2Ep;l*kR~cWkL`OxH
zYST5_G}ZNWWHeo$#^aeMg>XcQD&%?WBv`RFD!@)!RU<6M3?}z2S`Ge0z8;%8KmqL;
zj5|ai?&FgcRxxl`RZ&Mab)kJ=0KdB!OVNVHluw{<(*h+ZnQ$Nf%&huAu-rp0EE}Pm
zBS$PMc1v6YCfgBy4htR<^3uWv&iw;(e_$Z)7o&jVBM%Mcx}Rg|!#nmlH+YHHJ6Vg4
z5On6Yg_Nuf@rzr`o2!m8UD26cf29sGh*DKklURY*j(k`U+~ABapi`i_;elQ%xd>n6
zjk@wIt*^g2VoqD`0S?_ecK>S)soT4wpXh^z*Ln{0q2ZK@4F)wu4~S$XKMZyfs>tLV
z85(<u5a>Jv7GwQ9SXCo@)Yno@sqtCY^{1JgU7c1U$FC-o2?b<@k3FkOF>^fgdU*;d
z`$Te$u_DYD+4(Gv;Zy@@&)x|iUo7I>BVj_WSXdx+m>o(!r|xY9`BXSbwbAbsk~8Wf
z>w>qW(5YUCq5H)pc^^;coDW6~QMv7`LM?U_Q$4m%!fbuR8ylGhjEk^_nI4MO9@hqQ
z&f^SaLt_d5P~_S@PJECX%jy2x6gjG*bM~9VQadv*uST64U-VWgimL{M0<A4sNHZuX
z2t@?wm>)h&bF`VQ_XrLTAKjiPi-}rXi>pGhf3hZk`#WP;38|?Kx6V%wmMVOZ=Pz(k
zg#DCt$qx?)=X00(MZuE-fDw;Z?L8I27fnh30VCQEID#+AU$%Rut#!wz12-o#Gt&Y{
zrjGm+vWkjhpx7b?SZ^*{qs&4=Q{-i+PKp9~VqswcthzX$k$j7!m6DBmZEQ?0<9%^F
z63ebT45&^7-X9_#f;kgZEEmTiD=)7O@c;4NToLeSBmvKR`gCvz<q1K2+W6Bzz&4%O
zam~!gVB+Rh`}+0kzlbH&G;9EN{$A+<S$ICo1GI<0O}#K-8-tlZco_y$!sC18RV74l
zqvw|*6q!NDIy;!`cBRZ04wo^L1R|s~wk=^s2q(WG@VwLkPcG^JH{}5vlrhyy1&ARd
zhkei@0BF={Uxa2KK#u^}x!Rr?R-1W>#B+%V)9lFBLXrr-3=C-nAe7=P6SnuEB0<Eg
zkk_kN0SE_&2%v4GkuCYnHhR|qVd@VJC)5=LPL_x$$vit2KGiV$&~Cqpn{S3|GK|%9
zJGznT*G#5r#o87h&u&2C(aIXRNeMu>1VEZ?;ZGYY)ZYpd@MAcf+_KAFqaPYBRJAjV
zP{}hOe%aq~?V}*%y4_#1ma?_w0C+KB*kqHQl2TfO=gB4zs988UPljLmY1+&Ix*a(H
z+^h1I+BPOu)~+#QzymDF9`s+F?2To{8RP;s4%nc<=4e6U+jsAHfVYV%R~zEGLABys
z^QHvz6c##Es&!-8Sr9-<(6e^;I~@b*QX_C*-!L*BSrkMCuvT41CpR<{2k@C(aIwic
zw<1B;?e#DZ)N@UJ3g$S){kftD;`46(p+Z#$C;ZZJ#;H#PC0HYV+*e!&<2W4uB&K>k
zHhst0+V@466+HQ>7H0VP87tNL*TqXcdG}R6ovIzRnydDih0-`87N09cY`mQ6U;Ufc
zl==1Vgr*G`i`B^`@#<BG=Z@;FjT+%zRG_y|0D%Ed2Qq}}*15jhOlFFR=e>UO$|StH
zr0|53qdGl8U6!QYga`-0PDPJt_0|#-NvM_GIn-r6P?7r}{#_~NO%HPzplmknX!>A~
zO-+-jM1awzOGWTnGBr{E=)3;4v+Za+b2P17VL9|H>J2CR@aA=xa<gJxjN7{B2RHKN
zk~zSm-Qyg@3_W)$zy=sW@z<sEI76SDxvS%<Fe;pVo{+XOw~74oW=T1*v>R`A<f5KW
zy5$_4Mmbt&Cw$ge9C$Jx=vp5A!|*votZX1nq~H7eXaF5wI1Mnut?jZt4KXo4JPQ6a
zkbyPvh4=kLK?#6eH#T?9=L^BqdjouMJ-WP+(`(+SjAc*3&etWKE6%_Z0KaTT$#CYn
z)O{q2aFbzSiVRON@3#vy&0v$b>C838vdmWFe4e+eRZx)diVQMqs&_+KSsBEHXp|HL
zzd{t*S^@Rv*RNl-%TaQ|ybtGDIkLn#1aMN9b5FKv^b|mPDZi`vl7D`S0K)whGg&qy
zC`j4WwGt(&0doTgO}PM9gOH{HHe{CBO|Nf-BK`i)9!1efX&)5f*XlWb*4(bj-Vhv$
z@l4a_pqxd1811Q3OKp6reVcQ6?9@|>T_yDvGed5q-A$i8E13(c>oQ_UQ{OOa8nxH{
zGvtn)uhYfs>ym8W>#NN<I9wC!rdKAoMe>C~?pxZ4Nu){XH{X*_bQ|`gYmPNZ*p@R2
z&k80}&S8w{ldqK}QbXuvR_HMt1$TIC@7ykz9m9z&Bwz67Lx&&Fnp#XLHm2NwC>bKK
z*8#+sfo&C_1hlKG6ZjU8;^IbhPO$QIDqg1&CsZXIi=8ySdg~*EbV=sh^f5s)%FNXx
ziz^$h-7o^nHLn(4%BXBtGb&z8+7{_{RDw)0@4HFZe{gUh2ml_7Z{>t_mUnmvlNgug
z5S4k>5Qog>xKnnauU8&#k%ykZ)ZAE*F*HvO-t|D_91TI@Wz%2)x9MO?=C9Q}3*;gs
z^61@21v~NbdBctK3A5A@;PRmSc73NF9-T7Nup2iNW|nT3-yB2seTPvJ&)R${tVR|L
zQSt2t?OA$Vw!?RykOIy6Gzf;_{OVO2P+S;fYB#<oS~g&E_fu?7H>y{z8#Ph^@sC@z
z-7VN1i5O#rb8>Kuf{Y+6cz51r&I|D&d>c7BW}Fl|q@0`fkOL0~$PpI(1*4UAbA{E_
z`~a0nN?gM3fAmMkVi6IUF>Xbh-T&Qnhw$$FL>&$+p=)J$p_}soCFRZ*;09tm?Szs4
zPG~0kw?cDM^u2ow8%?7Xg4Ij3Aa2seb>+zCtuev&&o&3r1OG(-5t2D+lYb>d|JrGN
z{~u&%ccrOH8<#%AQS+<qpwbU6s#?neJhP`6IfPW#$`@7n>U9bYgk&D!tMJkvW3*_X
z7K#K5m5%{&daE<m5kXx9488_sakN=ijXr-eAx3@Z&Xo#d`0hc7)sw27VYx>;b>{Pg
zZrR>zjo7yj5E3AbF#M(L@T8!!_8(;@;ceS%^|dliW$<%=0J?n|=@X#lI4nxA8`M{i
zB7kNdM#}Xa+y^}9t4@Fkk_4PkCSrLxHxOXu@dCwxPCr@Wlob^nU12*TXkupOX{-Sn
zH?0B9N?2GARf_@OA-ChQw2X?ibv|7VVEL1!CLu7y=E}e$)OCH0DEgy@Gv$agzkdfg
zMiC?tO<+UN(9!}=c4OlVG=abe)BHq#*;cXg^J@XEP#tLKS*9qoaH;Xmwj(w+_B+59
zE}O$Lz+Y4XWjm<ak*&fy1?bgjItT%%h5>+qPL|8W>AS{r)bBqpFlf-xt+E{fg(4MR
zXO1!|siNNHc5|W;_^6WeQ}iWviU{G~AHlc)1_BoGeDgb$2ns;NdgI0J5Sad*lE1p|
zg?T=AeN_dlWPp{xE-3#uSP4i^fgoR7>AUe$Y!2tB{zFik^SRi8t$z))tWU1$S{bbO
zIOKL(eHnp+5=c)IVLSdPhYF}HbSRq@_@l!e)V}$;wl$If3B?R;0E!kH6%};^N@+$s
z^6bD#X##ngVsjt|f@v(e6kuYKMzL$-2ocxL62K5@m%@d(z0aNhvO!QAPv0{IJOV#1
znRm7ssOe0~`K#3VKX+8QS<*s6LR^-A5-9edPC0GB#5e*8A3#>d#GdaTKYr}&0Lpjv
z(UEfm?w_FmO4$oMhL<lxfzgr$eC1!3NCbU*7<m?u@^)q$rhtM6E_C(Hz>Ox~@?UM8
z|KE5x|4frajR+tv`ken4fSUOX&0EFXJR8(BiSVFIy!jJ=q(G+e2Nr)03iQGx!~eDW
zmuJtWT`H0OB5bf~&L<BP&BesTjGUsNdAUTpS=#mG8q{iQv;ZY_|BIDJx(=|#=Co5E
z;vbe@USX{)keyHdI%3sN6yC*A_p%RqUEd46_NuTP$!%Zy_U#*)$lpcibFc{n9hW}>
z1!qgRurd$4X#pygfV2ZWoqval-Og*tZlE5`73lo`g0ayWY^Hbt>>sIhK^Zybze1rB
z{Lbr2AWy@LiNKMjJ>cztxZ*sf@0L?vKTUp!x~xwRz=UAEph^ru7{a9%)dOZ<+z8Av
z5dOQ9m5LOl8tyX!zY`RWfMN@UsL!dXbL&>mr*>R{?Y8v}UliB^gDY`{g}<U1LENZz
z6ak73aK%zUxgod3ARR6_Z}-mGtll3#0>H^W(TnbM^zKxsA<B*cn$(It{}Gt74ODUy
z#}}cK(fMb!iiM%>c7F;*)Pd7_j4EREKHtbf@v9r?=)lupl}{E_PxU^I2fhpFVo?}5
z9}6&tfIV3{VVxl8k_$o%{ja~AD1;!o_f|nwr)||1s!;Co%Gm(GyCEv4qjhDF%|M8)
z{aq3r*HNa$hQh-jMzU5pNvoXf&cwvWGY7UVAP|0jkSY&3Sy>jlyFkj-S(-&fhH`5F
zDE8pjjZYKueCD)J?iFA?GgKGr=@HzP!%W~|%huG?piHkRSR$_to%&5s$7#cxI5i0Z
z8Zci}&zwMj44vf!7fXPH$T8yGmokg)tpXyoii!#<@y|GPQ2sHjBD9c13&1l$xqK)J
zK)3ND=>BA}k@(;Js;j#TQS@>k6dk>wRviH<cOku8?D}rq-sq=C;w;n=A3rMD+7?cg
z080U7B;CG!a1L`(6h4qxIUyw_g#je6g`@&;tN-b+)%~J>Zrm5lqGy0uZFqeiXW(@p
zS>+y^@wZ|^P<O5-3IKv=+m9bVUzfQ5t!$;x0F4sl44^Y$U}API*fTLPW&d8=G^7-C
z?g!<dn}|eeA=j6v!utD)!pcjebN`9^^qxWc&w9!mBMQ%yi|gEQmW^UiOvyLtiXDMr
z1fyPNINe`BocFkGkAH1xc@8RRuRd1*o8-g$_vZh!Tqaw%cj}K#sZOM+pPG=@41C^l
zlr-HlWc}vN8}px^9wj_64j|&tTi<ecWq!aoNF3jo^39fuBaTCGYD_haS+DL)QRgCh
zO*D;i?25wMTlrgu0lrym7q5-FjzMElhq+3+(EF{KFHYVfFOvUxnPNNfQwCI~uxD6O
zcnFrjzkgZD931;U`?k^lUDYS3w~+qvs@MO_-v1Zsk$mp=gMd>DR9!&Y|Es!0ph06`
zU;vvKROaxTcsIYr0QgQ-MI}mBPL9WFlnq2_9Z+B+U;3x4c;%3*ACOsDY3Y8zaTK(>
z`ufHI_KErQDGh{b5Y$om0W|ol`Pw0>NGJ+~f^@SKkJZl=Kz#vnmI|oxiHUI8o7D$z
z+X3|qlv3MXUm-zdi3Xs)cvQmc$HHJqC6RhHepsT*Lk!3E_4KHtDif&wU1h=!(5uxH
z6ckqaMp6OX@QdxUyHU&Q@4ri1t5?XNU$mK<S2zej0GkX@TfTtUzWD`}|JfrAv799&
z-h;x&87fv5mZ-MZ%F@yw$HwyNfK$N=@B8_)A0W@M@v2+kqNeSj3~=k--^vZEkUuYu
zcT@rX0|l33;Ozj2-{z%YS}RDy00$Ikl{^PC3LI9Qa`VW0_wM~wP10LbWiLILuU+=}
z%a`oi_ZhN{JFe$ulRKo;=Psyyd!k;B#?y(9zkjHkruD21w;8vh@I+Itp-tcFfN^rF
zVI+037Fn)SQDttEqobFtnSY0gMZNfA{#(Cq+czg}MtyQARWKDeiHLsx9uj{$I!R^~
z=_S1CVA?C-)NA&FdO6unbj`E=W&G0*ptjO^!WGmd5k$r=GKk?*3LX_7axyY9a)H+j
zzAGfd#!p<1|NedL2XRoWDM2^OBo+dY@JESfXgYrmU}N_KW1ogbS;>zGb*-ScH_A~B
zq#ry;lSH-nmIh%*HMoa?VgBc4O#kbbNYWsF2u{7t_Ta6@UUP2DtE%RBvu66ZxAMkb
z_uzkS@Pn)wa(UFDRE&EA&1oih<?%K8>5}Rz&5yCC8P{B(2S3P);G=@V@6NnhVeX!l
zE-pdk;<Ttf4|-hsWdHUrI#6DD*D0Hw`<Bnpjxr_j=8a-^f|s?qU0sAh?@?o+xS!hB
z0qQ?~_Vsm`mTBwbxz;5E-VGSaYZqn_xyO*0Ftl;a+l{$IM%9vH$Lf`MT|T}Yy_$x^
zQ|FSerBU-Fm{DB(Y08&BrXp{wwC)W)?Zxs?-_Y1*>8*fw+u7r78GDM0_(N2hpF|}d
z-c75oPXb^ISn`u>cPfd*e)$vmS3h|$HYVj0l|g6knRe|k@5(1C{AmL-VoI-eruda;
zz7}TltErO+5s^x2J-PtY#L!t*R8;7>A7fmOIaKX;#`v}eAZYv8&%|+s%g=8YBCmMr
z5wqLXB%~1)N=y<S9%n}4`(L|a`@BUBml+)ddy1#B_L}cbKlsQsGk+jvy4sA8M-!O9
zi1b=BR(P9O_%0#gT>*uR%&J3CU8AvKs5CL^`A23q@GoKXx<#zDFGLq`tQ6~ozo8O6
zqvFu!j|VdIXCT~b>kkrqXf<MCVVS%+a~un5dk9in7wkVpM~j1Pe|B62mse20GblPK
z?wX@uTDwy>gVV;$9Eq_I?&vw+=G$~+Hec4)ZL!@;f~C6WSANthRP^(S%Ab3)YZt$#
zJ`UNtQ*S+9U4Qa{O}92*?nb#0$*EUkdiWKUt+s(dxDg3Ka;t4sbaJx1tt|(wRJb*v
zR|HY;jIG;`^nKD)Uz1}&x0rnE>I*l0H_=Fvzl~~SY$Q51#ngp#N7n7Tva&w;h$l{K
zT5YZ1V(iD*t><5reysiK>}=UWJau00r^TETAM+5V+D;(X{^5w)9)q-`B;4=O)y`S8
zt8Devyj%QR<qKSI+EyhdB~kqT{X1L`d-~}EGc&U#dw#)#p%7`aKLh%bP<M!KtLHA{
zL5i+*@~*9in1q8}?-$`=Ap_Kkgq5+UBOq+C_TF!^WXz_JHE8Pzrs3?S7Sb$5%%-2>
znE`A?*m|7lP_uZoiFo>CA{~N*z%kNGRb@WWRMIe1h=;#@>MO8vg?;%IQwz^G>3D8u
zG&wTz8Hk`x5%zRO{6a#B0BCCYc+(8xqPDE`OZ_M7LKLY2A8lf$wCn8z){d$1n`TnM
znx$C2Qr&ywbjI6xN!ZQe9wIiL)UrLsx@)#DOaJ+-i#d-^ytwEqHhLNKG<0H4VEe36
z<3Vx7x3n{tT&@okn!K$d;s=43bO=SGmWlh}WWi-+WezoZ$(KDJr4_%T!iO>!&O1k6
z3|&@iCL4+!9`PblFL+FHUuEac<=Uk#$tvkf<ry?=s}Qu)=1W~X&+e8<8iVmmXirh}
z&g#!XTym<oL<44n_7jn0a5uX>_tXbc*oiEaPaF40FE{c7f4Q1O?sU?Oc-YUbktP~`
zef=68?z!KnaLFq{H-H-Z(c>2;ADb>Bd%oSp4$RKZ$>`R6p{*{Esc!O_x|A)CB`e><
zLiA)r<4K&`l!%XfaV)|l(P6U3XD%9+{L^8>`<H`>acg^b!S!;K|9#R#p#Z7TtDw2i
zGtTd!wf@$cL>9ej=k?rC39r~sk}mJ0OdUP-6}h<mcHyD#xuc+BxaUX1qUs0(W`U$n
zT))3GauGzD|LN6DOr`rb4L@z#1;<x8+a~%H()HRvAf%zF{}*6>@Vq^8k*J83oF=uf
z+sNtDPUp9XJU)fEcaalk6shrnV=?lTtRYDcoHlGIZXXAdeL3BYS^f@V&Sy6roM%1l
zz!*x5IBDAKKyzOXopbaurHaT$-oKu{;>AyL)l77X*zzQE^Erp`H0vCXJ&4~rPl8#`
z#ebB>Vg~c|<34&e?z0%ufF2V_?sN8(s_D_{eC)w-mTsbXSyBE)X6x~vjvhbd5mBi;
zpOe}l+cr2&6NPM!dK1g#*liPU=rWC~;#ITt8(v|<1vVH%dnlcTBb+X7L*JYS5Tm)j
zo_^P}CH~?gyfD?DaiZsAzR_w@&pVDqr2d`p-Kk$nCkF-L(R-wpK9%i0YtzF9%O-dq
zck2iVf(Sqkid1+-g2}_fLm;1AX?;S*#0n(|XpNjvL-)gBd?T1<i2X*ed54bgS62ZR
zhZSC9&sY97oVJuVZccZ1t*gRn<2T3+mb*o>2Kfry4$mvF8Z*8-xQ1;L2SB(DU!HDO
z@}13$8&+L?k=a-_!FjcD72{=48oo#6YcV||swZ(T5Tul!$|M2L^!4@Cs<dojR6Tkv
z83Nsqq(n?xGba+;hL=xUOND-UB`-jvrtS(bE$Iq-Ek-}NB$_Ps<rDaZOmaQ!cv;$j
zxbmI5j96WhZ?5@eINCQqtLihsvJ7u9rh?dci^R#V=$>78D7s|Z2B}^gx72br*#<sg
zd^YE!p-Jbc27*%5<ISzD^x8JZp~1ZqDX(y#YSkNXGA7wSm#ES@hBZu=UObuQG$g3Q
zx5UzY!5Hy)N=Wovm-_eNW@VNAI0Z$u2P;QMv%QFIFTIy`tv{#in3HVL!27VrkTK*1
z?>4VcT^<~5e%XMLN*8$#YT9^<6Mjqn&;sk-zHe3weql)^M`St(u93D`U%xJ7SE-7s
zkQ5e#UB>q2<bdM3<;SmoeEeuzGF@Is>vVhSn0~Y%9y(DKT1>Foe)+4Qg{#RX-FC4<
zZg<&^+>nM*Af#4F1Y&kV-Z8^?oz%4|-)`h&uLpx?bNGGcuW-KI?RC4w$owp1vuw!d
z=VN{ZcbDva0Wz=_N?txD$pB@@xw>p|v9T)p`jiI(K(;u5HQ-PQ^MBK2;}cPom*4rJ
zn8J@~wyMl!qL}AGjGr`asvk{$$hcH4a5iHG&#EqbzchGpuucj$9{j{wY-ELKwALIY
zh={(0)B2DC?>_biyt?i6%3S%WrZx0dKl`8CzwiU`MMLeEYWtfrjl|YJ*#w|{Z8zyU
z32eoRK<?@}rHX%O-S4HqFb~oN&nmfFQ%@|dH_)x0StD-H;`!IRAZ;q2+}704c;SVY
zH6#NvCaCWXT_@_2?;9E%ES_*K(RM4&J$16xy}VPYVclWj=D0H7bsF)E7=6F}xQ67#
zmrn#V)+AwFkBwU1x5#(hHELNsZ|}%mtQh9Ny9b;1{}Rw`?Cr}IZbPE1XFW-|gKXc#
zmIw!6CL<c0+(UgGGh#O@siWar=ru1#@#{b&mlQ8BCqF+l7g`Pe2IaROeZK(33e-8T
zPgQXOQ7^b!ksrKI_>4ci{%}B#AskvZW3m11iAQq;Q=t$<>@mgi4&`A2CzUWm-K&Lf
zA3A)J@Bkec-#O;(YApb=Ps@rUrK>L8Uc?(+&x>vjb~D3=uMGOfj*2Im2dc}bdZ<o$
zw{&*xtm+Ck_>YOE3_J;WZMdZ}bbT#8L0b~8U(pbN<+gEiB(qNIe35;7uKrVo8ecjm
zoLEXl_QEcdp!IQfZf<Zj{Wl8$6BPI<!q+Ky6T`qOy<l6Mot+V!+^(Ck-9G-YdA1W|
zo8*NqKtJ(BWc6LvY+I4Ww3JP8Y-olv2bn#Uv6Etph1;u+@N^FgJH(`I*+Q{E@x(U2
zwJX>Zc!}tBJ-zvqWu(q$k2z|Vqrf~%&&^T#Zd@!L7L998NFrB|=dCtSOY$@u-s<D?
z!klLr=m5{X{`Gjg))_PJ>P+qiLj8!^h2Xwz$}?DR6CPD>k{o|R&QL-M*1P6o#r~g5
z2NE-k+XwFU{1lYudV@&hdN}jtdAp%wTUv-rC1cz27ryW>-7)Vtmx5gS_PaIvxXd~(
ztWS29@y%vC1dzSK=kAg&Sk+cD;=-PNcaA{W50X!MjGUK(%KHWt@>)sRsPGWBFnjUH
z?o&lY5XFZZ<<g<U8r#MbU8<+6dA<wInOwW)8==A{yiRl&w|Ivai<r~Deh)q0RJr6a
zIkh5-8wq5yl8q;4O3y6~4G0MMX!WzZn~sq&5E$iB!;ISq7z#^m3_~&QB|G&B1cqE4
z3hAA<zwr@S#0#j6=paf@ESvrUjlUBv6qwZri#_ucT^&4ObTYU;xO{!IgVSk!9X9{~
z!13tr8h+wcxA72_s(ZLwRw`QAS_|LY7j~)+1YaDD!%HZS<m*>kQKVkyn~JwEHl-0q
z^kz*Yu3pDviNweXfd_#%5CqhH8?qgV(XBu#quIo_on&NY=B4$giF98MnmZh5yB~kH
z-L*Z@kaec9NMP~!7ID?JW>j=NV`+mAW5_Z4QZ%wlz-XKaW`Yo0yc($($)7C_5b|*r
zvJ>S=ystRyH8AtG2M_%)zwkNTYp~Iu6g!S8-p1I3WaqKHu`_uD3D0<-M`|$b9k3*O
zvMAfv95YaD4V!V)4<Odm;RRpX3gP#XSS|!caX5*?<phQfcfL>-!MA>*agCz9An?<Y
zJx4q{D~2>~BiKp&9p}>xTjELa`Yi)wHECB@YtJG{iEo(@_Iv1Sb61$GW=I@E2Q)}S
zVAxs)t525$qu%s%%OgiR=(>~t2>lP9x|v?^9z_R`aHrTgJMTxPNXfsTdGdJvN7v~U
zt!~X}+@1AAU%WPqt}oCXm04fw8^Fsfs?_`vB0c;dMUu|I{H6|_hBfbbMD3@mbdC)B
zt=ElS*96D5U;5r=akL7pKWJBgeO^*H4Ifbt{`m0^u-~}pec1Qa?n;1Syd>CVPoz7=
zWlzOr-YT!_9$>C0fW9G7^G<QCCAB-#x=&Ep+eMQde32psV%42xoY0w_a1;yFsr{(*
zze-rO`RN{I(lvk5%gf`heKigupM=D+{ev|6ufwMcZS=*iaU(J1c4o|K!!0JpVtkJQ
z@eQKJ?vHw0N!&Uo8Zc;?#~SpKAIGjBfc=C0R?VmQ*DjKPaNgMTkG=T{LhXeS(KzSb
z;v}~P_rF+eT19GTf^TU6PZrU05fSF0HZU{^Uw74vsn}nJ&nNdaKh7_F0l0Tb&IfT=
ze!++R?G7aaD;N3_0^a%LieeM!v@4bSV1A)AX!FG4rVfmv)z1y?*Rh63+Mga=McQ9a
z6&n?e!~j-K$ADqJq3q?d-+}+Cm(1x$W!CSJW?$b|25-irp;>^n3#;4-JR%`KX+8`|
zEA8Estqnk{hxb~m$HPS5Wa;Whf+nx~Z<RZH8lkV|^GK!Krx&g&FSxavif*`DQhP5l
z{A{u+M231+VlEEemZsTm-w`#Zv`|!D%SI4MI|xp+BCJ^b@KzVQpzw#zh>ufc<enFK
zp&%~B`G{6+P_nuwzxT1^%@#BOebv$23tsVtoUpHz?Oi9TWif~mN1mormkru@kMKhs
zzXh%_VD2x#+>b<<HUbXy<zCl%u88+6)C56?n;Iu+Qf5x0Ts{PSr1<b6EpzI2;6ee<
zesi3u`RCC_Q3!swzizL;DyFu67+r8c04577tFDNbs9~Z@PecngOXcU%)h5B`FS2_M
zE+0Wtw+(yyq!cUWhv+KN3f}~LXBhqI>#JEhLNCpi4q=~>wW#nA+TCf>(7ZQiNj@JR
zG5=Qi@tL!|Ab2E;F-9cABO?{$?h)k$7T(}bV6R=-1Rk7<Jmt+vi5K@qns)jNE&aFB
z)2e>4wo8Zl`z`mf4|MfqP7e0Ml$35(DZZcWq!9tURZUY<is!dywN3M@Vq`RBi*41C
zYu$mn@X^K)Rz%0prMZr8wbW2$cd1&`l}d+qs!010F6TTxPq|=Y>+bwDg|AG0BMU*~
z7ZtzQYGRyL^1N5gn0c%lDXuXs=yt90_vb?&X=u{Q12*gQ#m3J0M0(0glzIVgI$e~R
z`F2EyxP}H1(6V*{W?DtS8ype*Fp41iVt<$}`bo-fIig+tme^pMQ*qpJ8&smx#681r
z<U9S1M{OvP!Nn{lrzfO@F+|+bdk@@lb`yQonm+NQnP=&tHalep-Q8QxO#78y`uZ~^
ztRg^g+y|rDok?714U`i3y>;L&Eh8iFeV<$o(4<(e6A@EeBxCBft^!PjEN(^r$I3g7
zmKR>bOf-&6D8Q7u;};hzrr)$F3kz4Q{DNEpQ|P0up*e&AoV~G8bg^-k)+R4+YyABZ
zFkRe*sQ?3~!K(`_R8-jYAQlqXQLrWTpsWh5@7+|9kf8YnVRvA>`&K)(<y()2<{+21
zDjt3VI$vmDS%9{_{4%)a_wTD;q<ql**L`WrCainDx`Nlzlrd&w`zY~nk$^ki#O}YR
ziOt#*8zy1`)(CAt^Jpelb`gisQRa2ure(s}yZsZzS4#$5(shDohHvfpr)}YObW?(Z
zdt1(eiQ?O!L`{hMcVbFP_)MY2p`^r*4A7;ep`HETPxntg#fEK0C9@@ZuwL-`NI2#O
zMuNBV1lSwT8Dfc>!ta%38jBq-oZP;7gVNT;9A==dKJxWkkAs*hB5-<onwJ(HiMaqg
z*y@cFLJ%{7sK+*T_x-gtnp~xhzKk^r>V<qSJMXExb#ArJ=SZ_;cU4EfAn+hN>uX_@
zm_9wOOJlpqQu?SKJhw1vLh)ifT1UpQ+UD-AIou8g69jibL&uaLbk{9(s?mFL1)`Wl
zc=C;MElax^597}bD_}av8y9cHHXPAnS8*^f(?^%tdc_i$&0!CO`;lQb2TZOOlbe&m
zw!BwsCy9(sZ#@#Aur|Ne3ro=Hwc#ZPk}-IG;I(bCkzrH$scBn2g{&(NRP>jDF8p{Z
zSD8TDhdVIW`e>JA;VwB87R+OJ9jhcxiImd+!K^9zo=YHkai?o4^qdN(p>Y(gMpBx$
z#m3HVnksST8JNmtHGB*0j>x1<hMWTz@b8RrqCV9b_7rNpIk+1pw6RGcrRsho6SyB0
z$Tqb}PzYCCYw!&a{gXW+Qso?)pE_ksm>SQuV{A`VKg>1ePvVGDnaOzL{;QkvHRfbF
z5gK2cw>0FKdoTL)>i{D&N1dbenLvIe41BoN?2XH}${o&sE-nKuKAiH@DK@M{7F@hQ
zNxb0f$7AVV-EA{wH!V^cO{?WaxQ}Y5Fz40q@ysrdE-=Mc4Y7Fl0;yAQEkx`#jX)3a
zKzQLeJYaCQ$rR{wQgmvw(=0JwVQX|hdF0iGacb2`;yNlH+PF>i)TwR{SvI!)UO%kN
z8K<|RmJFPInieFVIW&UNt-R*TwimxZ+GOkY)Oi}$Yi~h5$-7JLjeCUI)FAN}@4Z{G
z4Z<ICqbN1v#*OGl_AgH6TZ@{CYe~RG_(6tCwlupf{J5i@-HoPrDj7#;i-AF;H2yZ!
zE7IAHN;j_i)v?H#wE_3c=5Il^2LZl?Z?lNSD&{sV!NSQ0FeTZv8S@lbw<>87)DC*!
zb=mV0=FWwJ!nJ%6)6HgD`lC;DtH?n*nhMS@1SkCv1WETx-a+dfx%oqV4b9ggb+kbX
z_ic^FwfT(7gz`BU89J@z@ek)aRp);{N%Oo6`Y8QP*)R6Xq=EG;NoM)vrtG4e;_W{9
z7Nai>WR=^2x_uaKxg&%!ZpM9bG>FK?$er!sA`6;(h5#Ng;LIBuwrum5R!99PHsYMV
z<^`EKblc2@XRv^7q8Kq6@sMC6Wf+~kBfpmFyZ0?SpYQugA6cvxf_$5WteX3|DQRe;
zAV*tpis*l2`PcLyn~Mgre^@myXYAPw27?le04Yrz2YnVvL6w@d&xU0)i+|GNj{yF6
z5J=E#p#RzGqs`P%KLM*p`Ot~3N4c8|iqtKMc_V|gG&Ca7I{b7jAjdm}G-;Z-(BJFB
zc{M9Iu5qg>8DC%n2n0Z^7b5eI{rDc&DIg5$leW(m`Tx{%<?&Fbad>hxu2^D7M2-)c
zbu7BZEn0G|2(hjTO+HDvW^D#Tj<xYYlpKi++mulZxnhQdvWRJ_O(EAHi5YV5_cyhF
z?ce?WotgLget+{m-}iZ*_aC=2?+D{*Lp-P=WhBz~qH@8rnzhobYxs<0MeVo%`AW-~
z(53^{YA~XOb~?w@qc<PFD!A)7=JDM3Lb}a+qA91lhL=t0141<lZimFER8si_*4S=#
z_K@E@LO_or1q%!T`;g#Sd})pRo){(5FWpUY_G-D)^254YN$zgudS@c`161mF5lyan
z7%AKIsD0K;hHWo;(9%Nkt*PYdqx(FhbV)uQgChA8xGsbV)hj`4T%+9xz0Q9;lGj0M
z!O`L?r<B$|o_=-mRD^m~$~6D{tjenzrAkF~mpBCrxbiSga7tkLwKxGUJSYf>BsDau
zUDYf%cSC67ArFj=bq+Ezi~086s23yax)O4~Jg8IHa~Yj`zPbT{rK$r$;e1;63=<ZM
z^r25YWv`F{ri~mE+C~Uv`|bn0Xt3IU&CgA`_4)WPtD?HSeq6AsRDfJ0!5{XuePvpx
zS_;{@dD6_8kUMueR?yg37IVIq*Ls2>Um<EKH(z`AnidRBE!tsYHGjFSXQ#4y7m^@q
z{rk+9L(}S7DcE9*sLR<QAxzeUYZ~l>7CsaT49|*;N#E106gP>X+u6w!l}&0c(N)7n
zN_X(Wn`*DxXVN!!ygHz?6VgNF0V|yKyk~7BH%wXk_1E?_&6o8=4>E!&$`#H*?2(sq
zq@ggsX)C#`U?e_SHiL}*BZK=}ePtO~taOq3#cQr@rJv|%-+2FspWt&nDWU*QP7LHO
z)Oon2Dk~@F2~t*oBIBM8R?A4dm6LJ%PWSXNs|FuU$P#@o9JxSrmktt|%p#o;e$IfC
znFHVvYSV%UvE|n9bM?s}rExUjx>D@YI@h`L6G4H#6*eksmn7brp`l1xSo8G0nv%jf
zBRs?ySzPg`IV%ayClm4H1Q7g%fQ;TJhI@iy7wmpjsmKRKz3)oE;E((8XY?j4KSmuf
zf%o!-tyHu?G|0?3GpvlUv4rcNCFJztNF+WujS){SWo`zuDxZ8>=AQ@GZ#G!M%eV*j
zX_A6uE$}q;a@o{ifBryfJg*!WcBD0^K$_#FaOScyGMU$}v(9<LW{D!Uv^ba(=jrIz
zjZqgCXErz1qKKKv5Sj%(IVQIp!;Un4BjjEn(3QRfyLD<%?L7sB@<M#XF7MtFdiRoA
zc-;0-oL8)1hx6v#?_8$4%ZUJ=zuaA2+iUOxI9hER`(x7Xr%@|-q#UxWC2m~wiMY=U
zuqjOx?*oYps2^#b?p|JLJ%OtGX&>LmQ0pgdPE3N}>I3KkXE^fkGb^>nY8>mtgzY;{
z0r8n=iPmrgOX)$7V08d9=8=aGf9(`#53VpPxV+HSw7EJ1N1Xuy>b)k*3W(Q?SqA@5
zDK-VirZD1hNZe$72=Sm}`oN9O(+>~<H$fGY3J8zrBCJ%Lot>SBi_ijVF|7ft-LPoe
zrT$|gJrQOFp6^0D{-3*?5|T@z!_`<SZ|Z={1M{~dQg6YoRDx2B2gTVR4^BUQ7}Q%7
zxxB%4itU~sX;T${i;V9@v!?ma!zF1tC6$%wjdKk%6>BjIT^dcFJ9aC?u74hzoQ!gg
z;C(Six>z+7L{j?EoU?&FH=lq~OQO+9Q+hPWY;6|`3<1)RXjnUBAu@HK+il5hbO<f#
z%Svx799V@^_rgV89(5s#*&0rPj6`8??-C>+;gW(=LWQxLEA}@HQ-Q_@p|Eb<8#iZX
zI$Wj!2Wui=N`GW61NC-_jr%Wg=?oA8#XlP#1=$iAp5Hb=B4%Z4D;zZxAqxcn#!V9$
zZK1<9oCsVFHsap0Dg*KY<E{0@xWlWU(c8)c%ohrRInQ5j^ZGxK?=P3cv<}x`GMO-4
z!1tEQKPc<!W`OR^0AvM-`GfLoCo(NZGHPg>LD1i0<ou9xL=)mVA@|3Vq@KX)DlK04
zfo)cV*9}Z^Vlyl%+7E2fe<~Wi9KDN#;tsX~-M~K^X#`enpcD<?m+>losf44&85aKj
ydz+)V40gDuH!@RXRld>D72l}s@4;j<w)jri->QxoWetMV0A*|KXjN(HNBS42q?|7R

literal 23013
zcmeFZ2UL@5+a?-B#DcIbh)CN4ih_XBdr(va1QZ0M21P(Xii93Wa3g{Zr1zpU>74)p
zf}qj~MQVUZ@1YY&NOGR+@Bij}=bZV!Gizq8bJm&}7px?_$@4ztF4uM4cYOOmPxIvQ
z3&$Z4$Vshxw+$eWLp_Y&qesABYD_ef!NZZKI-0k^Blx$kfkA<XV{Z3M;Sk7a55_Ok
zgq?s71o9_D>-Nn@KFKTNh~z=f6gqvL_q_Fkc&3QErye{kJ$vbjx4b!ivx3mtRj3}>
zcndqazdo*KhD24?5?w0!l@czPo%%H5n0?nN0ng>696H4$e)DWo=HZ#n#!d??^zPj^
z)Be_PoB}CSuOz$$ydDa-Ebd5>4_KcJ2q=1IkoyQMq32@EEI*ju?xjN)AdusKfR#K5
z3_Jn}JakwF0y+C<Iuj%v!gLY>VLAF70=c5DzLqVL>LaMh=lMu&D*+dcI14^2$W)&s
z=Q0)_e1dB(n9J%4=?Vn0a_7HsKxT+K(|^}n)jexiA&_)<R~i)B?^1Iq4YS!^xAALe
zHnf>tY*{5HCPs1h^%?2K@wM(Jwy9atU;%;DR&VT9$^k0v;MoKv&t@w%$|^=)z$|To
zGpSw4&=9V}Pa7$yBZ|SgKN4tqdeG?)W!>dDU@IDtPrv=evX4zWxJ^XihuI2NEr~Bj
z6}MVH@3%~9&WT5f*ml>67?_xt5F@;jRKYe<-mW2+&1+_iT0Aj66EA!MEPt}|L<-Q>
zLuUDH!~dSl_wo~S^VMYJEWb@om}<5DIdO}}#2dSzVE>0<(b~BMU$2nPDjOL+j4;sA
zsc9J&`g@<&FW-U$GIyp86?OE&u!>rnh3MoMl!}WFW+wxSgwyK0#=TIa1QJ$+*^kui
z<dPa85-0A89?a1UuY0oa+aiGM;GpnTe~686Qi1s-#4T|D+uFRnp^lDRWyQ=Ap>xEp
zele?ZHDGlmo{NVE-Xd`B*>+byo)3e4dE^4*H^e}DQvjO~MXN{5e$)tWb=FslffifF
zX;6N-z*qV>V97{TCoi%OE@0uy_S%4$R+(3S_J<nZyXP9JrBII>1dElPshiZjE`$F)
zniOjPy=k4Jrv$g>QwgR0*TQY9sK+JaHe$-_$nSaY;^1YqWm5Hdd6B|Fj_8w{)L?0_
zx8s&JHrgF=tD7Z#Z%%53rwB-|mWo6QuPM-mc}I$OBE4bcouC?@UvKfely^?9&%KSp
z=?c_}s)dP_l4Y}S2Z@-z%|0XbBGu`WiqAxbzDN|BM)jqo;%Aq0&nsNdo%nG=l$q&d
zBa>@?w(kC-J}Pkw2Zb7D-*1}?5wIdgjB?97f6%Q)S@|_I6o)w5o;kaiQ??k(;}tM=
z02OW#C~w}GGq?)#b#qzr+wBZxq8dg#TUi%^4;=)XmOQl$n^?Ng*Og_EsFM|kn&@tr
zD-3tI8soKpq-l;olj14fAF^ynO$_%J<vba(x>;SXtIq7Ej@WDy*&0MCNNcj0rErQ_
zbw<G(;<>oIi<cK@Pi@ou57P4BU6q7ZJVC8!Y`TOaY3rb`V7%h0Rv+m-OgFMho)SRX
z5OLqnMGX-51QF>vTFCktecxtp*N$k+S{F>s{pCD1(B&4mSL0~O6O9R}UEP{>XWVS&
zbQFCnwM#53g=)rK(1!2UrY+jgm0S3g1%))ju94JNGF*HXN~`eZuuIco*rX&}9WmO=
zT<+b<tD<|Gqs>+k!lv~vZ4%AQ3@mF<4dZ%@JJfD<E5pZ11{4?f-1W<xn3p2AChA0o
zECP(OCsZ;R^B?W`mAs}#dzL3ET6SQwri+W4ScwzP8IzKR(+}cQDz+uLsqZwrDT-Ae
z)`+_;J>3;3S}6pg^n0U)zq>ccBo6U`U76TwFsG`99<wP4uSNAvRYYqi_%v;-PAO_h
z_<An$9K|Y<N>{0l7d6QI$-<^3Q$JLZ1;lpl=MeVCU0u2(c|x_7d3l*NH8qS8lqg@H
zjI6)qaPeLp7lOrjhx;4hakvIlyM(~n=d1IsBpTGh{hLb{T9V->*3gY|`*XAUPJ^+C
z57bm(wIyekd7|+4m!R~tnSHLvo6M2@dRh}+US&JBF7J7hgvE?6MX3$?QKG;y!xqUy
z&9v3<{GpkVm~n)8A2~bw*CV0^YNwp9Tb;R)y}&A|ZvIT0H}PKVHC^^xecE>1GjdAX
zOt^@-mv)4WJOzc-O-wY$x==PI5SA<02IMYJxj*%ddw8iKNfo0|JP{->>ct8Iiq~&w
zLqc7(s3N6eCTuw>iLP2R&C{lsBjPl8neIg;@+App8Egg#hy-wG({?M{*3w*5g-#-#
z9&EJCHrR8`hw=H#cq_(4`Ui9i`M%V!p}8#lb`LMtd@%_LTpjL-Kfq~d-MO>b@K!54
zjut+Wp|npp6Nx~ks>n#1qSbN>R`yDh(C#W|nf<M@nUr~9Ax&ZIp6+JjT4br~*uKT`
zkVH3qY^$~iG3lX0jos;Z$lcH}?TC0Z-|d|oMH&lmZP;FBg}<Paz2aiULUwd@YgXHG
z+?yJi6w8WW4Jcw#hgyH?X|Vl#k_{emJ~N?lK|bYq?K1-+!q|kQYvvK#3%wgQ@~qqX
z0?(<)(W`A@N2(@hG*t}yLhRg*?*<M#M&<gcmh0G6lF7`JQekzMIB|>~&jMWVi&4Iz
z=oU+Yxpr1v_LBz_a#QhaYJ2IM?EP9GWxQyaTNv(rTP6;*E^IB;>g$Ufd+FMjBd7S~
z^A3sT_-NU5WuBJITjlD?k;U`%o>X0fMDy};yQ+5N%m<;6XvAy1OX*D3%1e|=WE$o6
za?(K_y85{~TxS!bY3@oYc2(^d>$QoR$+b5_hYGhTa#bQ-{T4U7p@=_yhia5l^UKOi
z$b9ZoE?9|O-2hRH=Fxs_U^!mfO9{#D?wiXYQ9SmSxx!q>Q+t0U`!BO<85+88ibbhO
zI=5YMUm1s@_I(qp$QIGcMBU2`o-Ov}2mu*kl<Ka03jMFeL9<k!Yf1dr&to?0rvW@r
zUN61c7Oj8Dn^v*&F))ZH5~@bH@V&A=O5e)jVEAK_83}4)TxFGI)1lhGM#r}GzmMF<
zU2b8+=Sy;4X|!E|ShJG(qy3jpwE5aZX^=&55?W-E&cp^b(iL?hciH@4xJJK%n4m`W
z#5YW?i@Q2fkvJ-1c9sd#2E29n?dU|?47w8O`Lnc`JK<n&ZnZdtiY|nylP;$-DOjsw
zw{f|T&52h719V52$yhoC3-6L0s=&B5y*|b*?_v>|nVWkR#QF|><Yv^bfq~hBEgUX4
zCw0{8g-J1U3B7iIZtdCkhamV#kc9dA$C4iz8Lg7l({*B0)43!p;>C*Vq>BA##n2e!
zxaUw-QIP}~OvlE?MxuVZZ+mG#%;x8Z5wT^+3N~9mE!9EZGWwZkjppK~oK&qaRF=;0
z&6*gh?&}lhuHDnnX!aU1E;NgwtYM*NMD*?(85ub`Il;fNa6Kb$FCuLD0;zb#`Q#{(
zRJAMN)P}LBmC<)GZJh}6Y_q_=M~&tyW~&jLI(?D@)w;ddzm*T8iq`w>v@1_+geCjV
zCFFkn+FFiNOM3I>&97g-D)^v2>%Thzfa>b%vTPt0w?TGY>k*`Gj<q^@S89kfeEWAL
zcVCHUOpGmUY`Soa4s8patlS^BQQiC@86Yl_=I>sk`4_X5DYM`7&yS{9WPqOP&ab+)
zfVGZ1qkQ+-7?ZjE`83>2TU5EnVjmlKA1y<Gvb`83?C|2n%I41K=%|+yeN&Pi$z|np
zQEPc*nJ-0=)ta!8XHqX3jx&!pEw`J^&7pmoo13$w_ZidHf9FK4DAk24?4oi=-!wV>
zR_MO|cP9>ZN%PrjsYx2_ika&#)aS*V)Q5ikn(0<vdqJszVUhd4S)uVHES;L5@bYP7
zW^FAsG|PZOSNGf@$0DiAF#ER1OU_W&FlWfFv#hKv*k8C8g=9b8SL)1JtV*fdM*Sdw
zd+Ga6MPZ~*_XM{$1lu?<A7Q)K=Xb>CcjGJ5S`P=zj_2}V)|W4WVE>4lJB0xhJi$d-
zdSh=oe145qG>kgT1Q{m9?C8G6QpVRd7xmF}LkhqZX-FzL%X@8tTgl`3xoeMmh{gTq
z#Lc23gfz3ieEHHbgdQx0VYk-^P%4F_PiP46UM}8;K!(vVvg!H?ldn~Fei{1iO@<hX
z#+2E0C0d}UJB&q&ii!d-8uu{c?(Cqst8VV2_VyStc=Jj53g@+jB~LJb*&PS_BqZst
zES<QySaX-QVia|mjaT9Nnoo=JX108N?kd~*&yPIPo=d;Rya@r^*R9pF?<e#TFLYVf
zHweM!ySv&%(vsgxHYg0iPyr&eGfdVn>E+mVj~U(9V9y$w5a}1S{tgo^H6To^c*+Xw
zKN7vbE=D?!zk*O-%LgGc>Dcqy#6Y4|JME?@1X3b->i?B~{CBg3e@klq_o)9<?m)UZ
z%eE_WE6IPwM!?A%W@u>WQZpNSKmPhf5TFp!h~1U?T6PG;JWqlINh&l)OThM;xZHs=
zp)vH@<suzROUrECq}ZQ7e_Hr0W-SinoBaI1U2ZbS1hLi`pJUsO!ye2gd-Q3o63r_e
zjfhw}?S8Vt4G^(&3k$ngxGX*`_UEj{{OxS$qYE|gdEn84AWk}Y)Bhy~kIVmS3}(zn
zjB+mQB#+$8s8ey536?x?un~@Wfv);Hf=XqM2LzC}wIXHY<h=I%oSmI9mEp+k3`Piw
zd`2bjC{Be69miS$TJCJivEb@I%~LZIsndt?URy0PF+mHIJ`!CW*bx`Es#CzJk}U8q
z%~NZRKz=tq7z{)Lj#2JD+g`U@Sv*c8H@;%Vckj7Wxs2C7dh{q0#3;*#hDy8(rf_w4
ztOgDT@1Ec8*O%N*1d)vqj|7l=lPY_^gA_?)@F-z%aq+7$+KlILkQ8kfdOn+$Sl>)c
zN-{JuiV-%gc>eHi7;nP6ckd)%tdPJ$Mr52%@wP(lH4Dt`;$3m8XhLd!af4Fxo0A=k
z=R#o$f}Wn9+_H|3T<UiQqqVaNENZ0W$J{h7>m|nvTQ>Y->-ZPMjf{=6%gO@!Vc^sf
zlsp}Oh6|)IF&~*}Iw9HlvV{*l^!EvJW~J0Sh4GEnv683WOLp%J!C=IfkSiIc=tq}F
zcIbLA#BRVM{T+6h{AzfuLS^Hz+_<|2VByP-YMZ%_WO$8<H>!HYivWd&P>%nrMDjnS
zFR82lFR?83$-sui>?Ebx^B_fXd0t6RN=m{vNB|d-k(I5)xONnCiW$#*c`vcFR0E{3
zYkA7<57s8?%N_c&tI`nL3#I+jIF+qV)g+*;)#kFZY<I~Ml4E0Ih1)W<qTD;g(V1oL
z6p>=qlNX}<^^^;@E+2*jx@2lF1<>~A(c`oUQbSx@l!*6;I*{g4IFV;2ExpGFIDv53
zsP<XIJ{`UL9(EuI-TTHNg`yqE$jiHu7#L}_*BZQ^kCS~NLE7F9fLg_+uNhKLzlQ>h
z=kdb=ToVV^Z~;X4ew#jWN}C4);r(=c?Ufa9LOm#TyC|LoOt-#?N!6}1zQC%!R?LnG
z%&XW<9p*J+>r28AQa<v?d;L%WBNcK^FjE<@(tE<!erb6bPF~E(LD6Y=>Vp_YuuYNr
z69VxN2?jWuwv3hX+g=cEW5AsGG*m->$RS7|G0#mMy57iac>~CID`&tofZWJD!q2iz
z-dWxPKI$=e4cupS)mR;uBrge|vlfVQN7#9Xz38E*grw}dll6Fjuh%9uG1Wh7i~PW=
zDAmFXzwd&j?s}R0RsaD<8r)Tdy4|0mBn;-l6->D8^Mljm+0w7v1Q`c|$jrRFJm=NC
z{G#@^BB}8m5}0e!cAX1#SHXcMUGYCcuEM)?ygqiut6I%vta=QNQ$@%CnbX50n7OvL
zRw>KCp@rkIX+m~(wkksfuFY>^3YhSnpZ&?>IyOoEz%4E1wKDwAONewy?7=1*pOQ?u
zJ+3?Ys$t*QSh6B~ZU&&IPJqziq;YTi$);e-8)6VhwB=-AQE@TOP+v6$n8Ex9>F#@x
zRpGn*YW_a_zB3VQJaPqwA;4saU)<D1+R|?ZNjVJ`Fv8$_F3E0y*7|#TJ^^Uo6kZ9E
z1ql6w){Jnq#8=F$b8B?knlzXfh~&ariHV6)&Z89vGa^tekmS0d^n+x-rCet$VPF~H
zu&nXp3tiwa;uz5Mk?3rKeJTLRI*iQdy+@LF9*yQet_Rhulyz*5;E;05gM}UQYNQ%q
zY%>Mr-P(~B-s7IR`S~3c4*h+@!|^(j^)d_&5`V*Hw_OPWdEtOod%K>U<l?A|w-AkS
zpG&Zx`<fBf{`*@He9+WBtF*LKBYbP695+|e%-*3Tk`fJ)1lTGO0L%6GL5>k_V~`6v
zm1IBl{cUbuo;I)#iR#~f-A_{sC@x@zG!7nT-RAQii_i~PGGt6MQkX+^`vH>NpEmb~
z$EEdPf2$wuO7LIQfi3meoD@(~8M1)u;&D83FIpM7^q`Aypc^^Z=<@veJzZU0{EZ-1
zj&8<>#>U2G85>mDWUfHlfzf3dX9{dGQkX!ek))rsezNfBy`vS$@B>DkV$i=^;gm4%
zI-!prKPs-2G;byj)m_$FrY&65cvGQVq@Sjy6$$W#1gyBQ&>1cFa_YMZuwSi6KzMk;
zfd^X&f$Y-IWS=e&u-(gxSSfXeqSg?E4$Q<(QB@jccQwgAEud=sPxOSJlU;Xm!Wn&)
z8)^HU_M^n&A{WA*IVx1gq}X-wcMyjjIOwINCHJ12tXWxEwjnKtASEbK|090GB~MJz
zE|%;{3+XEKgDul)w<ip3vuX$6owdOM1qPKcH&0?npi!4~+uxkzk#hUpc+Il5V*DjI
zdYwCW?)Z_2<Fm`=6!H*u!V5tht5Mvai8_JRV}OugoI1?C?GiT>>Hl|XXIKc>R6WJn
zXyLYZoMM|1;r_y|C<<PA`?2@fOHG&R$v0l4%8H6i0G3q!wp)8XYm{4m`^$?G!S%1a
z2!Jpr5M%(5gR!e__tfnax^~%wtH$D8v8v!>fSLs?q|p!NYYsL$z|g226I2XhtzKRy
z_L^27`EpZ|OYK#LL4jECUuR0sfN^Q{{o}mf^>xcpiQmE64NOf(h{w-`t8CuW1N7^W
zu~%Kke0NIH8j^%bfTEBfPQ{nnw9Iw6BH?(O7h}~o40H6l0!ZU=0D(MfkG{HE)H^n?
zteixdiPG=>!39fRF2*)A7jp$taRuh0Q7dFr4ud{1D5ob?eZ~7fDAbc7R-*mCAdgZ!
zQ;P8NP%&<_(h1&rF~no;KIJgbtnCgZ7F|xq#Ep~xjPc!?v+az(!9b%7EJqZ$@$~jw
zXF|V2GZRGJ+fX4r?wV;NAewXWHeqc5VokVUe%8#VCiP}$&M)Sq5r16-5fO96pBp1q
zR4vF@2S{R(!T{hpJDFH|8tJKQe767zBH@P3-Fe+(K$pKgDcyq}_cFhHKfa`S=jHUz
z*;E8^B;zxv(_ln-dj<vul&KXB?d|O#o38dg0Abz|BqksROUh69OtTN!cg9D6-6Q_T
zr3vPM{n<Es038!l5nh0PGW?=5L0Uv#ZT}fVO2Mo(D4*bxSb1m4w5yn9aCK`Qryu!i
za4^<4(HKlqal1$3gk`xvU>vv$nM`Jo<pdR9_ilyh55jE=y%|x$%^@5(p>UPRC(Um>
zmp(x#i3YjR#TN7de_|f0yX*%_FLE1Nnq<*YksBS$B|n!TaNwr4)z<M^M*U_mvyua~
zl?d{$?QDCDX`RnYMiC5zUl|!0k6nKVB~y$>g>44ov9SX%6~SDns8(5Br*vg`_Ep&4
z(n(*J#WZ)SEKH0f1EEAW<?sQQ_O7%H&H#di;^gZyn@c8GGZp5#XJGLCsW`e&zND&3
z?%G4*%c`~XrGkBEi2Ym&0Vj=9F(f)bcCQ9A`>l>vCMd$~jEk+tIkzAyTS%ZmWG}r&
zYOUp23Rvh3aE@$`1~GV#yZ)gP{j0n^@&tc4J|wH6qQc|HeF&w3Az{>b->GR?_>?_S
zxOJ%KWiUw{<~tk1h-M83vy$`X1hO#$GJs(J0-mfdco<@hdZds(-<1>x5@NFFpsAJ_
z+M~b{Q`OJPJnS0qyLV=hFL{SLvp!S7RETa-jW2M}rftb;6^qlF+%`4ZU~-oU0#Vz(
z4oUZHblm5++ECO>`-V1sJ<`5g>CaZuGyMn(JnSXKS9p6sS}`%{WODocT(7SpNZ-SG
zWQvyEw;PTplT2hUadqojjYU5a=KdvZT6b}ORJyGIDgsk{c$?Yd%XsaFal4LK@s2Xv
zj=l`da9+QK&w!li+jp^)&?&U|OaedEzIq@=zw+l6T}j4wtt8Xl*2$VqIhdi@wsEBC
zPt0j+K~ikdv<@nUK$SRi`Tlc?M`4#0sdmT~zzW6PVH;&zg+QvAy+5+8UAIj`EGn^m
zCP?$Lkz|TPQPDj$FAu3&m;riPXHhoJY_oh8+fjRms`MnDc_@^fpA>|>)-jqrU+WPY
zU}P)0`S}|yy`|wZz)9qV4Bsv%tEgjVF3$i;+Hp#gHv+&(2Dqu&pG$&|y=<8U!d89g
zk>MC`m1^hP^o2@E&2>9hsxxJ}YOG<y!e*J&OSPiabIDzf;`iH+56X)&*L3xMJ!b%S
zu_Tv|ktv$q%a?kVQltAKbfsXHn>n49RV5wsBV0RQ|L9p!2C1|~ZMXD(lAJhza7YYA
z6R<XP!%kt%#!sFtM*b&ICP+%o_X4Q~I#(B42h_d2EVguRjKlL!yNWu~<va|#uG${d
z5#nZRH2Hpsq+O7zzJ}{+$aPbvZ^mIe!%+v*d`0>RQkMkw8d-OrUkPNsrAbn!ugk7N
zqi0j#W;iL4Lf7;6&CoBS8F5>~F_KQse+F!&+f^CixNurLYi1CNeGofuHSjB0|K!yU
z!}4$tYgz<Q{tSOyT3V$$!r|rNJ4Y5m!-%hnh8^MRv*(~31zKSXDc7vh?o(`}Cf5h4
zINhun9$XbRuf=QoYeDyX2v&a0pNe_crKh}QCzrXSe_dIAE2brDx95f>w1?k^GOa&{
z&L4_Kuq`m~h9;EU>pvqT&^FYtTj?xV3vg2!eLoGp@q^2IG!Q~5TwQ&o=B&P4nf|`|
z>23Gku^DC5{RWIMw^Uo^z(0Ed$OaEY+K#b`;0g$n#m1R}RK#}IC##@$sv@hM%fDZS
z3~t|}Td86^VpRvq%cLj9k5kY91pDuA-cM1IWt@W)@?fuS$*7nCiBf^y+nKB=tq-Mo
zsL?jU=K*(kiXl*PP~;pvo(tC4lKT3tE=(z`=Ci#Nb)e8b9aVbGwCv)360PlyMr4bL
z2+TKII-BM9C6H$g)5woR#l*@1Bpd@g246s0=oJF6sY{M@;(1`4KlOVUA1O_FFQ4X+
zQB_~;>bcqQq!4C);g_`U^u;^b%+_4TSu4U<fzqIEmvKg21hZU(DO(E5=}^7fQr*O-
z{^Zbw^eEj_6-DaZRUQ+erOIpEGOavEncjhxgp`$NLp`mvn3%REPf$Fcr;Wwug_z8L
zboIBX5n&q6IXl68va#F9#ALq86(wmSQYV`~(JLt~vZL#$w0RSUXj{sjN!VGG54e+!
zoXzI-U;j3ndek~iFrBHkaT&Pq0dbxlCD!0!M4QjYc4vP~^72_plauJEojaMbJEkgz
zc;<%FKx3+&E|y)R_n%N)8(-_Rg_1oO06-66704VMcw|9PQy{>@g*K{-Q6Y+R&<%;?
z4(62II+0yfHLR+wV~#36WOcQde_>~NXg-*$eXC{%v|6xZID59{x;g+fW?(D;)T_p!
zfVY7hm(%}-836rkH+4zVGEfx@>%7sz8^u63F)@<R@Be3>CPgf#9fE*#1S0jzADj%q
z<FamIv@ih1j^>!9ynOemw<2?89TFXYnkczyLm&&6kEgwASgpcKJ!yK)R}{`8F9AR&
zj{hWJf#XJif)o|$<yK7Nkc9;W_YDpE|H}onwYn|$>5uY6w*a7oaVvYR{PU7(PaF*C
z`-Yt_airYd9>5$27ng3pb2b6Wu>*yR8VV6p?l##}6we8Pv=_Ol4-E~?g9=MKC=I}W
za@z}Q-3{kY5$67X5S^!hE4cNf-%OBoGBz%<Nci!d%N|H*4gi{KXlU5u0EZmco(x0|
zVJ2cg0s?SH2#8=hSR}RfsRUSR>~ZW{xBqKigs7v971CWB_UEcNbQ~NU7T+ZRe1aFP
z3!ZKlEAIUIFk^F85~WBW52D=}<Tg9oE@|5T^hJQgGXO<)ou~uoWtM?s`)adnz}P8j
zKrQgl1YS~vAuhdBgg~s*O3dR*@h^6k2230LeE}Vf0B4)63Dzjf1>UOzC~3SfgS0X1
zZ+jIl&rrEQcR?zkX7_;?NY<_2vd0|-i@OUT6(e(%w(uu`TrVwc-`)A?q6ULl0Y%Xc
zBp+_M7mrf|c1PIGUr!gCV}S&gf$+zGzFWbP{udchb#$EmM!xPMqxl5BGhi~ml}7h3
zs^1;e<2eoZ>wCaomkZH=hNX^7l39Vez{V|g8|XbHFwr{J+S1Y+0Jaf{04QHGuX+*w
z=5G$a&b@%Hn-ymq%f!K)ysI~L2ut&%?Yh#v!dZg?+D_z0p)fgPh9$**Q@@2Jo*pj2
z4Hv5R{M|SNVv+q=^-D#N+!$Kmd@4c$3jVeQWKI=2l|UcoEWedwioG#1Jmwotnj$|N
zN$f_buwUarXsWMG`|Tp{XgiF|zKn@cB{j*Yn~3rDXoRqLX>h1U!@Ctl=wJ?mE`?Q9
zav;IuVDudrh`&9Y*;LW}fdhAarWHw=%K?0rpK18L*K*J!h6u@MTw)mGUo0mW`^r!n
zL6_50M@L5`#9f6y`F!sg5EyHbKagviGN{Rcq-Q9d?Md7FF0#qs)vGQhj~>ko+k|(4
zL&9e<H@#ZwJQHacCP6y8auVvR;JYF@n|dfPY^ldcAzLpc5tPjsyuA%|(1Kz}wxAd`
zwg}8(C0xnlh#y0~q3uoMkwt}t<Fjkg+BCJ(pmRw-<Pdo&zqosEW4as^dg}JOT)R`e
zClbK16r}<UvNnaC#RKT0i~^i+n=)a)o6m|F^mOfd2SeYkgV%ca)qlhutA$^$fWa$+
zebbTx)(pbJ<;u!RZUr}MJXPrKhgRUXO3n(vd0aqV1H@m-eWsNc-g%>jx{Eh7-QUzm
zMi6VbRsB3tk>o`^zJJ(|K&Jb@YAh89=}99K%NPbBjkEMJ1;SFVuOWz2@(=geww=-A
zu>jJnxCH<^0HF*4@;<m<(Q#I=gdM%&4GYu%?*x_){HjV?QvQ4EDx0lB3}?z=XK<N2
zlYx4XH-nDY4i=c2*0|fAle9_(Dp>aC&lgXgJei^@0Q^`C#P1|fG0e)&zRb$!@c`&y
z%1)QTN~*k;qmJ6+Qd2%_gv@=?U~dDua~$9qQjF><^q?6!18|ScKRrAUx$rq;6R>Bs
z;(%2rcnhEROs}RuS2G7R83=lpwZfE^zFsq{_}b43yqlQ5elAQOMea{f3n0r0WV`oi
ziP#?OZ7^7v8o4)o4tVg*oRj}pf^-jkjZVdW;E{I;S)OhQCuzL~c7_J$%0QBpPD4bu
z`kETW5MKAze}?ivBoG*q7?p55RbdU#q5Ps+(KrittsFK+u+<IORVR4#>w>_k70HmE
z809}CZ7Xeld=2UShC?6~;L~f5Oa5fk)2G9s4gKL{An*m40{d+ipt;F8kN(<)hx4m$
zg4)4+c^Y*^u4NZs32;-2`(!Ye*F-d+g}NSB(!HzZxgijCVayJrI_Q>u`aiD@t~^u+
zRZ_LTqh{A3Fv|%SmhC#<Eh%7jfDYP$8m%)a{QM0;_*~~=#Uz7{0`>!>fVS2MAz;}p
zGoa$2Z#&z;@se0GZwFdJd_m+__Phs<QbK84>$Q39)H^XLPz$}ED05ZIz{qHz%4@*L
zWcK?R{Z89B%UEt{JCA=%{Hz*t0NxpG3*1=QLI5-ufB{qgsRl9u5hld|<W|=I|1;~q
zmk_&CIY3N_`4=V7^S{Wl1fZkT0QgY$y!Q*ARfG+25QNM>Qebh@%YI$?*-mL(XAp;4
zJdn141o9FC1ELueh)rY|Z5Fn<kP$xj`i$PfJ6~`R=3+pD8Y#2&ofU(PaF#<rb~aH+
zCv0cS+fLN^THZJT+zf%K44&R8Mg{y$K?jPFo_@P4HRIC&Z`8K@Q_P@m=Rjvc`)Mmf
zy7e0iE9zzf8M~~BiTYOGUvSNTUU};khyS0f0{(75()uSlVBKbjMgX*|xh;bR-W41J
zFZ}xYkkO#<&Xj=*K$v7;8#N%h+X0)lya6`G!zmc;#!!Y6<ebfbjYR-4UmU1}dbYDw
zd^Di_lp7vl$h4V&!hwPqu#WqY&j#|05dOLg0Hpw-Qk67;aMjK-m{@-!|Ew=lD>paS
z!8{F&2K+896LgR;$QC1(dtnJNe~YqUA&tx&9!n60SV;!TUy^HsBIw(w*&TKIJF7>5
zuGUBfOZ=}CF3rDt?=Ym-+;#qi?q49s!49Ul7-D!{etthtF)iG_vH1Re#aae*kOH$x
zQ4#@#VwBZDb^-n~_tyytMmtGVX2%dicjHrHrlykWM1YRBzl+9>dm&63sH?7PoAI_C
zu~{JV0=d@_<GMAw0d{&5D7heLes)&YjAY|iQDA5^a5y}!*u==FGRYeNA7_C-`iF<B
z$Lk5;00<znf=-P!{1@L2JYtC<0RuFQA18rOam_x}*A5(g0%%&WYyEU_b&%0ak|}GA
z@vE<@su~|<Bu7vZs7eb55%X3t7l=EskVAeQ0l1WUR=;u(zU=^T1&WO>sI-tO$6=P^
z$nS+v!Mgk*G$`bX%&s?c=tmZqmR|<4Jb^u&1!7Hqw5R)8e%!-QqHeuORn1NZ6n!P(
z3?tVNAV1Jk+n=eTumaH!OsPuQYrNJIwAkDOB_WJ|fD;3R;azYk%JLu`dn+AW0EJSh
z$MZ9;VEy=^OK}#Fz!<zp&ij>5f4pNz$9(+OV1E0(;mGJ><W4@LQ;~tGSx%f|bRn)b
z+<AY_cBVA~++tz#@49lxp;}Ov%>{rRq<&Ca?FZsURT`SMM8KlDK_QA=b$@fV0}cwS
za{%-mc;WU22L^0`eb?*^nlpMvK)H9!t@$MA8JV&&Wz=CkpPy*?zZb1m){mbJI&!kz
zAtz<<vj__Ci88O%QAScpssu7h4wypCcApNTSLA~LR1I{-v;s$rpAJ(lTCFa!0>g?M
zMSJZQ)vTUZ_Ig1CCR&tepLWBWIF{_c*;cbR^&U>$8Hz*j9<SZNM!~uyLEE$U2G2hs
zl&kMvn)?t;Y3<~w4VgcUkSbN#&mYd*u`J*xQ*KfuN;;1i7@;n2O-B@8Bb`?s*!*rY
z;?Z#uv{e`9@3-a{^twF;b=$j>C%E6+gM{)-?bzY}re62|B`Ti(yD{c}cn3U^0<Ay=
zg`wX%!%_dX%>-!F`6k8Pzj6&hF{!SL>^Wo^2a1Xebq!RA7*5m=t^-7F_je&HwjFuI
z571{uOBWG{-Ohvmwl@$0sWRRTR0CC<zdc?DM(&ey!l_YST)1$-01$aE0;sV`|D$s3
zy-yxwKw7|k>_Gi-zdw!ki%5b%6s|`HoM^F&iNRML^kwNVidtTz9Z-GK0@P@Db{Q~*
z)j_rOR~#VDGMeV77!b7SwYutS_dGo*r*yLVtSkwANfX=Ib-=H+B4<0|>_JRZPT&P3
z<{_GW0T5xk#l9@iGz0bw<OfD;iUf=>8ltxSxeCu{EUO%{!sJMv0>S(K2uXdd^WRi}
z#+A~SmzJa)e|;VETP_6sCJdy`C?SDPoR|No1=jAI3lsiN!jnsJAt)Edg4r|bYxsJa
zJ4k7aYQF_2S>dVm|A<|l%i_EMlaBw#S781J!$Mw%DnPRkCl@_pjo5xyaa8-enPS>y
zMao*#@%g=l2GAGxfL}K$lA|7)Kz~%`V7MHBCEISo$|EH$lODZ$UMoVkI|P}1x!k$C
zkz6r>JSd$BqhBLfl$-#a;SkH5(i^7rqNY22LrX?&?J;dj18vUes_^+PIcII;OVnn#
zD`9#1mL^z>^-!YRB{A6X-$r6lM)~N_N%2Fz1MRz|W5d-59QsMXd^bnRJ$2?rre^^5
zn3g`h6Dw{W3y#xuD}W|2Cn9-Z`~FYl)pznQ7WSfmz_Gm=$Lr<OhhEcxQ6qyWm#O3`
z<QiVShJ=lTseY1?b*4}<VE7tIUZse@BY^=AKn^r6cw9Pubb?HYMVl0w4Qz<y7P)*J
zzgg^~_?JBE#<zedBp8VUxlikQWzjF4`*nC{I?cmfYX9Kb>S!F=9l&&Q^}POAJ?eT1
z7g*4h3_v{?R)>F0ENZ}lCw#~-fM6KilOcRwj~RL$<2lGr!$eXz-S!TDQNfFkQ&#m#
z=bAA2uGpEmHDMTCdlOHnvieDk#_CD<!VK^^%c)$upTL~;9)N`DIFOf;N*$jdfku$b
zN1~n!ia-`Mgu&7ncKto{)M}Hr{lhU(UVh|p{WXW*LhoRGH$ulUU~>^ite+3k17E4K
zJ^{i25C9g}V?)K)oLpRp>sz4f*tRi{iJ^&jpKmZH3;{m#s8&uYIwkOkHTDs>F;HIF
zUEdCG?q;uAnmMyF5Fm|8u?H|nS@lGn?OC9x<$%SWfT{Mx02>;g0M$yRx=T+x);_XG
zeD-40eL<UsGvc-K>E@lCokdXP<7#kER|d3dMHZy;6~=6^#W2E#(;-mLRc2&yMz>Ul
z_^Bi;V{kxcE%(}AQQrn0NiPAw#q!PCF3}udsgSe`@X|Say{PK;3t!;Q6X1PJhh-X(
zITdA!NkrnX(b+`v#I32(7}Ewv-`HZS2;2NP`ya-HIPi4VnPBc(Y=*9F-g;zWWZ(FW
zDqT%mXFJ~q-C-SKyzHm{CU39dp;V4MTwp`t24aCb2!E>@0KR6EFvxHcxDhQq`kAv~
zS>9DA+TQ$_G^P6*p+vBQ2iCdylhd?!?`f>_b85pZPE6kGz}?n}(<pb%H14J<N0%5Y
zUmQ8rlkt*Cb=i<mE<r^MV7K9xc+g%xn57@CmfNa#tNKP8Eb`q)qs0J&W`24GWWf=L
z)yh{dnGPivxO|)(TuX`^=rSLha9_<@%TF>x$+xXV)ybB7?VaV!y=RF-at|Za!wQv^
zsKD2lBw5@7aJAcOBNwoenlA&`mwH8FI4!-8RrZa>%2<(_<Vd80#*1&o&`jgti(m_f
z7^X5*wAz!^&;NMay3GGa$BnVRQjURpk-U#mcr9XNwzb)d^vK5c*5OPw3zGEw@sUY!
z+)+~gB0lS49X?Up0qN0PMM6iLK|iynCjmKKxpd#G>5fE7)xKAR)3KbYJlR0j!OH>p
z_@~uW5`SUo7+P}~$7*!R2>)ibCnyZZDS|9LL%-_L5Nj`&y)o*HlXKMv4aF$hZzvpy
z!E>O^`aTGoT>33mq_N0n#AYp9Tlc=UHnA78pAlzz8kuYW#t&xPk3uI*q_oHz&C&%>
zmlh}bqGJQDB5r!Poqc%f(d~yXneOm?<9w)@Ot86Y5C6u^i4`ppD*B*Q{pWGfLx%*b
zA!Tc=f=&^->Cv%7L?neW-FM63vz8X3e}?K0%JF{0t~TYzczr#bYD-+@YtVhKrL8>$
znCQ0S``OG);kzpVuUyMw_t!p-emln%DAd_fdIu5b>a@$wU*o3$?WUXxY5q3#<NYFl
z&U`shezKDqdwUvJVEXvkzo22Sp={-HajjRs2A+#JlH)J_1U7U=;@Y)svh0_P46Xb3
zGr=KQIhR^Y7$;S}c4j?<8LqtL;=qq94*2*o3Hp8ZWp6PII^n;i+pV;GEkENeRNU(D
z8MQ;<tYP$YGEre+n7ez~)>J0K@FA19VC>5pi)XgB`xQUj?d<Fd2Tv1wJ_nhYUyFkd
znl%lspXXP@Z*bDBc2pMca94ymx1lCbAI;y;Zt9GtG+t*@;pcEcV@4^X{2ll*W_S3v
z>y6WL^XUkc7MC~k&yy&_^-lsRuWjBi5i>|FP{Oq0AgPmkPElmj=l8OMX!<4V+`x0W
zHZ_a6(VVl9zlaw38NayEcwH~AD%{p<xVw8D@?%NK_0wn0IN+-a|2eCZE$!Q$Dkba1
zFK`oD@v0wqlq{NeR*Mc<Z_czK#ZMO^8=z!4HUSAa=g?ico#&y5C<7*O15LK(=E427
z62JD0@4^*RZ*t^+haO#ds2kp*p{Ez7zZDeDt9T3CcnI6M>ZSbl?c0jyy?+tMF$h(C
z1qlhgV7E`09STBAQi(0FWBMt%XyiUYmgn5#q-769UiK9)L=0tVw5hrI?S~HqyYcck
z(%fU+`TJd+Q|hJ(zBgg}U3RQ3y$H;>1!tBveq$b=R~19Dff9#jaMm4m@Y}VU&<mL!
zm2=SeP_9Sn?|;2I=v2K%b34q6?iNx}O64C?dnKfjH(^D%&bGO^=|B2ZQCd2xpx~Mc
zf4E8Ulk$=0p!-8hODi)o(+A8~O^qL82}fq-_0Frd6XI00KIea`UVnU~Vo?*9jTM>W
zcstz8i<A^z&ba3sDnLeBp(B_TCtOS5W@c4?u@Vc-)bTIe@O(Q_H#5(H+3ZC<DGAu{
zF-Fql+uv<H9|(X1goKO8l{$-Kq=!!L+AdaXwn)TvEDya@N>Rg|@lr}r;0-gYmyB2%
z8`v>+S#UckS&Vv*mTSOOSQ1L4z2$9i$e@m*o_3y_xp1!YTT&nPp{?C=_{hlLZ*%^v
zNvRyUt#^~9A~fx+oq%;)3uy!Y)AiW<4_PF>=>@Mv=DB;ve~XG|X8t~!#iDto8^?Hd
ziA8g5OTSvO_1WE#Q=9H#56sM#Jm1{?cp)M_{vJ>O>k>thqoZ%%x^?Ku6Y|49UtLuD
za*95*{F1op#j0O)q1go6ok<Vu0N=L!12oY|UA=nMSo17OnSLl<i5W@hN}lSijq#N{
z-MDRT_?@Fl@&axXNBe*&2z_pU;xs<FtH>y7r1p2pyD!6*#$D`ABB#!A>li#09LIQ*
z9FcDOij%X;U*E~Nt@2db-lBW^tal=O4t}2pYR<Qnk@5s7+v7mcsZvVaOX^~Gjw;l<
zI3>&3BZw?{+6Qxs<79de9rcrzECb=4e0clkrBj5u6A`6aw|`zf&F#J2aIid$pNb6q
z;P636<sB>e8%xSHe?tRT;tR!3CsJtUgleN_XRtfAl@ncvQvY_T%K?fBCRV<|IubSC
zm+lD)il*N=m@H$(KKqdw(yFV^l5RS$c0j8|^^!aDMYG*PlRE_xVJ@Dp1MSdwhA7Vv
z<1WnH@fAGnFVlAKJICk&@q&w=i?RuI6=9ZXam#P%bOVp_Cs(V54p$y*p<i%K=-&Mw
z<)Md<e*}i~zUyv2(WEDj!kOs|YklLIYD1SKY*3enJ^|<VW9QabmM*SP$_fn)O}-3v
zz%%_$`=yF|HLsg7#%0LC@)dJvR9!RHj8EtG?b`?HEEV_o97iYpi}tqI!dA?~%Y(fJ
zPM<n;><C!Y(#@y%^1`q`1qB6>U)~?R13qxuKmCvy<MBHyc*B442O|Yx;XA<^E!oC7
z3bJu6ONYS#27H`7rwxv4|8Blh;uSui<yfVBrtHLIV);#;^Y8DekJhp@BpfTZYupLC
z_@K&M|6bsG$>46w0zJW-+AI~XKcp<K6ZStPxO$#m^HEtJa2$?3LV_Lt_@zKkSkOKC
zNh4y~di6r}lH)?aDGQ&7f!r{u-|J6&CKVSD$Jhj(c>a0p7KoG!+>K)&uiI9dM$5@0
zA%jG6zA;nN&2{qoLqj1a=58*r`%o;8o(Q^#ew^|(@6}Kfl>d#(_@S@Wan_&TAHDQ-
zmaBQ0@Qq81$8N(yBIggQ{d0{vpC3UW+|G>u7ofq)p8X&+^s(b0TQw@vRE0G!^l0^3
z_;vW30WIO9EbHD4t&1b6N7lMS3U1{`?S-@i77sO51<gILkp8)JTkv#QUtf9vYrf&H
z0!!9=%FjKIAG>woRcyM=`RiNuJ`P^%XK^ifzduqF*|95VrMIE4uQfvuffm<&M7?4Z
z4dk~|F5SF@K5CW1+&>sERP{|1eUv8MQrCJto{$q7dhbsVy>HrVG5I(q`U)Ip`<=A2
zJ|+Lj$lCM5sR+kse6N_Oop}ANe10s=aclTMKn#tQ`S}8;G-Zww+_}wCA>+f(oLwlZ
znF@bbpl3+;;+29byco9kJRLE*FYC>rF*l+6J{*@m{=Iy^{|hS8GSbHL+C~V=*0)(H
z*C-JqRV(C54A<w(na3E)b4)Sy?XNSOJ$>o6TumM2PwYaA^&((Ti{v)<kGskL9=I#K
zz0!?u**9cepQZ)R#LPxL^ncx!Yq+RPz3|TBhjB|(|C96A_oYL<Z((UAp74QFv+lvc
z>uoKxW&O;s+`!l)PSd?R5$G(_6yfXV8^?67+nswR^Qr6UZGPAWn{BF;MXF84$4f*5
z_Otv`<6Q!#T=(?|G{T?MKgQgI*lNGMFyAxWnV)zm@2|up#ZJkAMkr<5u4I1O@cErD
zd)+?BR7Tqt%l{@P=Ub@G?=dgqoQMYx&CIm*^z^dKU0hsZs@|EK$&KKpOyNGNF3px%
z*wfy}pnMkX&8m0J)?IFyC$%0`x|8O&+&?bIV?@tSvF=9C1<=LPZN;`WGWDOVe>B3?
zZ@I*iFV$_Hu3O(+BzY%a=c&k_q*lE9pfLL~;tXSnm!t_}VF>%$0j)#kPp;POFk4LU
zxu^;>@fwRns@tAF0-8K!&((=28&6nSSYQ^<$(iryd>$#8{16IFL0_(v$wdrZz&Ttv
zclT$m-a1&A<ew7kk@#0#N4R!^v)*?_%nUk<rU{&4qWkmSrSr(q-N{tGoIAG<=W=Zq
zY^c87y9I3G{PJ!i>-+Td%sKdv?P<Gvs)sptL+tYI{1x`({LK73MP}1{{Dz4qm(xI_
z3;K+JKzRRRCTMX~Na=oO?u+n5Z`yneyM2b{bxOa;@6}7JzxrEDg#n!%6LfR6cy|aI
z^9${4{`o*-ASNz>G)u52biG0l^U29Nl$$Mj>nqR8@2QQ`ZZCa0{`{Da1c!gk>4svv
z2e%xnM9!q9vU`()PbitC+-<KJC%$`HmEAiY#n;i*^~3PJ>QZxaGsv;R!or`DlPzpY
z&Bw*%<aRSNS|--3wXasiY<{mn*-X&xyy%#?`4)2)(-k&uH$N9P1h)xsavH}ze(@8O
zAoLFc<z{uxs~vhFV)4ed{M#A*rh4z@Z)XOD3UyI}p7&TP4p%h$UU%VNQJW<k&$|4>
zcCKp9oTIU1;*ZZ)`Es%f3JQ!_);3`C^Fu^!+}E#HlJ!Ko5@j=V;w03<50p~$z;_At
z^7nRf;!B>NJDK|9^=jHy|EsL;Z(7`RHz7sN@(`X^5fA+znH`7E$73Mf`=^?k!}&0D
z{(Iw0c(lfli%}ZB^o#|P<zO!eMLv@&#K4J#qbvgAe|?e4k2)NlR+X^TBCx_Ceum-D
zY|ZJ_)x*oa^}u2tJ;JcGOixX@fjtrt0bfQ>9p*oE%AZR)c%=D!@2b~b56t+|l_`$p
z!NAE^`Nf46D~HCef4A839sCf*RmVX+IoBmkS=R)+0CLOWWGVUkX6z@b%u<9EsIe?K
z|C8Rkuih(}r~E{`dGg@F1GiBjtgUSWO0az71?L@S(3%?{>JyGyaeKZ1yVb~2|AaQR
z6mE&A_Z1_pav0$p&T?#>8!UoG_<cEKc1YVSOwrZ`R#NcM_52H%O6t@Bn;)Bh=aLMk
zw#7O&&3X%2{wb_rI_)rjc+SuLh6&#ME`LHvNy$6GDrbCE<%NBq;r(zKI%r?=QQ;pf
zG;aZYJ?G_KoSzg4`y=&r#M~E;aBs}XyWRH>hmiUeSG<$~q+y^#D({Bw&k0pC%}2Kq
zf%%>9@NR0+Nf*AC{bI&`#jfo&MX_X`Rq#uS2gXDA)QwlNu1_6ASK6=VcvLu8P!(-$
zZ7Wlk9?%pdC5w&s1%-rwHOheY{`S^lbg-L;heu;4>7*QB_%pjnE<$b3JOBE1=R7vL
zPxx{hMg$RPAPN0Lbr~PR6*u`bVD9p1yzG;yXnlvlf&_gWmn0{bQHsH-=h`~;v<D(n
zCpdea));X4*Rq{+21x^T{!K{XiJzqazE^7>e?C$O4R}MxMNW-a8sB$#$ZmT_G5Xvr
ze?8PsOa^aqZua`lM~g(w`9CIe&5jOvHMqWbaZN8L3KWF*9<B2e&T(@u&ZACk#C3Mw
zKLV2t^+q&kr=6SEfaaeuva?>_2*Yvv5Bzzq*y_uoG}3E8SN?|g^bxrD9K?>|%U9n$
zaG_@kC7S-jfKuN1@!_t$Y07t~d%vOSW^X-!HTQfT2}#s0Th?0nV{25RB-xlv9G=`M
zHTUxGHBWY`;^a4BOUt~sTdnPcEX-)nNyfjByCCcM8blb-_N*rEY>d37rj}ZAo$7PU
zZmoVx!uvAe={X!z>zj|0mT#o&+f<MT8=6IRo7(w8EHO9qRlaTVgna66V6Jd<_Do({
z9nn2_&EezM&6^)}?~guLh0@M1t8=aJ+%<&4VczCxyDg65ayv5CE$&=z9tF}}Kgr!B
z_*9EfQ_=o)t@MjL+WnQkMZ^mq$FVEu<4NN;Z&yTxJ{P%htnh7Un8?Ax(5YU%CYvw>
zFCS`b+Ge#d@y7;T0v)w73!AzmQ)pWCCPH_uMCtBF3BBsT%KV|PfxG3PQ6@xd`sd4`
z)(9UZucnyz1&aDjzBcNT&J+s8{%~aEK6zaOv`Mq&3u*lUa{W$GCfns}!M3@^A{%KV
zl9t*4Qf_gDcKv+SMSxpxk<@l3kG>ZMr08q4@CDh`=jSSJv~X5G+z>h2*!>RPl+pQA
z;hNxF1?nwdcaXh(qIn=}f?81<`RAM{{L}~}_Sbv&+mV)2RM^w+2xeBRE*c-+w9=6?
zAtAxICAXx^{5TM!euHma^&QyfN{gmvS*<EREh>wc8~;H%`36(b<8OWbwydMJ|5nj=
z1@rhH&f9PAxcUzqQW$+Jbj@v!Qad&WW3A$UjgWW~-&h=V`}?6&r%r)-X3Ns~7`6*C
ztxHEwA9|#&Q=fokxm<jG$d_Gii0X6+V|;&VJ^}YBDOFqR+bJcOhrDVL_nnOFliu_d
zS<Dk$uaT0HOr_FCnX}SfxXB`4j!532H#?}uk4rOtPgQVQ$Rn;?J(gm6V@CPK@IA@)
zKWz;pA!>Cig`#+GbfVkkV2Lee$0g@{htsy*;%0HB5gm5%!F}nw%n3d86tW*eQ1JN^
z!(8|s+qth>B053kK1WLM%?rOMdM^H6C%D4*LaQ@(R~{cQ8Ky0?M~1HnN!r8>oN^Ku
z5=c#Bfy+CNm~3==t`=nmWO%J~j(0UN>xhjCz-Vl?{9MH(--a(=Q)`{hffwM%M*aY;
z4J*oM8G94*F(FB8lq*NP#pZmpuSK=WO}C4|p)yy%p6ji42j=7?;N3$GVN^eKbB%X>
zny>03rE<L0)wZL2*|^Ua(=^C3T@YsC<ig?KaC%~oe^aMGPv}**OZgTptC(q>WUK$w
zx<t{5$4OeQ2jeU0Coiv=e3I>_%bw)hh>tb7(>n!f2hu>(Fe1Z;c`ZisIJVe%@gIpv
zs)YVxBWqp0KO%ZQA2M6&Bw-u3q`Tht-TFAre>YC8nR|1^n9MGq|H42@SJ}eav!K5U
zDXAiR(u=oD>FwcKtFG?JcXDxMmz<Qv`iG?ac61<h0UJ7;^O<uiPhm79ZI~)SRBpED
zwE+r(5!s!`<w#Y1GB*ax()n=l4r|<v>-+Egd_)8Vjes(O6}$3E9u=t7@okpU-Es2=
z@o0_?;ndjnm{;UzijyZVwpBEc%f7i_C$aBh&aN?mi>Z1w1S;i(^JZIg<Fh=YWWr1L
zXgLdUS(J(2N=itwtgHG=#{SxN>3#nz+45}l@sSiX?u8T5%JL`IDaEUtJyRyi{gaD}
zB`$#E9D3iyZ-GhoRk`jddtlE#Gw70W>ll;swAt401kxwtc=pDQdQB5vIFHN3hz!5|
zZ5438Gqa&Hqh8FzNpv1%xVOoo*pHE8fzBhj1!{F+*}b~=Tz+L+O@xHDz6GKJ?0n<O
z$?J1Gt(|D4kH6wzyUm9EU9-g7d0XnFNr=Vpfr{C8nf&K4gby=I(JgmvxWmCs&C_ZN
zjq-@j%6_P_l2kw)&@>{aM@#m8AD<&8wfx>?7L&Mr_EfH`X!>jgU_8S$*KK1IrLKOy
z1Q-wPT!Yn#H*+4X$C_5=9CJ_Yw!eQ%cU>0Y4DT?Ryh7T|-thU52fcp;EztVQQhMJG
zN3xh8M=EJw6Eu)6WERh{MO>&;7AQMjs3&MF#kS|i_d+L#mHdi9fKIo*8^7dVH7&*~
z#~KpiD$@6|pqi+<>L~Y&)`Uw)Qb22kYvY!OYB^k8`~(8wi3dei((glg^@Vy2#NIGT
z4}H@qH&W@iVb0MM)o<PJYAh0=zT5xmk`EgP+PkL4V1sR4orGc9vtG_IO7-^YGYzaM
zuv{LF^%Qt?dq!C1khogM)xStwjdRS5Pez-t6?X-{vTyzb|5g;fc(G6;<JoZ{-XwQ5
zuZo&q5yLa#dA=FLp1-FQuvhi2McgLQ#LsHF^>x&)kf1>vI2e67fk4u^=d&z|vdr=I
z{9c*gnB#6fEf{$FolpiZU;n=Fh4s`i-3vUYi+$MQYmZnBl-W)N@r87LG(k={d7j=t
zrx{ByvU|SGuh!KKYwrb<l+e96WZu1xWyiZ}y33J+q9>*>f1QS(SZ@>S6#sTf;b2Ff
z;|%vX2hR@|oW-@X0&>`mTmFdnGvkn%o<p-5{P0Ue;(%7>XNt=i73c}6TW(GyYY$xO
zsk$mE_F2p`bH3Tks%Osgn2P<s1MC&S)6ZmB&GKe98lwlUduLQRouiL<4v+cRe*e)^
zUO(V4<8eUr#!g4{hWD+t@l>OmojpD{g^$}&e_nQJZZ4}Zx~+NVhA+WGZQot+;|19=
zrP$EJ+?+iIckI*-yc=r7(V<w*m?k{*7M*x~+yCf^BFk=#_E5^&Y#)ufcTnEeNvLUB
zQh=O2$9ZFtqVGekEG?+Dk=0F<DOHLLJ$w<2)HbTRJ}C&F$h&8mh~_rn_!dO{>{~bB
zkXfbVx%B-j3i`zJgSHyY|F3~lw}o3B*ef8vHyiCcCJO}~spwEgl8$BMG`~wR76GYc
zBlT{2@U{iK{U?zPWc(|l5<%Wvjj=&)=B}YV(d<ACaT86~7|XL<?-jfvTCVi=b!{-d
z)0c!=`WrtB2@1NqCpMk>(mr<m$*~VfIa8*GT4>kKQ)T~;O0GN{%C(J;rK|^ICzFih
zkis~Yk0}&`5QVZvmQ&PG*0E$wh%8xJjqGBOUG^4(`3}k0O_ZH1BTJ1feD|F1y1u{9
zH~-9az1Q=+&;4xg^W4At_q$J0o$&ysq@?w;`>`8$DRa$-8UmzYahb@~p%kUSHt|uL
zkC<hFM)Ny;FFD?J(j=P0BnOxGk$>XJ%hob((IBxJ(;94IZ0x_4!2leQOWU?sC(#<i
zP2c~iHY)t4$Mc9AOEP7hJ6`W@H_|N4f8J;4^1-p3K{$si$KHDc`XA<Q;@iPjGu{zv
zbjXi=B<q?PRfDho)XGqkTK_@&<clti%?`JQReD`)f@zxpVmWtCUh38;b6CjI;XC8X
zNmifBE3m1vo%2hcTI{4<-Ana@aX8$s1H=d(AFjIhZe7^}xA3X&QYwShcK;RDh0lB9
z${SySHItb<v%7IS$D45U=V}LjZ0my()6-u06X77I38~PjF{xsh6!rV7R%E_$=S0h)
z)m|5&9@$UTx5Yp2Xo*syiPien{B|#lIx+{X3_0W$8dk@+w<bkO(g4knvt&X=(XY4m
z_d7TE2b5SDIUeriP<WMOp7$)uN^l<a8f)-QGi>-ns|5qvIVaaAm~>3VQ{}uPhhidj
z_}$L-N3|-ohy-D~CN{d+y_CHiSIhqv>}B^$mXnVTn0w!LO+Hh`?~fVMGh2+Cw5Zj6
zn8jq|x->=cX?&c3?C$|9SNQq)>jBJL7_BQGtPH7I!XyxQ#e%G4Ie1jEXzeMXOC?(V
ze}!gIa)w`pto6JWoIN@<WX;PbG@L)RRTzFFp(H=T*_BrSNn>@|>32)>^)e|4Yel=>
zfNQzP7y55-U0hu0jzZM2y|KAb*5^H5z$bD8kA}tQFLm|ITTt_W3ybld!WkPIi>loe
zG&eWT0FAem;YMcw!KqW;*RcpdCR*!sSy-VM33HI6mATlnR^L@zb7U~X-tJj<X04^s
z=)Qp7Rjt_1Iki-?;h50X0oA~4$^4oM#^kV8Co5({)ECk~FhMfC_mf_H0s^GfNz!wS
zv+g?)$|0FKf3nWETCCBkxH{ddpm(+(Kf3fB*O6|&*kH{0Y~PP}o#(*V?3229opzUJ
zxj<(X=R^V3i`3vnqox}uqS1Kie6Fur!E{R1H4`*CNx^OCrRfnMK6p2M2JqW>zTBbd
zNop!TB5d!6wTp|(qMB#JfZpl~GaPp_vwAVzbK@qg9(3373+jfVNvm>*+-bt@54`RR
zas=^DCwuX+@Z)WU_Ak!iaFM3xL@4)pP-sof1`B)eV-PYjGM71LAf{RkLDmQ5_O-<a
z4J8W=6{<pwe}3|KMRT9|9F(^Isni%5I$y85O#ig!xk1%48u9Swet3km>Qqf#fQ&&n
zwc;?2I84*__;$u4yOs=ujz`Q=7%ws@y+;Vi@yF{ZoJ5Oa4c1_<+P$LcHz|iREK$*y
z_hm5&X;-^+{JvgXU;l982M-F{Hd5)jq-u{(Zznl~F792s*V)W_tUG=K+G3G>s;b$(
z*%P-X7Bl}ygfHgQ#QF+m%Viu5D|@xF)<b`!yUrG+m&Cci+z4q6B(Q!(7WU50QtE`n
zuV0J9H&MWlZl4A!`37h^hJp+#KN9E8+>a)VORn_gx+cclZ7LNlI_;2mrCq|EbZl?<
zT8PR^UM})@JIZr*N2eLS8w1ai9oZTkn*5i?8U*TgM~WuCv;;#unoZpyRB+u)jjhqv
zZKDVAwrs03S0QxXf}5lV?pv$$bTQr9@zSV=D$RXgtgKXn_aS+fJix$$hjs6;Vx^?+
zGTa)aH(Qawx=tk!x`3xjzIw1@U6N;nD7ww#ci*wj2X>Ix+~aW9`SrNc{Y7Q<&`n8n
z?(|;_lzl$2XK*B-AnbxNO%IT0<_qT@o?%H4S9_3=Apt2ztd0gRD+3EhyV0L^6MvEe
z5kq#UEc0!reRcP?Sl1lG+>xdQe@Pg<e11QrMfIDHvh~lP!){mh-O6_!n%rc&XdW03
zLrmi9lcpfxkbt}e4DqcUN}suHqHrh(?>If==1skKbtv#lV$V;DTe}KMD|DGBFkq~)
z$&z=UP*m3PHZ^KoQwEirvxz9^_g%Re!)%%FR|You#|mw<6))C5@QsZxxucv5m26C;
zDf~VLT6A_>q7Li#_TYkOYCg?Ix7@L%GzluBWiK)*NVx$uy2hu`Mu&#HB6A45Ewqj=
zugMqGha&R+*^rRtL2np|8P@bxDJk;pI+frK>x7-*{_`O(k4&0-kCafi^539FTHt%&
zKEA+hpriJnX{DLOh+&{DegS|!6{4d-v_X;Wo)b#RPoD;KDvT`=`WI@jkzip)B!C?^
z)Cs2jzd|Ezvgs@e47;XSAj-h_X@h}`|9|$t5yO2?!Txn1{<fW+IEWO9(>O+D7a<{v
zV?Agk7pT>;Aeb*%8)GmyAwNR<Qlve!ixU@y5~3I<1c02UfqzkW`deXE*i@I&+x*a7
zI^Z({mROiCB=LfLc8eFJ9RQLgFG#G~AXr?YMW*?0Q1p12R(F=lr@b^5RTyI*%K)Su
zy)+~eP^W|=?n{P=J|tVtW~@z<K|pthv6g_oVLy&!1U&LH!6EMh5_t#6dTgq%{~{Uw
z)4RZxN~MClS^%f7zt(4>Wc%nM!_>ulobHm%plHV#7_38PF-t(adiNp5ot%>L9#TCR
z`}kBvbNv0*GGYY%EIA3ay*qyJ5%xnM7|!JTeH*x)`KAPQhSot?RuH<C(s*5kn^b|t
z>7{06T`)H06cZDB^!RZ~M#cx&mcS9_2X##?HuLXpDGO0*G0v^5D{ppJ9+U(7mXxa>
zUqST{ehJqSWXeu9bE7cCpCKuMIzjywvl9rQzAsbzr!Fz3!a`D&tx*_YFIRF>`=r;U
zWeYjR<D9e7ltL*5g+v%rK4IbOc!pI-qU`_TLB}N+#s1Ib@K3939kkSHP8Y>>2tyiV
Ngps}(u1wb{@;?Ptb&UW3

diff --git a/docs/articles/vignette-05-tree-sequences.html b/docs/articles/vignette-05-tree-sequences.html
index 7879b8064..206139e1b 100644
--- a/docs/articles/vignette-05-tree-sequences.html
+++ b/docs/articles/vignette-05-tree-sequences.html
@@ -33,7 +33,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -162,12 +162,12 @@ <h2 id="setting-up-python-environment">Setting up Python environment<a class="an
 <code class="sourceCode R"><span><span class="fu"><a href="../reference/check_env.html">check_env</a></span><span class="op">(</span><span class="op">)</span></span></code></pre></div>
 <pre><code><span><span class="co">#&gt; Summary of the currently active Python environment:</span></span>
 <span><span class="co">#&gt; </span></span>
-<span><span class="co">#&gt; Python binary: /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.0_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python </span></span>
-<span><span class="co">#&gt; Python version: 3.12.1 | packaged by conda-forge | (main, Dec 23 2023, 08:01:35) [Clang 16.0.6 ] </span></span>
+<span><span class="co">#&gt; Python binary: /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.1_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python </span></span>
+<span><span class="co">#&gt; Python version: 3.12.2 | packaged by conda-forge | (main, Feb 16 2024, 20:54:21) [Clang 16.0.6 ] </span></span>
 <span><span class="co">#&gt; </span></span>
 <span><span class="co">#&gt; slendr requirements:</span></span>
 <span><span class="co">#&gt;  - tskit: version 0.5.6 ✓ </span></span>
-<span><span class="co">#&gt;  - msprime: version 1.3.0 ✓ </span></span>
+<span><span class="co">#&gt;  - msprime: version 1.3.1 ✓ </span></span>
 <span><span class="co">#&gt;  - pyslim: version 1.0.4 ✓ </span></span>
 <span><span class="co">#&gt;  - tspop: present ✓</span></span></code></pre>
 <p>Now we’re good to go and ready to simulate and analyse tree sequence
@@ -340,11 +340,11 @@ <h2 id="scheduling-of-sampling-events">Scheduling of sampling events<a class="an
 <pre><code><span><span class="co">#&gt; --------------------------------------------------</span></span>
 <span><span class="co">#&gt; msprime command:</span></span>
 <span><span class="co">#&gt; </span></span>
-<span><span class="co">#&gt; /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.0_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python /private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpQbon7e/filee17860409253_introgression/script.py --seed 314159 --model /private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpQbon7e/filee17860409253_introgression --output /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//RtmpQbon7e/filee1784cccb6b3.trees --sequence-length 100000000 --recombination-rate 1e-08 --sampling-schedule /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//RtmpQbon7e/filee17859cbb4d3 --verbose  </span></span>
+<span><span class="co">#&gt; /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.1_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python /private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/Rtmp1N5UNL/file10d9023ac675d_introgression/script.py --seed 314159 --model /private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/Rtmp1N5UNL/file10d9023ac675d_introgression --output /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmp1N5UNL/file10d9041197df2.trees --sequence-length 100000000 --recombination-rate 1e-08 --sampling-schedule /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmp1N5UNL/file10d906f4b6ddb --verbose  </span></span>
 <span><span class="co">#&gt; --------------------------------------------------</span></span>
 <span><span class="co">#&gt; </span></span>
 <span><span class="co">#&gt; Tree sequence was saved to:</span></span>
-<span><span class="co">#&gt;  /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//RtmpQbon7e/filee1784cccb6b3.trees </span></span>
+<span><span class="co">#&gt;  /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmp1N5UNL/file10d9041197df2.trees </span></span>
 <span><span class="co">#&gt; Loading the tree-sequence file...</span></span></code></pre>
 <div class="sourceCode" id="cb23"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="va">ts</span></span></code></pre></div>
@@ -431,7 +431,7 @@ <h2 id="loading-and-processing-tree-sequence-output-files">Loading and processin
 <span><span class="op">)</span></span>
 <span></span>
 <span><span class="va">output_file</span></span></code></pre></div>
-<pre><code><span><span class="co">#&gt; [1] "/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//RtmpQbon7e/filee17821d8acf8"</span></span></code></pre>
+<pre><code><span><span class="co">#&gt; [1] "/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmp1N5UNL/file10d90648929"</span></span></code></pre>
 <p>In case have the tree-sequence output saved in a custom location on
 disk, we can load the tree sequence using the <em>slendr</em> function
 <code><a href="../reference/ts_load.html">ts_load()</a></code>. If we’re dealing with a tree sequence produced
diff --git a/docs/articles/vignette-05-tree-sequences_files/figure-html/introgression_graph-1.png b/docs/articles/vignette-05-tree-sequences_files/figure-html/introgression_graph-1.png
index f91829f50dea45168910ade5de9b7fb1302712e1..fa23ab17b515f30f35f6404b75d054ccba659efd 100644
GIT binary patch
literal 27819
zcmbrm1z1(vx;H*4-3DMGN*Z*llt`$k2uR1G5s^;m1`8Af6hXQKq(KSkP(%czOQc1*
zW6|-w1NXW2-20sy|NqCc&$G`a)|zw7cf9Yf#(FL<D?xgY>L7+;q>{I9C}J2<C590U
z{<ROjV>&|o1pc+}p0vab_&0p5i<fx9zYbX6R=35lV|M6&1pS8eP8fCule}?F*(qwe
z&s8hfU}tw>^*d&BgdpTFpS?&<NSI#HJJoloB_`>rZ=b7~Q8{SXrgNK2t!iqDv0rN7
zFW|l4H~(@Zp4UP%voq)M&ooYxU+O28yy_l!g>SCbeEIzsOG0IXVD{MX1@h7Itf}(3
zrb14E*dp1JRXm3Y(3M#~&yAo({|CeJ9}=HJ{~n~g?X~wakN;2K;Qrj_IFQ+ysabDX
z5J?Lw<MDWm8ymd-3e$a3<wI`P6eDu+aMiRthRst8dlF!jzW)zzNBEr#E>!GrRr6S^
zXBK6wrw)C#dIIy;!0eVy5@M3HCHt<al^!dcyu7@B{~^q)*mrC$8{Jm^+mw{sRX&G1
zC;Cb{Sl^cWYUsod7#g;Zi}=x!aqf1B;ir}4CKBYtRF*~?%$7Z{?*^y5J$BZONcQck
zGv)J)(|2+zH0O`()~jq<*uvi+JdCAKc*?BP=L<S(ac_Ny<;Zs$PzZXnl_y=bpS+WD
zz=Pc{iGPkPj{lT*{KO@l*1QSs@#craL5b`?Z$!`pD~HXsR}v8Mx&JitefEq;fNqU5
zrXmJiDPQA|wehEtxk;bI`23A|PFtf*eixBp=I!|^^2r|`5)K^2QX~UKS6+6yZ+`UN
z=9G&kt&@_AI32`Lt8mUjYdyxY*x+?}-}|ISQj+SVz%eW5(LhoHOlvWKP9&}Va}~w0
zyZx(KI04_*EbWpzC1PF?AB>vl$|{z%`&Pb%*IM=6JrrC_#-@7v_2US!UC!L9hxEVJ
zHR%rUU~ykBN6U8Kq1(8hulCzcNXDPwDt>cosj1_g^apc$CoTQREa#22*9L|@1Y%#7
z3WSf{g{oT1*xGU<A*PI^{d$?iN78vcxq*zxD~t%%@4V~@iixMB)cN-}WurNA(#_iH
z`F1uMCpRPC(9LA6ZAJ>I-^Z=Q*06E#G;8nlbX!dgPH54(IlVYHd7+66^Gcg97~_WP
zs<iW2T(rYSIV#I5rVV0|Z^gy)9R>(8`y8(Z?sDmPmMrXZ>`_csVx%~nl$Jq*F|h6?
zF4vvv+Wd4palm0IBlhyGy7c>f84I>a&HGxXC6e^_i%dOwl%z_V<gqF_M4B#U8OO9f
zycBx28N=H8%G~k~N=fkzZJ62Hx#c)84J7oD_ok-|x<qW?nPx<)o3CM%%){)p6V_$?
z$~?3M8Ch>bOZ;=)v=z$pw#{%hZibqblxy&EbayeYRM`_JC#Ev0ax^$xHojUO%quRB
zw9u$1e{=YBT+<4DJ8MbbZx(_zUhxf=@Ds<lwwYw>)>=Ex)2Mz@{S=riiND_DUQ|)a
z`Lx}!jGNVlvEs5s;;_1+o!=xmeaeBIjV(rkeVFp0gfjyjB8m#y9x?Y%QFI!Nj7SE#
zXtlI&&vUkwIV;&)_2+$F;)-1G&M?X_ckUm;d7n*4k-w#5nIyxZhi80^t0$J<mZf<k
zOZ(cTpD}-WT*}LHOJlON!mTH5>!9EgE~JC0+ar8MP2svWQ}*~X{@LhZ?ZK-QKYVFI
zMD+uA+@z8Md|#`1y6*&GL|EF*+831fqtcuRwIy}Rm6g@>qL)H_97G};(l;+377vi3
zTcgOKOkXauqp(_k*UEJ5_%4(5jVHJVSjy(d4BPF}G@ldBMN`|}NfgYLN)j$RwEc6U
ztmUyrk)c6a7KfAB=iaL&?&#Y$x<3mXGcsLY)#sjlj8UvDjk;3kXjObJu5w;K$dct`
zUS1*UDG597ovG>ozIkW$A;;yzRr(d5$sO*ScFUD4-wrivXMFd(##y3fAbUPe7Z+*1
z6zE)6*N`3^M-Vm{yX&s#yb{YOsY$lQA|A4O^|b6nPe@kgGb(}5`>%fLGH)+OryCXO
z-C1Axs$N5(-Im~g_Nj#Sd=0^!q3kAI*d1P}R6gD*Z3Znr7nc-dwp|kFcC{(HIZo1k
zZ)n@5-<2+^u6D_{%C;F(Z=zzeU%B8V-Ssh}x~_JBQ&PzKdQA7YI+;Q7=){}b8=m6Y
z1_pjA7Mpj9O$v=%vqv%N)=#vpn#YqyE+sMdmwAyBg?-VDtzf8x3g)C)cD0Y#%-We7
z&^7xh5J{mVc1JAw+bf3cGAr)bAGjk|`|{Rb>kh=0Ni7wyCsF%2lR2-LlD)ra+W86p
zv9Q_R-Hbxwb--l#Ez;`|xXcUIWcW+{uY8xd$F)4&_8#$_)yQvu&Coe!zcXKxotyjC
z9`ic6ec0D<!)$fRUg^-zRbp#BDe1(_DF#0kDfKhmgr1p^OOzAe=vZ9~PPC1Fo_i1<
zt?U2!CmUM^{^`J}wVNLp=d~W6X{@E+y9F<y_y@+%-r>s?+FS;l8XbJ-Q=N{6N7t^l
zUscB3y8A7%){DA&e}!*JDTWS6vl|t;JhP&9e6afb8I-Vn1FNN*$>`z2C+j^+p)iZ&
zhoijX{e1$%X!l`aZ7UiRspM=~Q@u5Kwo9V;tuASgq+g3Z1M{N%as2aJoJ}q}Oc<sW
z^(aU&S6u!>TRg@dD1v$I4z6qJ^gSoWtZPkscoaFQo=iFERl5>&@rt*0sr%U5+nraM
zC660*E0Ef7i|<yz?2k7MEdJ^D=O$iF39&olH7G)a8hgrxP6x$oE^2OnR1nYiHI`3b
z-6y|POd6CmY2t8V?2YU7gCkw!cPgR3j>2X#t$w@^ZC4s6zp&J=606GD9_L-4Di|D@
z9TM)k9QkeEhRppZhP&lW>vDvHwnBCxofXU@vB96uH3Ym9u6kLn$!cHLH+_3B+iAz;
z9JGq|gMNv-GwW9vb*Ym{)qPads+NKxe0<Wcm7Vb~;g_AEzSS{ahJ`UFmfJ1;H4(iV
zvhgA);$w)p{UCR5>%EEP2NMH+NzdvNKSnnM{I%mdmYpm#KSh8I%0)bRYAV0!_rvZu
z)@Wld$C~zhxxuIPrz1mh+w`+U{r!p1Y-I~>(^>_yA0oK(WvT1w9!e0+R(Cb|ru#jB
z@TzvrIW?Fk7*=r=hA({0A41g(U;oWF%pVgg=l1q$acbn<*8loKA=xaebI!7`vE7Ly
zNF(1oS?_>jx<UP$cV7wNb;6z#pGvva^U<Qsf@m{kgwJ`hS;aSC`&vzv2M-^UxZ<cM
z(Xvz%<??P!be3z-EFT{o{81$bL*-RIRxYm2xU&^qoU6MT?DrZ1M$MCaOGbiYBDPYp
z{3>*C+V&yc1PEN^2V90<S1)nMe4F(x`*Cfyv@y}d&xgtQuhX~&G7|U=9+Cq*@-hx{
z{h7}Uoeg~+$hP24ZJ2Zg!c;}q_z13%$y@K+y$j#ThNC!CYbWeb;bPd`gM0H9^U~#U
zKgv<XlV{l{gl@a2sHk}4ji%jP*=n6(u1b%Hl(clnn>X+H%sN>vUQ|4P>xFzEoro#l
zzPLL(k`o~pvQ{56->#S$RTr~;5$u@r%eJP;zF{PKJH>7~&PZGSeS2C$JM`O9J$wx*
zyzQD$o2!@D_Oeo@PMJnqvRs#@RasJkM4)MH5Mzf+VE84Yx(PcrPEMJ&cQS23jM8}z
zhQDg~6R^Bq2>ESSs6+iGB%3=ueSE}nm_rvA%x$szGlb?+-D!AbmZo;PyFOL%0{u?n
zp1Yd^6jW4fl2TGqt@?o?@18t)Vp>!)(Uy$wHi|WY_sZ!uKS)a2A+p+=P%z-3sa<TB
z8gdE~_|J3mq<Z<pu!&+>v#stdx0Uo+y!M>5uU4&9<IxGHHz)euZdC1uw{{CpJuu&r
zATeSm)R8E8+)+K(up#g6C&JYS4<3AIZEekSn9=A+$H@`xCu^T67~o3p$X#-&382*|
zcPol<`gI_1(S37;m5J%rwQJXAb4D8?*)LsEBJ)X*x)6Nfu(-FL6Ae+n-~%-w`#Y1)
zq3t;d>;yXtW~tJ<ix!-3mUQR7?^3H-1kHr*erK0TslPwsmP{#TIf^Fg3$BgFlk8un
zdEg@13|t6%KLP{Lt<W(sF{x{4P>B_Fw_ELVWET`v^VVZwWV~7Cf^XegUmCF%<TYxf
zX^0WgtgWr3>$-SYoJ?Z<L99doje_6F7;QcH3Subf^XNa-nCG-`PbONZBOybl%%x*N
zj(|Qy(@ss+=s82`d%>~NnSy^Yf|)*T&%XY(>HN`|wYKA{wUqae%YbKwS&X%TRCIM6
zr;uI08iCd6efo0**wsEmQ$q8Co_zOB2m6_yD%oW#5dh~)=E~E9gHOoBifX&A&UP#e
z*YMq~#MC)XcANA2(_B`3{`@)B6$kSooxwU{VkeL9ms~BIC?Ai!IGm`EaLd=v&z637
zwPH7k@<&s{`dc5HMMvo%{+)5D>+Jf1Nme+iPkM^Oqd^>sn^TFg<+k|40xwtx`UR&7
z-<rAPT}YfyIke4-VOFQT<3reSlS%5i%j-4t=}R&Ktn)7vj0{uhsop&-SZ>|c|H{Jf
z<ke|R{%bq2AOph<)pQkc@52w6BMuxmurc_UJUBEod3xG><o9pmZ!gcz+SVPDE$Pd8
zyq~<-NBBy#EWG}w$NQy&Tvn!24;*J=5fajXC$Laxb?w?ij+1`ZuU~JZ76+cOSt{zj
zks=o>Iy(%{AhoctaNmxBHLFnYL`1UYxrtp;`;ZV$)%=-D6Yqv%X38Rn%x~{^ylo{D
z{H7^5#;IWPfxWR!Bq7n~G!H`g=U$v!uSj?Ao${Wg6VXV&c+t+KnO4z<9lzZ{o6hf(
z^loHqYKps}`FpqR@{G|`d*K89-1CH9LZ8}+3r*VI1!<8VsaGfWtE{|w^XAP|_>Rjq
zzN>F!KN(H)m*>YRjvagJF!R%7qBSWf@x_b7oVw*&a-!>rHTD`G%~Q2Y9BPNG3k&#u
zR(MwY`t|Gn2en3OB6)3o&&8?DT%5^{Lz3MvPg6yBDcr)V2z%=?PA)s`0gY<yQp+h`
zB?Z0A^v+X)KELFc>sd1HlXDLq%d;Ljm-#b1wN2@r)X6r1PL||2{V1{wwRbJ#VsA$d
zwKkCL$0muDJT)cYVx>d5G@uR8komwrC%JVp`paC=)|mO3|GLkz>#ILCaJ+*D5AvFI
za`@9;`PiG!MomV0`gCez^cAy(U)8Wnl4YY?r@M1Q!ot*&PI*nrYn-#t9^2R)uf5r-
zM8CCRx=3BCbH-q<ywJ7%DZKSOv!|zPxhy%g0`u#Ei%CS5UBykEwSRbV`+;}6Uu7jF
z74-Dtj~zeGXWGsJwJoyTEFQ!V+3*e;<s@u^w&r+oJXBJe)j;`VcWyFnHB9iqTiIx#
zq$f{^lM)lh%Qt70e8_0iP7uk{u2|YOG|jY2=S1_pkE@^47HgV5TP}aJB`cPhOm{0E
z!^Rv~*rhiN7}TQ;^4Z=ioR1cK@Ok<+7AMPN|5x5`jX1Uo7h3wxyMCOXxAvzNkRAT&
ze`U=6p1yuPsVCR*6DJaX|Nd>mE`j6QH+V0<th1v>^Hrm}R`$hVr6$j^-#9id&o$CH
zeIiWjl=!0mkibcHKL^4}?mKc`52dS96BIw0<R(C$!sJgMyBillQx_aS&wIt!);9R|
zM>ehi60Gnx(Wdmxo2_3eq{>`&HFkcTn20q?tfs%MK$`S>Gf_L7LU(xD)>ihmxMw%*
zaO!S|<)n)30#PO7x8=)~l}W$e+(=!ZFqC*;xw*s6XJX)R0vlwOlGe@+)p^;d^`ve=
zx@hyqf>}?lZey-Dm*>aXuHvVMOvh}oO2#L2a6!CgV-W{eW+Yni1(Ip{ZJ%>If_9ct
z?V6ShB?vK(cX3x;jLPkF*R?r>IW$>s-cZ)DI@S8wF7?_IJB!lgZVg<ZwQ`lMPUlPI
zaqG0op=Hq(A>Y#t7}p6H-`X6lw1JZKI<G#ubMcw!;>(=OS_|kUEwbzz6v~GWuuju>
zY7$_kSAVgvT9oq|zi$^TUUcYOa<h%Fm)}@_7Hk)2*N~q@yyn-GBt|AWW^y?x$#ipu
zDYZp(PCq+=j)OxnxceG5FNvVX!7OglNCw)bpa9z7DEq;Bhh*{M;$-%-*;Se<gkExM
z8+KWXv}BTs3M#J35&=P?PI4|2Hz(FpLkg&Quja(y{ao-B9@=P}XT=sQzq;qq?P2n%
zWK0H#Q3XIk)2j)FSvc+a9!o<B3=0iV=y9A(essO-nxtfh?7@RUbe@VX8`+9qJ=v|s
z>{NFeQrlR|D%8?+bxm7Z`Nl@Q=~{79+9FjA5w5G(uu8dCfLXEjyhzJ88U?Rn+<%5t
zs{Yop^>Q?K%w*w|boW$WFET%yDW@vE+K(HN|H^0f)|=p=iY1V#Q)jxH1`cD5w-EZ1
zZ{s*=(t6)ynPp-=-nS>pjBMW5prTyNmc&F-V=>0ADK<^GX{2Sep`D8%QOaT}E!1%z
zix2UsN|G~xa<LTB%%0n2o40p2@_udjj$*5moFXe?tq~nr#T~mrgiYgRw6@v7MM@hJ
z%&Fq$4b@~==wENX0RBqh^o)^rmdk8Y_n^P)BfjDB`o3V2N5$$L4@Q&bBVA2_s_o{z
zAg;yj2T3iC$xZ8;oRt%i$qZ$Tf$eDYuEta+PRpa`YS$r5o^cs1h1RoZvF&0W3apa(
z<=%g@l#pk9QC1uO>3aa%1G?7NW~rDg`tJkycVMZ<Un+fA^Z>g`dkrmv_1R<Il5?Cv
z{1&`df_!PW)m>b2cwI8OUtw3HKkdO~%C`#oCUt(T?Uae?*-|Z`MJt<pJzcs*9$$Gw
zzV>qCOIv!KusE`_XjKa3IW;9ccYW-(!h&DxJMeNHs7iXzlPc4*xOtGsEB5@}@2L4x
zmFQqzLjKt#|3AdS|Mx%HQvd;WCk<CVB+@8z$uG8>xaYtUl=%KVPrgapE&j-cX?gkm
zXfX$Cpt;-aZm&*7JNBopn1#N1gMg@6XNJQ4`-!(+oQkItaVeY<+ugEl2)j@<95qzs
zV>aF#-$8fCE26-DO0_T2DVB`xdu!8WpZ4uZkAv??B52<noKD8u&R&y=-gz`zJS7YI
zSS_M)^8o8u1O#xYsi`)CG<-G)`LY4N24t4gGZO(#{L$mbeP!2OSEgHRn?#b?vrBVs
z0V7<Q9;nbI+9Yp%T~MB6w&Rr0Jrr#FyWK-%eeDcMS9<cp0&?CfG-1y!R|3|KUtYFj
zWMVRf3C3r$4QP8Rw@Fm#-o3Y=OkNikw{H?zWf2kS>oe@i()NA+JoVKkQHr+G&dQ@(
zU&K<)G}@tQ2Swb-Yu??~dl=)k#hNSrkBbKy(CkaW&TH{A@8Fc5pI=BsMB0}xH~axJ
z)W6|;-8R9|WeEV+Wucm~K4h+^K<>hY3$VI2p?p^TTH)d0vvgyz9vijM&juU^+%<9y
z{SQ~MIA$JamNnj7nVAi(8hY7L7t3>1d$A)_cgmZ;AROeteB<_O&P613_4SvooCte{
z)U~w_zQ-i1c!Gn1-U9M5+1Xl$#f-RQ6kymClZmF01v}Ms_ut=+<{P(sPT=6+xD#~0
zpoikfk+{#Fuj9vC#L0!{y<}rWGsj|FWxKn(srXETYn*qt#>DvSCzS(pDz7e$Q4p?$
z^jYGp<-@x5qHxA#Jl!)69of1A)ADzZm5Txeqg$EnOCPQYG^_L?K(ylE!PalD&gYwU
zq#bm9$bS6D;JV0Wcv%TxdyNnGo(kDdGWb=$efI1xk*)bE7bv%S#&(OIj~!5S6CNNR
zcm60ffO_=@xepsWd*zDe<;$0;F5Z7ulELAe-!miRG#`gJgJe+KM6li|Dq*Mh45>y<
zF%dj^kKQIE*u%{RPMth);@iCsEDQ{ZpEvnU+rPO7HaE*VI=i`*{u&<ct3f3Vg~sQm
zh~<yS*=Dyjp4Vi{*c`Yvm(mn+U}N|@yWP3DXQDKow_nBk_u&i5Pb9tl<>tN>nvkZ;
zef_+VV{5;@n2XkVL6$q(8hvjomR(i+t0FtcjUHO<?RvVmZ{N1(nJAU*Y)sAO0P8XS
z_2v1&0|yjJ9A@(DCKM$VZ{6~K!>yae9P8FPOvS1Bu^q(a-`0Y?u2!>9-$0vk)Hr??
zK1k^I+;x`;zgX=q>)`P5+_`gXGA{|%vJ{O{rJ@-rI!}9>-G81C#7`#iT*CS0%%ZRC
zWbO?8_@;a8RDuLKdu1K(0p%A+7;rM~%8H(LDB-v4O<7xWqBwcdv@=6}k0E>v;k-rh
z?;wtf-#iQzM0c*yREq?C2!pmezQbi_eGDJFyCo4VWUuJqQLf{;Z4djZ?%s#?@iC&i
zF@el9`0>87Q)W9hTBO&hq=F7gFY?(qMCI5xnC&jmn$o)4=l2qY(jmq@_{lKWWb}JD
z(CW5B+@*0F_$hW)R!I@pmE}nx*Z|6=rfJ8H9cxcfI1fgN`($(cV+K}0p@{hOyG(J4
zEPcP^8xYW`?bw$D0JZCj1mR><)l^ccv@su@PnqdV#+%8_X4UkDZP7}qxHFni9eQf7
zO-uH3i6H%B?E}BXPp9byO(tPsEdX$Bg%5@sojuIW&5MSb3?4mt6dV<$WjQi_Q6_@7
zWp}G-*NL9j`oOmxRUvJ2hZ7w&>eWq)o1xt7p~RkV6wgF<Z002nNc)4rKf+66LnW)u
zsVdUB%XicJma9AG8acXQui{E;DTzQy1%+ch!b8VKq*Ti^tNY~Fw?DBLziG2(+Yml<
z_wsaWfMe1(U59I?0?G&POE6SiNAv0~Qx&WGd^xY<M`bk@^-RS%(bn_F8CKY;8AkyG
zOpz3xtjrBPSzLekG#eLZ5JNWyf8dINtrzL(jiJGP{_g1|F*-4co+1y5TQ{j?Ssg3x
zZyRKaFdwShq(v9z_M~d4Ie_-cGJ3Mmlc2FEZExW1M?!Fub?n*7P^>3zH^t96TYcai
zvrjcQV@_7uFpmy)zu_`6cg*jRc94+BxtmX8Dzvpfj!@Kk)Rova((GvcVo`ly^6K*>
z-%EyS?G|kx+1VCmu|%`qq*)V3d&t}vlZTXz<qNMr60SJH770A&CXR12njtk#5JzS-
zK|-49?KxNh_dZ#ulq_qqu{;qWWS<5-rFZ1hr%!BbY&P`!_w6&?T3bLf;@tW34S!Me
z+X!ljR2a>fX(^sF*gBKBt?WC+Ktf!o^D!ruA18OA_Cr(7zT2V|N=+nNV~(RVcaw*L
z+pJUCC`6V66?BcAx(;ZRamwU&zBB$Lz|EeaLqse5id9l-7T<e0+tu)%cK1t@t4QsO
zYqSxHj=18OMdvnu4L{qbX;rcPQOI#NBP{GxV42N~LkX<ryWO4D3iL}<0+t!9O7D!}
zQkVbgviHeJ{3WGIZX=jz6f3&AZp2hgz>m8(t#Hy>HC$$u*EHps3;qN&W?+tixSIwc
zAze2#BOAW*tA!Q<;<CE@UWYsx->JK?t3B6WA!A$Nq%knPbHn*2v%QR83{}g_1H7@o
zd}&v?@Y|#uxtX?4$HPrK)(ZIWM+b+|yiG}f&eu`mXf*@<B~;hLIca9HBR$!$AtDvj
zn#rDm>;QVPESwIsk)Qc(u(8az?>c4cRn<*psIsm|r9RNdk$*^N$LHp?htPxy3D?)v
z(Iy4;&8CShmFl{w9HO7Pr_A_JRNYSA`)1BhEpE+9?l~f@U2}t)*;xMC#d<lj%51SF
z^C_-^5V0;Q>s!twU4l!6y?hzPaWSpZ4L$nh+3(#37G@$oZs@rej4Am?^+{G0i>4*-
zYh)MbUB5F?yPNsmZKBFG{k*c)uRnJLCaU8#@o;KB(?p{f=UahdJ5HoGo|;mq>K2!$
z=Q%5_>ouQ^Z(ZV+8X`>_(~(V9)<~PnE|TITBvjVWNDB%Ivf;L>GfT6$IRadLHs_q{
za~iT*+^u2_i;?5w?eAyPC{Nh5h)fvbf5!GH%~`t?<k*J4@xO<!dnT7RTf-Q!P@CZ5
zSf%W~C|<ndHdS9dQde?+sE0`T653spDRn^uFl9%^#?l`Q`{EQikJHiV9M_947P+9{
zvZXLlNr01!ul_}qv$dSt5xOOnmGv#|%+sen=|iW*$K6}%WAhYbcPvlK;*ty6hBO)q
zrppqRLvwNiXp^$$2%?gVLU7ikW9s%sI#t#&_A(NTW7{?rxjn^oq$49EA<|o<RRTFK
zkqu0djPeAXSGzL^>Nx9K-g|ywV>zQn(5YKbWS+HZAZF2tbu}HV^X&Rq;UTnD<{`95
zFM_}J=x2tfY<6$iE&oZJDB)L4*(yoJnXx_CcJ%b=(@0c&Clkp~yX|9RZl2NF)@DDG
z-xYD1j!QY=R%?`VD^6Fc<Wf?7#O=OfX^#62Iw#Xl+zt8tSgDVBd0Pr(iuw>xca&Uz
zmYKM2SU;EYmHy2mWx&lDTs0xD_n;g`n~k8f$|TFHD)n>8L#1(^3EE;-9mO+qQ{Tz2
z>tfiSEI=a{xz4m$0YhrL`rEdr%)RNdRbq%P4HD(#<eZ?SOfH!%mX;I22hSD`lR5Sc
zetJwnPoE7MORD|UqNjY-{vYOwP1M@lYkcV)X)gH5R6tOMJ~lc7a<8YEwdch8H$IB?
zd{9=JbsNyXX;|AaHEI23E9gjQ+Hxo?9PO=CDXE!_)kgCC-z61OH+GJa?DsaE&#&Fk
zAG^?Ta?W+2ENvivqUofhN5C!~@Ge1<=w^nZl2S-;Fhech*8A_ltV-=4^d(@<NFF-L
z!NY!|Vm`T^yRC_l5mao2w}fj;<Q>_by|*oX7RVkyxpU1?d-eU>DS^yI-+nQ=4cSBG
z;aO!h9>EQ!ZJ9XIwm7pnoRf9p3SE6+N}((T4emo#R`x~QWQF>G@#r&8op$`-<{If=
zm>0QP&i&_xjZx|*+n}BBS@srcy!a-zE$26C=!3OOo>b>GxtmtryZ)+kyYL3{Qeez#
z)T3X*Ggo#JWPk5+NSZ3|+T@#`NGEj7Y85C|=6v8DUZi;RtUvF{XvIWRL3zei&AvHx
zi3Zd!EB6x<6Gz!Tpn(N$&BYzv=cRG~)lHF~q#3O-WwKSN@wq=?;hKtzyXTgzV?$3k
z-dxpqc5EjAAr^Wb^B+77;g=-HMb#GW@xTf-dl3+sUS`7zc~t#W0#~BXx|+nhY_G_w
z8s5r8e5`W+ofu(tgrxvd@VP7-U#20m=r2|8DX_x1xD=~qX{7?l2%}}fl+{-q+)R9a
zO2CZuB7f%V+X@O65}5j2rZR&P0U${#f~}Dclm<|2-Iu?heQy4#euN63Z`x0`v(gG$
zU#+gL2Hv1>=gy1nc(84b@BPw+qeqXTkir$1Vs2z$xMTM10Eo+Uy2IU3C#Ko^ma?oY
znI<&>Chz~JnTc+UpU)Ksd=zw#?+ON#T|~!qh7WxwPBsD2WBLzzRr^q?;o{*0tv-<Q
zLjd*fr*~_9@(s)#(minFwc770C+<TQ(qQ<yYKR4AmD|0%SkE1K1Pp@m$9s&ytG!(r
ztt%rVV_jWct3jwXQxy-y6Y|ZvxbiLgG#q;$_;l+LV)Ds<HdFcfC^b@)Px{@IiR2Ff
z(9HlcGx|+>xA})L?knoA6Ns@mO9Xr?_Xh)K<l$*{cLNZS_WphIaF|L00VXK0hc_xS
z3;NSYsrd^7vrA@$;MsJ$gR+fZY7&b)N`+yfkM|z@{I~OgDH4Hnv(DvkCmnIu9)pT`
z-5nrk@%Bnj$jg^^ZGu4)vB)4Kg{N0I@h%csH%af_N!j2ravDK-NYR8cY>x_5D8OS6
zKMS`q)&ltnufW3f(MUYRU8HR|JrH5juL&?2D0lyvZx#y7K_=h`U^V}}Yfw;DPE<)(
zbq==p@%|c!{eZ@qZjxYLb4U)U+)pFoq6~}CVNuXGsvvs*+Y#7CYR8%Gywp3zbaOxE
zW%cLRgm<@=q;B7q2gEGl44QTM+OI?IYePP>LoM)B?{PeWSKQl<7cX8sK}9vGKnL?b
zM#Ob;V0Wj$Wpy?awEFN9XQ6)RMJfThdc9_rV-plCY~85muF$&TFr8BU3?BU-T<f*7
zy=gX57ZTEG)mN<2rR(X=d5ajXSnn^8>P7F?1+$vAz4NDY{(Y2`UKAO8kokF;lbegL
zWsWs&O_Ur_eFmDwUZiH8AXBrT6Kohh<K{CBQJ32ixB!3yQ-{hAm281Rr~b1WqOUl1
z0#pbI4@cT{>ihTd4Uqyk5P?seIC0;>37(SkURfRi!UwW9`XvyFkzoO5QAdXQrCf0$
zKvnujIbf%JiL@$D1tcXIq<0G85sd=NOt^-0-3kv-1ehfL0xtxKS+Hau&BGNH6?sis
z86;a*2RyUo#I}`ljha*q4c}4m8pna^9!|pu%SJ68x7T^ac1t_6b+do`P?KzRohwT|
z$*E!dJ?uhIVv+4wn(bI4f|B8THNRIV<Uc-IME)$WniJK@Gk$d|fM!C}vN`VI%1n>Z
zEn|2Nhxe-aYPxEsvb_8m$-2hIPIzkV6`r1s?v|EWt{$-EQ^L9q%)=e2b6Q~7)ec5x
zX7;mZuhoZg<0?G1VR;gP@JR*P*x2xTY&rEgk3};{fPmGfnne5qEC5}woj`lRvs_sU
zdHs4gCXg7z(rSpvh2!eEO71+~N7_m*wiTBo9SWOGKByf$H+W*QP#*l3r2u+-WaE5<
ztcIYVW2+Cki3%=)y;fG30$gN2NaxN=qZ9PIYd#sO>A-h-?WqMU+dFW&oarIqG8h&m
z9dc|Ceg%wDDS${p%bl85DjP2?S`(jJ7EipJ^g9B|7s>wpq3j@gTsnIUo}3)~#nJP(
z3J18sKTWvhpZ38E8W^m>0iD2*jq!v)({m6*gD+F%5Erkem1;`e@$J<C96T->GKr@a
zs@Xd33J>?RF8SzESuaX1trXCf((R`@ef|CMYEMA;Lp-W(ig;}zIkq!`&rEFsS)SOS
zQfUlPspqZ-9{M5Ey}g%+`|8w3)SDztIWjuRcJAC;reE-FircUcO-v>(_Yjkiu=4Y(
zf=ikJ_rtV8vJV^N&=EV$EEACkwSj|{&Ucg?F8=CkCYj|=`{|9la}0zY{CdJwv1Oy}
z&~5ND*C?<=E2uL|JDq@NAL>-X>Coc*x<o3e`OT3kbb^8+2{}@FRo)aHRl<(5y`~3X
z)ou&)T0eG}?M-Ere#64S@jYR7yd_~{n0`0q%3P@u>bu~(d<~?}M&8DU8oEoY_uyI5
zWgi~|i|EF}!a`6Ym}Le=MmSe_Q1sgJOxl)V^9D<MFtUO*hH7iBzs$ad*6|xqY}oLV
ztx1Q_tTCDwBG`d3V1^!pHJKRbE|<Foo&^M`f()@)=&`*~_x*c@e>3!w(-{E-u*1eX
zGb3Tg=$dT9)=DI!6)Fr|1(65FWY^vfc;P4Ib?DF``{^GaCVfEyQoi@~#c9+XbX+DF
zXMTJTXO@dmhOGcwCK*UP4mN(r?)Jd0WDxA=WRMZZ<9+BZ3czNX3_B#{b?M^8c6eY<
zy}j9lgnnw%QB$X*MxtpAx5>!GC5tE(9&Ugg9-M-Lf+=i9jeOH2B)HLw={WW~Co@dQ
zd2YPkM^2aV>byh*4IgZNzb;#^-2V{?c6Ly)VA#&AH??Imn2l%(&UEKWIFx9kQ6GZS
zNd+`f(nCQ(f$$;It%I;hf7=soP3JbT@$sohN=hO-YBVeIqJggvbY#ibOeUWmk#tZ(
z|9Q!vo~f$su|5LyRq^iKI2bY%r%sKxc>kuvynLNSh+V*lrxJB@O2VT_m!^`=#ReB6
z-~W(ke|wp$<7)F5@H18x7PP7E$-u))tDK@6^zy<^dn=ffb(pF^Mc*WW^%tfU^U#{<
zDZs%v%ZF(Ra^H2h1<jA<3LvMTwSo?i3gg7k2W}vG#GZWqy<Qb~+)+|flpJbtP{>-!
zdqs7rNjM=@DcN*=aRiO70?R(Tr3QhiCeNMRkLKNPE4F8>Tt>r<zLOmXQM&=IruH>6
ziiBuXc$D1=q~nAD0z5i1P%)+C<htrY*z=J#_lyE!C;8x%&-9fPv~Kv&d#J$#Y6ekP
z(e)b46MfkA32)!#*^bHjQ}Ia^^t+@pu!@T61hs=<ZvpcxQzJjE^9(d|N4PBlf)KHf
z7W|RGSatl4!k@dspEIVy(j$QznSUU;r4$ybf#`~4-Ptwh9*MB1rua#q5g&WlGImEJ
zt?Y*V1v+h%p`pKl8mzx@;|9yQbJxA~jA0i6G*W~R1J|0%fX8O$NoYvGnb+@qW=qb?
zQwHe}Nwy$9@ZJ0J7s#kp+E%B$fsH2L^gGE4qIN;Ajo)n9s!kwh3<Ol#(3-+erxt4H
zy8v@{ypy2@E0b2p?l!&W)(1Mztp^C*82Z2fh#tkCQwv&#t1EVOC9RT&=h;zgZ-!K3
zkY>P!UEPQvAt6DUNR;f@(<gSamy8;N+a`d}MgS#y2d%{zRIS+!G!oNs*`3Q1ZFuP5
z6VPfja_&p!*AQTpPffdXlFD~?TsgH0<)N?mzIdSwHoB1Ac--^j%xoee+I_|Lm#*A_
z-N_5x2c9k~WFw#*1~IsUUc0f}CdcQxVs2n)n0D>)z6gGch^~jF8<Xji?WwQ3#-f}4
zj+wAtxS+rvDH9(bUkvhca7+xBB&gFau<E>epNY|ChsuTx1lG5Dehnd3S^dn<uMH+?
zo_-A(Ja8#6q&vWMh6$Je9kOkD9=h^UDIMW9glNrRIbb-1Fq8r0HEntO6dKTZNu;|V
zK?ZI^bv**g|5f$97DRKvvJNc%_3K_zQj!S>9BH`hn0;h4DLqyd+N$YBv*qxPdtSWo
z!V`)0a`(*>$B!q(#hr0fhj=r<8&)W52<bpHCk?7>b_2F0+T|m^zSq~c!R_|dpoq~2
z<#k&^%=!al697GAL-X4Uc#GCd5jz2&pd;^t`Y`$Z?bYn?K2j=Uc(y9#ItmJ}beu;h
z%2s}wA!%D8TJMw;xZc@1WxSp{Zopv>SV!OnutL#C2*vc`18FY51qkC5{PgKlNt?_R
z=jCyQe2X5{sjh4s)PaqyZAY>keV_Yk-$6366+tUNg{WX8y2GDCW9&dzfWXCMly7+@
zG!HHv=1N#@ML?*NUD??%V^9a&-duI;8XFyT$){STwknPP{rmpYCr^UYdrBPlq#|e;
zaK~uUz{^S9xzqmRgMK}p5H~t)asF_XbwxqX%+GT7l46IMv@RW28E?IDkaJa}q<r7W
z#m=@=xmHu)QoyvI9fp?M`7O!-gPbulGb|NQ9p%O;V!_Lo)gdMV8g&MfTucT$3>1F}
zn!2|ALqAZ(nXVO<7tt_n&uu(Kr;(8nJ2SHcN|%I&FlIxO*iTMpNABz(6b*<I5+!dK
z@YCH&fYIgc?D7I=F8_Ma_vjHJFd4Wh@Cp7cno{yl(G;-xSDn$&wk#alxB37ci88>>
zq^5(ISHZv2wO=p|Dv8LAp;7>Rz64Yyx-~BjMuwEl40&&<vn_Z;sZw1V&_WY{G!(t=
zy!849a=X~rq!5Xi%C49d9Rqg`p-{-QfE93sQ@dC>DEQT@_>ub1DVXugg%V&Zq2UAt
zvHJ1W=IR_yODhTr4R#+}>aX=ugV;kDcIgLLGwHawM@`)D>H^V_&GH<@vo*smgsW46
zB*no)hwwgPTc5NG9t_`t4Gh1QCe?*Re`H4hpfrVON`$C;F`60hpOe1qS@IwRgKLxz
ze*uXcSha=-K55`ioV(Z9PXCty5Ee;aUth&gE^TCZz*WOPoI0gyWv;8Ef%_oJi@FV3
z1#a%9^d%@+G~cJv(zDXyzX1-Bq0A$H&j2v9e#m+pDr+_eQax-<^_e`U(mRIT{+j?u
z7lbzcMF6B1&&tZm#=)Ts?%yqcDn@t(A3B%uM`Ex>9YZzrp317KZ@^$#hA_z_Jd#eB
zYxS98Lv}*;u)EyB$1C35*+fQ3R20qj+HfG#O)>-p@JEbm`FaIeS!+ewi_CQ_|6LE$
zk>`P3)%xP}CE>cyRaHO{1tByt8#VUz>(@8bJ6cT8{Q%~Vs0MHy2i^x@#xa`*aLcvZ
zx1Z@a4Ur&dBM1Gc*nVm>3mmI(a4bV0Fa`9sq#n|VIxz3$VAcwpzyEim&dA^Z^g5AM
zGTW*oCz5n!wy;WDz-j(&bxn;KIGm=?)${j3@m9c4Y=MrN;JGzS|FhUW{i8+CI}lFB
z+f$XGKu}%=!4KG7sh_oBa(np!mjUNA9i#)04J3nH^;jh%{vGgqAP0`xkpVbz#DM`&
zxyl&+NgtX-Yx6@Po#eDa%0QKhQ{oU}$5Jez-4GBE%wDdb&}sg62m>t1eo2Uc4oVT+
zAtjUTM<j=M?^b&G9Ym*xFt5o`Jk9?eY``BJ8nq{+0VED2)fC*MVi<#rl6xo`Zr60G
zGZPF-mHd4N(5qKgFS`7d$G~UOdIJqjc=pgw!6?d;gtb8GP{#-Tn#q8pfMQl*Hmgcg
zW7t+WtP1Eled$*vZrm83OwaDj@6t&R3kx$P2dL&b3U*#LbWI44x&%kB*VWdxgk8}8
zPIh_^I1vkqiRm7q<N}mzDf0ygj>l0TYyj;@h>5=qS+nekC@bkeArKetK1&3GjI+po
zr~XoB0PDc`Ui0dIJ*=Kd$?FT!NmNa#^Ad<Sj4)0mLBuLxbD!mhn-S=<Q?l>|MWSgz
zh`?jEXso@+)~F0db%dB_Ie05D<Kk>)PI9Qpa}p3@gV|v64|!A=x10kMR`&(q`Mf(k
zEs%U+UZ?<ZWhnB#F*Y`abG-}P3rFw1{Hhxs{ooSdFK3Wm?*?#}uU((3*e$<!=@Nt}
z+01i6Sk(xCCpa)<Pw4gzT;i#<h2evT54VHx^zQxp_mjTRx1#ihK0gH@T!=XTj~_n{
zCes58u`P)sb^<ViegLm0Qdi<Qmjxx54`V~)5iA)OfSf$=F{`ASAQ?iO9ctYV#2w0C
z2+Toy=wKm#7q&4pHK>7CsONy~p2~EO1gqS0Lfki&FYM`b5kTM2bb$F;Qx_bB^kcv_
zCXG>-H>PxV(_nC`XubrfSi+@UjB*R0@frFM!on^V?_W#LE?0!korDLB(3xXUzX#(7
zO+ZyC26UpLrDg5_xPFZz+ttf+vu7X;peX?XMeL-tAm9w!7NEXfx=3kaSIv-DRt`oF
z#+#J7&XjXkWTm&rR^}Ph)4(9i8N{WL^~GnVTEy^!=P;mg&o-dax)$*Gu!+}lq6*+~
zGq8ed4sJe^1QLMY$%_Jg1?d@B!8t|-hH>b!ipIt%$hLoah|(08oy*oj4Zc;H+c|0#
zO;^okxY|hWimle^BTxg*Py^6Cb2Kucfwn_nStUypbcaiZ-&t_7$fpF5`?~9ZwTFj?
z3`+f^rGOH&zP@hSTlk?nevhT=kU*7&oj-qG9FoQg@aDjU`bIu{`0%sTIrsUif0L95
z)Ya7?o~rb>Q65MU2Xj0F6xo^nvYtOeEC4_VQkxF^b3=kWM~Nk3yl~))35Yz&p_5@0
z=b)L={SnUob2Cc&XW{HW@F|l&s@Wew`7kyA9oy0Gsd*j%_NaL7Uj6gKqLj;(=2g(J
zX2OJ>b5L><Kh<TnJl+EE7TwQBtAQ*?o@8njDMEbZ!XA*$J2XjPNDT^`DGr`6QUPF&
zAiWtZMc=@{u4hNjBiQ>l*SYu)@O}5A9&{&~f5^^@za=3m0jR;EHD25Y(IueZE&m3L
z8~EX2gs}sQf%5#Lw5h;YgK}}W)&*$08A>G^(dCGQ;C~rw7oKcuYr_LlQ3AY%8l9qI
z5NJecA}28{@zD66JNUi;KVgHkcF_G(-Ww}3>A-E$K`{dG7(4CY8j=k$(Z3@EFumJv
zm;M(N<@6=P6gXuF$fOO5-2T^>0}y8}y9FxF)nYH=Fbe>JDeeG!lmIlsLuHWM*x8)J
zg8*c@GTjYLP#JVE6u@EA<p#<EiD!)x2Mhl%t{clH?kn9!fJuWJ43(hPpzqNoB_~S<
zK>`A4urR||&z!jmWUvk8-=WSG!2sktb{xZ?(*HbJS72oM9@{Q^fzw1Jt0Te+4ZI!L
zTSN2}9LR>5i+GfZqF`55j^L%Fq+|y{00pl_T=L#<YBD9asb}>efBlyq_Z#+s0;|}2
z_wEUMUlRBe0fJ%?^4KbbsZK#dGo6(Ul!T%7-UlZwc5ZGt*aowqC`ANNDQ(_`jHd^n
zf&3=wXMlT40x3B)5<%%G4x|RX6S%#zd*3C>Tn_zH-hfcs({bDlv0}PpKH^ZlEhvx+
zp@uAro8l?|kviaaV8$uK6k1N%?W)*G0^(DAeCQ%O4%%$0l1N<`^I@#;h|VgOOA75a
zKN}=*sG^q)ph2N?s*?j16(ynq0$n0vYlzw7Y69uIya<Ty=XN876Rr=9Sc(-w9MhS~
zCD)a@ck&dJlxC1p5^`I6Pv<oF&<2l$=-r*|zHIQlvwwjs{sJ0tYjzJ$l#sm?q^9NJ
z1OY5t+RP`I6T*MXr~p1Ye%HRH_$L{48T!^l<CVFLH?DAX|2G$CKQ1SHRuVSb|4au8
z>&rJL*~`EzaQXd8!mfd@OZk)>2qEp6ngXi}paZ0WENZ%*fnmFV{e^9Q--Z!y3LRn=
zmF(39m%)eo^PHCqNvU{mL*kv@V=bPLh={SMJj}Acv<R&@=PJU+aKZz>F_m3XPevoS
zN4$ZCAoIO7Nt#g-X71kU2)9MaK(jpaZbj&bK~+Q6R^f#pcLK~S8k^%%dw*54z_P<E
zyKMB;2|x%+&>N#;W3!>Fvk3?k{82vNcbi8DSj7OaQ&Ce(5!)IfpDhL;$_7U)Kuhbe
z$p&Q}GS_A>X<==UxWH%j=4$}0CKQbAGe%Yj<sqH_SG6D7*C4JyG>9FjeUKgO*i5w#
z*PDcw4%c#Hah;&M3S0FH1FbNH(HC(+|5g&YD3t9sMETHr!(&T@<19gmfaOigd>Rbf
zzK35zXKN^RoUTC&#X+}$tp7xZJ(Fxy3Idx93@IR570;D#w#SR_muv-#KLyUmlF|z0
z^*T>XOfXbqUQ~cAwt8)wQej8B;5%@9K%0>q{5j+lNLzcr1`CC(J?#1-AED5OdVha^
z6>x^K%2qV8v%6tCulAQ39V933f+!F#_@TBj^wiXvmTR`*Ksuq~LNzaI1Qi4JfQ{}%
z%Z*`_Ra3J)@QSpTZEk=Jp#-%N467a$6=efIm3RN+Kaa(Dx{m&p0iq&1C|Q7A_%B=s
zpe_m^Knz6jh-a{+qN@IRiv1^8eSr{8WZM>|94Lh|Ngyc1f)4w|ff!tsxc_3FQ3hl5
z;7azDIO3o;wx_AEgYa<&f_Es_7eKqYJJaXt_8LF5vcN7$o<}3K{{_XXb058A)77Gu
z_LH1`)N}sKpjohu^0h=dZ=o-Q$C=+Aen4n1B=LcUgEWWDXaD<1r5_ohJ#rieH9pw{
zea<tjCbW=@T=v*udE!lHd}gMPH+FR@B;2jS(MI=^g#d20&9S@@e4ENc5PghY(xguB
zU`uZ6e)1eZg_sWI;M=$75G$v>;?NE(XHRSg;Se1{5ZzhRx1a0RhQ93oiURZU;&Yf*
z1M#NgvQ<G3pZSmXOm~QAZod6T765S0?!pGm;QG=SI*H&i|7pL-SF56tlMqxDmwv<1
z0*89?&ATtgJ|)J^Gy`dpYiVggh#XA#-!?_Tk&)>@OwpD=`w2$t_W%&%a!2h6bU@c6
z1H)`=OPpLJ(Xy{t0W944_;25CL$Jea@pmnXCjQYXk!A!-pAP;O6BAgX>r2Jh6&f-U
zl8$Xi*?vXtEAJ!V66}D^U;{%Zm3MY7f~X+)1qKe_(rrEJB5?xFgbXD(CZ#jA<+-y&
zPkM@HFz7MD!>}t@lH1^ja~7m0Z{D~OMhoPnVi<q({b*BcHaIZo2oux_8=S5K4R;xQ
z7Gz!kaPfj)7&xU)fc>Qg+W=5@v}}wpoR5U}eADIXm5`8tv<3fD7q0uA<Xj9pL{5&(
z8O!o@!&nk1xK^a&Lfb)e4@9|;4i^Q-i?EV$&e~p#aF$ONSQ5$wLL)l^N*Qt|LG#H3
zNbDAU6dvB{_e)Lad=84k=DXlc5$?&iw1vRbfe9!Da#FTR5L;Uz=R3C0d+$RENXBrO
z!zf`Nytkr`PE3?+vSrB(zod%KjnLrWuqfaYyRC7AQ86)DuCqlF$X!5D!sHm2$wxjq
zpe$}MoB{bA1nYZfzijC|K4=ZSC98OnZFOpXe*RE-4bKZdIncf)mBHzSG6bbzszz+*
z11yaN;;>uSA0M2D<9ESfVJVP9n*i6i3#<!++-|BX!2MDT#~Pzi3<gYBk*}Nhsbs)=
zveH1I2&t649zS}7%s|OjXeg7V`(bh$)1Vx)zd(OVVj=@r)srB0;0z3s$moPq;0P7S
zkvg!!=^0O9UM^1V6yvaZ4cny|8F%1mAxtYL4zOA-P;?^&tYYLC1}+p93L;w`v}l)K
z&zYBDkVFQ9fjv6+*795!$OJIU*&yNwiA@{i-DpBxh#ms5Bh3vq2VkQNTMq!tkTKOi
z&IH=)u9LP`Af(seWvjn_HLda{rNcd@gSG*&G5h(!Yq8#F=V;q^sDNUVZQDO$!wS;T
zf#ZPOJxjSNA)g@+Jrw{JvP&Oa$Uwvbh8YrM<-`fGBjd<<Mu}e-!F;1|f~`$XEOT?V
zr0L0h;2PxFO$3b~xtp4rMlYd*r#3c=j86%{k47>zV7DoJmo5&Z8;^~)Sae%H<n*1;
zGVI{WL{@(2qkyVNzku!No!7^Ru;wke8F&bjWi{CyrkVII6R^D&02tlR*8n{H4q!}2
z7&#ePCrnzi7fP_91#70n;~=k{o|l&=GzN7E^vgjKsEb!Jc&MPdxNZZ<8FMPzBj6A}
zRh5+ucy}lKTxfj~9G6Ad2LhBxN`W5Tjy4*66&gc}`yU#GC@E}VK=?_E#hA1r6jexc
zbSCK30BurAOSND?_j>HTShKmdGJ$A4oDBt(0H7B|^0P1Q{b<luN0(R{dC4beX)^(E
zqH-$!<;#BJ<jHp+0*Qk~iMPLn-Us6fhZX}`5O_PyO&%+kEP4wSP-%hymH~Dve9po0
zc>H;ZfHh$ttnd~385Rb%W9Ikos1}`H=mN7w$QllWoeva#3pQ$%(OK{mE*lsau(7f#
zz#=vT058^Ea3P;ZUi$_8ukDv(VH*K)(N=_4l~;ltjc5bNs~|Deiu6M&5xm-nFd#3x
zgZ2dFTZ{E7@U+HC>qmZ3!n1P&n^7OJLJ-&Z&3~Lp$M-V1fuw_Y6O31xJ1}CF#_S1X
zWo0{u1I4(icuc>8Qi)0ssI=$-RP`Kyl>jh6l7Q;6O(O6`S4+?wfH7o_Hb$G>gUK9G
zu`ima^j`HrC=sb3cso3VU{M^u7|N6adV5`+6w3SnGTQ+650WmODe=%zc>$|S$;f2d
zdO)N=N><kL^9`uz*nJd(h~T_^`*~Xk(i1>q5d>Ax?;aey+vmBvjTHGgMc5gGK(I_j
zHYPd#c{q{^5f>*W-he2djtErY+zU8w*MB|(_7_xU=idW#0Et(;Wk9`<{|hmBq{(S<
zg@T%z?>MJ}a2Uv3AbPy)zCeH(cd}H%g+2TqD)P_hN<cw6W@M#+9RQ#`JudDlTnvJ(
zKwrBdzcPR5#9io<m2j#Nj{O?Ise+?N<6$o4o`;s;SXxew+MdgQo;D{*03wUMw;7Sr
zcAeSl4s7h~4FN<@FvW@DagZtniTy3q@CF<5;H!~!23GnwjIwr=yF$_yr24<wIF$4~
zv4h+3@%u>>_20F--y<U^m{QQ^5Q*j)#P7yIfkJ1|<1BUN2t*yG!{DgpGKf4H6`Or?
zc4N_w#bcllCxPgI9=C|g^1O*Wh7ooHr~4iN4`Tr4I`^M)ir_VY&yg|AEgA!ha9mXE
zx%hk3?7{tRN4XSu)KKA)5L`xQ<SMzH-2buzJ`VI@GK&6^;R~R1dtK(L48U$gpCO8r
zJlj_?l~d291Tru(J^=eRPCLPKc7>}YgA-(Wr|cUwGgU7~H%cFb2_}rJY1pAuLUt+0
zf&_>Hhe*ef6a)<!!B)g?!EEHY!iHfDXx#tnM)x;%`k80)8qQnVBiHwhrezE0*C=H$
zj)E6}q8k1HBRn0-i2M&_C$E2>1UqvV9k_;mon^ZNroSS9WuP-T?+!uuDR!Y4lel$@
z4Wv-Wu5`XUE6(sop@kR_2tJyCrdCaE!`eqz4MAL~FVX^JdjL#KclH3^f7LRfN*C>c
zc$_i$S70jcxe`>UE>{{B`*+Fe8bBYl4+BXQb-x-q*E=wRkoV6DZ-=r6NKxG*n2?lD
z{<D|h&nqB~h9Y#pb0<sB**Q3@Fp>bKAG;wTf#i0o3-_LYOZWQ%B^_NBw9h`br3R)X
zBCJvjrV}nhEgq#!K#w+s-_^elvo0UmpMX6PmjDABRN!%d30+83e);l9?l1w?JPm*=
z79w}ZpzG4ImC^qVS8ntj{1`aO6jfsjIV<o_)1VxflHr|po!sdW$VO*4{9i$_D8B~C
z1V|T@eI=Knr<7oNF!@v|%)w>1H%y6Cf>EZbqeJHkBhmOiNXf;MY3aS$RaI5oYca)H
z4e(}Mi$})cXl>crFTdg;Z)%9GX-wGwm=2|ZPZ+xT@qX<wI2eLn9)<PtfdzxpovRNE
zUH~XT*Bu{HzBzjZx<W?@O~q9r%-RIH-+u;6r?L;tXQ<}_7ZVg&8+>fq^D&d-q_qBT
zSfh?o@JVOK{(acU{G${+Hztw+(Wfpo8(~<p;Ns#U%7?ixw@3sv(psS8JM_azh<Wf_
zg1z4m(}jCD+m=j@3kRhB92m$QICKMsatHL+`}r|FPXZx_08a;bvd<5|a_qqMBdH3}
z04T7p-(by5Kfs3^06Kw>@-hP#l#a4WqtH4YC4(SX77>QvSQE_80#I#$F2cDfQivx4
zTE(wXY-Hec-8+~G0M`eHge0}Mw+oHcyf~czhqnH$x;^3Ik^0Xt0ZkT8XDm6x_#A{h
zNJ{?|4z;}&@Nd6c0q&IlEA$-oVOpZYHINpQpDKkp4M~p(L9lUls;7`x8vz!=5<&_f
zK*G_Yn_QvNh+l%O3nDwpyh5+|HXQ~X#|&`!YOie|z;965hdn&bf(=#l+ZtM@6u^(g
zr6oS+#rveEFWv|`K~K*uiEJh4ULD{a9Y0P8FpMt+gR^f<+AitHW|K(<h7R0D6(k)<
z#3<|NWTm9A0K*HX1wpr>IgU_kyHvLkB`mdMla<b%J&O{H=!gfJ4+HZXqLm<!&(~fA
zymSSo9fD!d%E0jC)_zb;0Rl%Gtz$3p=_6>#F)zdHYa~BLY>MdN+#g73%cs(J1VL|u
zTt@%C{erHR3$Uns5JW%{28=e!?WdC(HITsx4i1Kc-1p%_37o!Dd&i^r5`0EPECdmf
zK>+wRVzbT*%N%;|e}SUHqK-icWM1>R>httz>fcbHuzv*T($7H*n3M|AxnUR<X>o!7
zb_X$405JW7O@%N27#%u*HvT@+2Y$v#dF+DN&H{NE>@7%hJcqrr46H=~A&@(FmeoOC
z=Bh?=9|SUiXap`yWtGT+A_=GmNm<KZsjb5HNCGU46QlytGBSc0PyikLxu6b>&BkG%
zr)Gnd4@otO;A~LE4xSq|Zx9gRd9|VwS3oN#AyR^LJ)(+ReX>!4@vwJT*x9AQV;&kB
zDu#E=Gi;#3&-CD+usql6b(fRHwk6?5vvY!g&kw@qqIElgeVGSvQ)uisOd4oiyn5#V
zOUMD|g4x{HqzmWEWMYK7nc2a|R)lDZL<=b?X$T4)ODY0q1`@P9TQbMS%c~4-DALYN
zPt1b~3c*hqq+ZpmA(ak_3(Wn(8Wcf*G(-J}%>?8<RDgB?u9=Rhn9ivOTifjSHptn+
zjUQ}^wCs0|z<lRMku8K};o+D@I~HjH1JggA3<JD^HwD8L7}TGP%`AndH)|W(=mL0w
z*KS-MNM{|-Vfe%W<L0<)D24<$41My6+qw;?+I_?B-g@mW{gz<CBR}^7kQXya71ubJ
zA)3`+zOX=c0iLvw^J1c98D16SxvZ~r<6!V2HG@u{Qiz5I!*p-|!M9m(bKTqL7t|rD
z1Q?d@qd#nj*IguqAK>#epcPG_?hEaJQ4*#u<mFyLOJnk@J_j+ia<xuK4N{e^^09nd
zGNbYy@|d+2%mb|0XV)g<GLaMtiau&z@X;ci>N#JAxlf~y;;HGkD(P#i&cQH$!)jlO
zS*4f%8+2@Fa}SzLfuVwz*xDcw+GQ2%4OVvWaTp~v6+0VdD6$57)gdsNGaHm@ASw3r
zi(2aHtW`X`kj0k-N9_@v`?3N;P!M#Xn_}nW45_}Spr#h~`SWL>awy~P*ewIm6O<eV
zmYxa}V|Ep!6_6wvw>KID1|S)=ybuf9LdEcVn8NV6;C}_fO_gpDW0l!qI$l8h&<6=Y
zicu0m-*mRGUcG`0r60=Ytg1mg28`(hI5+(_K@LhREfq!{F{E==)?SQ{8za03RS#Ps
z$PUh-!j*uap8}N<l;(?H_X&RBbP20OJPLg_K-|p>_=B<op1WoIR{a@(zPeZ{f=*CV
zvx9;Hf-6)LNTZV@p}OdinQ~}$4K<^LAADv_9{3W#Z-6GE%o56-E34!HQAtPH`Z<GA
z_%SJ+@DFBOJlr|-aq@Sw1UrzK0#fxB(Xp_4Fi&6^(P<ROn1HvL3YHSWvT*RSRHG^C
zvMr>wvLKOQKi+%|7FGqKFTG1(0V3sy=}y=gz(u+{i=jwo<6hs~W1$h`N{6h$A_nDq
z)FspeW_Hl)*T{F6vjM=7^<ZsD9XdKhOI)IHHHV>{JUMW@*Y)Y%-*S=*d;oNUVZ_AH
zAK>fE8F_#p@I~+tCB9=>+1cBm5gR$c{MsVh;}Ea@HRxwzLI?}}^*j*_2PH_!zz1~X
z!_f)#%rX8U2vRD7LkhZ!@gC<BQ?hyt3K5kd$oD9)2);UlB&c;F1(2fr2Sowar4BJe
z<CP!i7y<n^5GvqnPpD`C2p_=%xquxXbfNTr{Re>lXFA6tF0~&N^?z0u{^oE|d#@M<
z`aBq-DCh<!24TdY5FyBjDj>prwCqcRbZ9s&1TGEz_uk|g<n=&pL+s$JxcBWKkZ2es
zA?x_F%(VajzvZ7hLo4a=_P-suSJA9(`V)NoJPpijjdt&^eu4U^3Pc5RQi@{~@wvIV
zk$V8*aXLV@!TqWLUIC|0CIpItB>^I)b0&~Ad4ZL^8%0M#{>G>N@WE}^yyi4|8*)%N
zCm<xje+7!t<Ijzpg2USY)V`z4vF`8@W5g$sApWlarc4x)2d?eClpDzW8!!N1R4dQf
z*lW{*xxDwt4<Q*5R5z63NrP?x=Qj43*W90iK;A#dB$L9+kRI1MMfnMOIysCp2{NCm
zz1!#8yZyU>kWfNdSwTT>Pa{3)vv=RfeftkME~G<a$#EJ1xjinrfA6mcVObO*+s7!$
zEE_c#Q;DJpe|`Y&tH2ciu^xf8|EO+cPxC7MUm(lm5mLpN(VsQ6gOb;qRImY`i2Hx4
zyYg@-_rE<VEh;*{7Nw95N1RahopKb-DP^n1APNn#l;v1QoupD|vK(Yd*0I&tB4b3N
z#2`y{LSrfWSZ2)leZD&Hb-nL<z2~p@djELmAAgu<=6Sx~=lLx6eShw8b4M+_qDdQ_
zJ#yeOBzs}T+LFVD3_h_NH-V`}P(E6<IG8eEpvSRI8bXZ}=;)wWIzzDnwDSZkUveyD
zy2)C&Z0Osy-jQ|C;3xDViz)3uAEJy7qBBUA%SVWy9ftNC#R#Z6!cngS^}+rCDpZ-l
zXtd&5Iz(bvrNKbq0|)R^vv8kMGnrSR&IY(HMjZSlzNTmo`NKM(Ib*1Ek?aa!pI*-n
zOQfJ>2flms5}lXc+IithZ=}KZNr1%iObb+ESBu&Mr_BKrQ*s-q48_q=-wxO(0Pjc>
zm*W{bJA1?K>eV%2A<(>cp~nOeyfu~q05ssbk$^w<TtVP4ptMj30xc<2XoWPW>H(}m
zc9G3qxIF}!mOPkS=yudqKp(z12ct8V(2csZGTY4TL4phHGpIFFq0EFukpe>u5ZZUZ
z>Ny1%{nu<BtgNZIa(hAxF=@E+KVH^eWZG`W=<2!iqbKn)Fxlz3?CHU3(A>2^O^#Sn
zko>@gL4$LBQmeIS3*YQYLN6fGI3iMkEKW@Y!n8sg_=jM|t3+VG&*kUSxH!bJf}FSa
zibH)4RB3Q=AP?4D?13~7nO_oy`+fb-5(rtK4utI~$E^Y-t9ZE*vp4~Y7N4j4(K#U&
zLzfD*bsOmJsk0w`oqUOeeIPZ3#8(DA7K)hRi_pqWfHHA%E<j=oq)w3bhgK{xK?Kze
z@WizuIX+}=4ev=XVgQ9u8?1W{)L?DPi9k|k1K^KT;oGVahl}Ke;Nhica|0lYmuQ}?
zwcfr!3K(<5bi$Yo;117W<E6yJp7kxF69Kdiids{sl@-B@0?7L8`^La=?nKewfRtB0
zdm0+LcJ&TWh$C89O|A4zk{~Q7P=6ghX`^>@1fgX5f`SSLqpO<?4+^~2twV=nFgtdj
zCkcwZAQS`P^*?7G_{ZkW2<Q5L@DV$%dSRBrK8>}4x;|59b(xM}+8Wzq8YEC3qS05B
zs)SLR)eU-%DcJ!*fyK1+&Mx$HO8CsVuS_K3qX%{rTzQYfNi#BDE3unTMa_2i#|cC0
zhS9~qAsCRIc1*huF{`JNqJ8mFhSJJvPeMp0zFyoX4<6oY3NXotOF*ls8}#MCz@CAu
znou*%z}v4^lr?g@M~8FNd~M$-AaiTrx?n4I@xJ0+T5L>^XU!{VoaCr`L?|f!2JN7|
zBuf1|k`GtPfQef0u75duJ={vUbBA1B-h-m_+M$2%;;YT(2lsNZWAg+;<Ia}kH;i{l
z7Kf=E6#^&TcsM+D%p`3++iCUTT)*FnAK~LDnL+b27E;ZenMHR)kid&gsykE@%)%Lh
zhWxdq?j5o2i%uP-c66d?qW270xSQUPTcFagh`uujTPZT2YmAeo7USa7bGakRE>VXL
zjVG#XBUq-l%AP0_v+A00X%iW&6J1sY8W`gD4qnxW-~T#dmlJ(5jHK>G_EXDF<I>f<
zOj+sY;yhwJINQy5m#kMKC-<67b)Uw|YEf1rs*EE4NTSBF4E1?ee}-IvS=m7J77*Vl
ziFM_g`Eu391VSgQT!xRn_w8$*>6|i9?v7=+nsMd?OHQA+Bqw%vSiQh#Y}1>dan6#<
z>sA_?)}M_z)WDBcH2L`9%E0jpujf2Q=Lv=_MB(k_Mdu9e&3|?t%aTkjr;l6ZQ7Gf*
zJ9_p{q$yudb!jHm9O4)e$Nqdu)~b&2WX{H;tJ3|HjA3W}30^v(Eflig-SeweZQINB
z$KBHk2eQcPEaF*>Jbfi4;}2R$it`_%lIT_j6H}cny(aq)q*+FsF}ECWo3L+moS9ld
zA9nB=W|L5;ew>C;octx$IIT#|q+hKvUM11bio@4Jiez3G%kvY*R?Es(C~(zl?glGe
z>-;(%ObZ$Rqh2iy&468NK3sV#vo@+_z#`3?d!Frooq-V$I?~{eXIXf);TJTF-KI;I
z{idQ97+7n6G3I#hADw751`fH44(KWkHJ2^6nwz(m3%sb8_|7trmrm;r9JvJkRKufW
zqDyDbtI{9uehIfqxmkT;U-9=;@6{2t=;GO`q#~`1g_-V<3xLtX`o`J{xMAJKQ#;o<
zo#ak<WHwj_NbDAjPxw-W&)aBy5Hu`6XFO7F`Iq>%$cojzZ@ZY;nGs>4F<nIp8tAAX
zj6RRc78g<3T)bFROngGBOKiR&hlM@<)QcbfU3sAJar8iX5;lopD9|I>d+ancN5)!<
zv)84AtuI8)(iaSLRt{>hpLX6y-?ZT<pBvV_L(9zX;%cd07rV5RKwe?5mr4mRE9P!_
zx{6Mx(sb8Aw0N;Y+gf})xBm4oI`T*14S)YP;ZLmY8$fY;1D6IjUGJpmgAA+{U}>rB
zwv~Q1QhGv}r+?~HJc>yN4?;#NRL#$~$d3)z%yUX{F^A-OqeJk6V^ORy)~<1<ue{x@
z$=?rT>C@%bq8bby>gqyieeFER?H&9MCSow#B++jK=rtrP`*D1}Z5sdl%W0#&M1q&N
zz~X}6Nu%p$T3Q67#~$}<J9|Mp9Ta5CJ$4H-phuU(r;Kp~v#!J+E_mK5{drGHYzbE%
zYi$r^v)3NxWiXo$Lkf`jkdYLa?j-xV^V?pgqiElGvErjZjQlwJWT-C-4@2vBy!dj#
z75upuG8biI>eqr5PcYOLYYU0ruvJ&Ll+m_7%6iE#_oh*p|Cti2giH-%xsJsQJ|<C=
zo8J}U?Limt(KgPg$IiC{SP|c}`3ycVfuFcoTI*2AN64+{B|hu#TS!qa{@#+=kD_^>
zjlj{Mpf5QdA1+-wKQZK?pC$g@(>`{t6u(<!C=aJS(K;qtcx5U+?uBMvI~@)P@_|0a
zi1$=#hdGwLa9U^izM^evtK6G!ra^QnbF^fV`udYO^Hbt;(NVc>#s~WuZr~dt`}*Uq
zV2%OF)ckzP{rVw~Z%fC8rMfai67psUF7TmvO}?Xzot*pg-TO#32C^|8!&m|%q~(qw
zJoyc0Lrzk3u#P4KFTQ^FXk}-$?2#Uxqu5$IC$mNc++vbM9L<N)A=pg8(WAZ|bc#6P
zoGWgq<JjAoO^4;z9ZJDGe_&%k*OpPUB$0dL3n|7QF8;h-<dA2mvW$oj{r%6R?4EDC
zDy}F_SAC%Q4%UiDTqMa)c0Z5li%nas_<Y3BchFUU@Ge4Pyy>eXGqC+S-8-Vq)+}w%
zIPJVtb`3d}=5P4DTEnu{8?Hom=qMEf{uHRPbS1LUAaFkTt7a2dxo674((9VO1+a5X
zmps4Uyep>l*OA~5?4Ho8j~iENV>{a3HCAYAFkU_!>2bgvDSPhu3fJjOdAhIU;t?eg
z`$sDCT5LYr`QT4;;nf^^oOP+=@601pR<FAJ_OsL&9H&etKmo!6=Olwoxdwa8*e(ux
z3NyRBwOw;qvt(uatUY8FCEC%t?gp1(YEip9k)N2wDkIBTa}=-#BteC<Cz!V+kEs!<
zde)#NmNMQMO=As41O^Q0u|}u8qN0;pyOVN1`B^)+k1nyJRze-@Ojq4A<K}d-TQW^E
zw&F4Mmwp-1meap;fm*>;F*f*(>I$o0DIrhAWTdB)0bl!2R~G>lC*T7}TLc(i$@})^
zUHklVquR*k=KYq{A79{xGadTA+%OyoBg^|2uitxDd<c*iCdIKz%e*MZ-DSBzz;Kx?
zZ3l4JAIW1YCfJW*12<{?s)H$gi}yGel-`}a5Gg&^PtV2<@0RvXj7}9|Y)R7E^f}_0
zQ8-uUj76J~(e=1gZ~KHy?|`V?#v`%Rsg({92);A1xFDl{f+Lfn=v#B2OUWzVHDXy!
zeWELD)9ZchX#&9Jqjio2xFQEe2%<Xt%Np-~Rj2LkF#D@vFOf;LFva8C^2+xn>c8Xc
zzvbANl<=gHb9UgE_se?g!?iw13bU-;j>27?^E&m|eL7{;5d|NC!&)HHlACAV^YQcL
zvTRPaMaMi-@v2BoF|l^~vZ|Tal>EYR&OzJauGhB!y%1V=wt`M`=BW&))~p6QvZC~E
zkrs0svFoP2oi~GID5DuM96ZSs7igS!I>Xug6VU;{ipqWI+=g%e*yA8v9lBI1u-3@o
z85#~WrFe<n%`SEA*IvV`>>Fksg}1+VYe#HTObjE?5_-5}Q*`JUkm&?~M;DFTYz9ft
zqA0?kpUH=*PJHxu?Kfow;e}20I9sm=SOt&KgZjmjHeL_06f9{dgv@aLZuR_ayoh;x
zQ&+fTA9s1bv^uk>ol9dlJzc*NnoZ2L(kVS2T*K>da$6B$H^g>=7PsRuDRu16#4Hso
zlb$mXL4e2auj<xycn}K$kLmLe*-VhY7Y!*H0Fy+DN|?`xt{*W5ojz(QpxE~-ryt!`
zidV`Hy!p%%e@$Ccvqy_dTV>PMdN9iI<1!9YjnD!W!q0YSappgDmEO$Ad-a76xr-rP
zTxWA!ZoIE%GG+D<gl>|#*kk)sudeEp8miy)pyd3ze<87}k*9Ilwt&keq*v*6NslqY
z$FoKGC#v;U&)LlUWZ5rse8@HMcg~d%YgMD)3M*u}#;0ig`IExneO4zVuFnbI)tNWW
zjNN*Qtz<#FC4V+1R(ObX@q}JO7$%wLGgn!lt%SyX@ZEK;DYD!8Q+M|=FNHN&^54PR
zbkr$U-Pr;YU${rqO2LA#yC0KKT@MShI03U-Ws8x;y8vrz2Nt;@Y*2+&qv`0i!QcVe
zt3kb+m%Dnif%;n0YQKEsY)3PzqvDbDuX4@80h4Wo7KtuRG7sf5h~^VZcSHECr>X6i
z$0)aqL-+;pt*)Op@+)hs3&OlvDDk<+_rmF&@&#ui^@=0UB#e6k>7w-x*e1J|7ttR5
z=k|2Yp*{<O1vowQhG<?eHwQp-vF`)qd@WEv3gT;nzh;O2@J?Yfv2s+?2`dN8*X7yI
z$Md60>Qd0nB55Sr@v~lxba8<zJs-Ld-8maao-pu!gHc}1#v-j7!}or;B=3T;ESSp}
z1<KZ)mH|=)kUb-u_;2LT5NXIjcbf{DGQXm*tqtp^<uZir)IC`t>U5sGhT*^hPw%R!
z#@?!|EvlA{8%1kS;oX0f>1zECh^uLDmH0n3ze_n~+(SO-Rul;$mJY4<WZj;cqxyz7
z2(_k;l)QC6ArZFI5%8fWxu7@4{33YS5E(W?YYnc+nXs2&CT<2k<Rnv1-Y-A`A;S<M
z&<cQDng?{RrZAtVZE;HPE6-B+J_XUHU#E2POu}rTa3Qy3wuWkmG*dsN--v`-Qg>&;
zUQf%JErN>R;(=!Jfiwl)h8KT=|C$&#l0`~ftv231rgT=g>b1d{SKcMEezY#B__lOT
z?M=^KdcHYpyl{`_Zl7eSmmFRU{s?!0H@U`VT1DX6+q%K7GP~>wsUyyNyoWe8rz-MZ
zrMzwt^}Pgy>5q!Z098zg>yPv&QaX>3V;#KUVuOMlSjd<YUW&YYZtIbN79gtaxCQWl
zaUF)XaKpx88no>x7?nyeLGgyJ9}9}avH=Qot7R~^qy%OPzZh)9=+@UP2x11Db`5Mq
z4lrvolQ3PlCNn4np$^i0k`%AjMf`Gl6TJEDf4uqLy?aeaQv#eIaG94pWlI1h1M$%~
zNXJ`cN_tv6fRVj+>oa#9l<91YN<{rg3zF>tibMc7SxD6_xi(;PeSQ6I@F)NPe$AE!
zq*fop0R8R1!rqFRmRNwpK-&y+K;njA)A0hBpxPCfW}62>;u=s0&i(*S|Md5D`T;0@
z94I;#j1)#CCQgd<gZFo|tR)%E8A(XgyuE|L6lyvXc`y|xcovXEGQ&I<N0{s>1tA>?
z-O<@8DJO_a)uL!H;R&K$XJ5d6E#3YO2Jf7y2L3@K7;6NO?iozgV_~cVXipYQBmm)~
zLI8E>8bk^lc#RD@d0yGuo<D)Gu@(4oM2b7CE^#?i#EJBH7kM>kw`2sntwF8!Wyfe}
z6Hr*7-mjz12&=dp*bly3Li^VKX)7?T4@z>l7tr@0k9UaWv7l8A+7tdAe(K8jT#(%H
zOP7!XBuH&HA=B#}N70JQuBn3%IHQUHqW>e5=FX~sug8hEK8Lz5qVF)&Kl@wWiOaYD
E4VU+v#sB~S

literal 27281
zcmd?Rby$__-Yz=829fDBP!yOd76^!f(qMrDR2rl$I+c{RU@8KNA}XnXG$`GmsE9~N
zN~4H_?#^>RIQQD$-rqWBfBT%j_I0@Cx^Mwwyzl$O?|0WT?kb#<VParqpin4G)YB)G
zD3oP|6bfDY>XrD3N$2w0__p$roXkml$Jg|U95;MhV|7}?nnL;0hWwAN)qul+LivkA
zJ$X#UA*ip#N&B&WYw_^l7pn<Mg!H~O*ROE%7n$%RoT;``R!=y&;dQz|f9fw~9^+n#
z-nz>P&pgx91^B(Byp@`4(#+Ft>90zwdLkxeWOC`=-$54*257!_OV%q$uPYJ=o$M9l
zGf3=}aONNVRczZ*Z+>WP(H~1UR$yfyziR1UVZsWOW#nJZ^eYaM?}dB+cYa{Ctl4$8
zzo#nFi;Ig(jBN#l68d(Wg<_Ss&wX3_(D$OxWo7;sw}kl6Q`{@sXBa6$>KhBmRVblS
zbZg1?p*8>YAGFiD@rt&KTRbGD+Lz7u$3%@KF>M>`s);%4C{aLh8_)V7>%E2lI2|3`
zSw~(9MUyMBKr)=Nk|N+WGIGs(o4D-Wg4uI<C$!?kW{%HuiN;YYG>lV>%iI<c_OiQa
z59UsWynOlc`Is|Btn}d8`Kb=yk;%z`hQ8^$KIV<74?kQQwplwRlvsd=qob7lkl*j7
zLfbIyIAec8-{j-yxd)H4_R4ffdslgey74}W5w2JlCcMSt#r4t9ug%xxdp(y;`$aJ3
zcw85HtfF?Qe&}M6pyj|Q0~6D@8T-_Mf8O<Mxk1YQj`8~Ovm;Lp@}|^s(qz>0o~FsP
zIgfffj`wulSWbCfb%MDq>1=30>>D>8wbM;KbsIwNb5fI4sao~<FM2J$y{gD?Zu#-*
z=8pj1Udyu~q1)&wQy)Yf$D+$EtEJz%hRo*mM)tq+fApy1&FC^C)~u{7N9)?&?CH14
zUTsg_{@~LQOjdfkepQ-T19yneGLHA-8G9HmP^2EEH1#JtIP^SfxZ>AhF=@1t;yrka
zSwf*c-#<CE{kpeK5?5V8t4@FSmZ4+XZl*6W?iBZWuKoKT>M2A<T@F%I;Eu`s84-1v
z^2a2WoaWyhot=)3rE2;Ejd2?~H)Y3mta$FI&%u<nuTejTad9E}+++29gUpX=KDV_s
zEX18k>S-_8$>x0F<?&@%73(POGec>;d|0!E9j^O$>mt4Ml#TsWkGXEPUA^Y($N5#M
zR~PBTM|vIR#~b5<n>73FRBI3U$0piQ9z^(UsJeOAXEOX1wV;RASYY*1MY}j#Lpk8|
z4)e%U%+%s_*%K?4Q~C2EUf9c-mz&NtODT_h=;c!mqoc6f&psdLN=P<o^bhFmswrs{
zonGf+KN!rw%uo{~pJA-GLwOaQz?r(eel5Qo%wsz3ntSpab<55B1ZO|XMd*Zz-0M)z
zSdn{!PuFcL-Aal>&2aKH7ssyGU-mp&xMnUn9mrM?*#9j%?T`#}^=uo5%EeQ6JeBE(
zOgzTAP4;Jc+b9M-9)ENBWTq;QFM})RS%$U3qu&>3s{1q3I$Tt<!l!tIc6M_3iP%`o
zO-U&a)h-QCnBtbm-9FzRHqOZ2+g%i$HmGp*4QI^a9kyVWHk103wl#u^Y^E)PbJKh!
z&YwR&F${Lq-*V`4>6MKvwOMy#R832Y;`Hf_oQs*|?i5GRWl+E7ZC^2z5i09q8JlJ9
zvam*IJd`WtRPvln(`EA&wrM{`b~R>OzHxec!}0awrk>o7?DKCF)U^6f#V6^y?Z?jX
zerd8$eTc!d;6`F>=1DHTxU`Ce>QkbXm#+3YA1Qi};KS&ez{RESvP;Kb&rOZChpu{&
zOW7z|OPG@CtY>-6IWMAMquq^;`6sGrEQQv7+HUJO95Xgb*4PFb`mr@_v6E@Jc!+xR
zY}MA#6G3`@34(g(OuikZuld2;G^i5zxh`#jjhFYKo_@desLS|_9Lms8^tl8!mVV!L
z+x^N8YKN##)eV>>v_1-7HLLq7|KWPxS+-zD-^S&0xs6xq-44{gc=_Yk8m7YVh55E(
z746}0&hFl2ZJ!ldb}4x1%s3}>D4y&O=X@5yu727n-|GI_P!<hqr;gN$G`4E}Z6{1N
zxR$rAYhR!5I3I}};-E77;KH~>!L9s-;pkt^*{Z&C*+~cXJX*JTQ*Vgdyv!+oe@BZ?
zg{-sB<O8{EXACI?+M>(}D#uFf{UVkXcWQcv{PHv_cF9)tTd=$zp5XW<Hh563-rKsR
zHNnZ9sk_~x+rPW&`JJlpq|l<H%N+jxXA0OeoPS!@IF<CywE0TZHdmx`C);xO3+w(x
zotb@bLw8vF7=@>L!rsl*ZHB8XG|4}atyiS80dIL$DJsM#!(ZR0Bd}LE(B_@swwfbC
zv98;e8CHt)3dSV=44N3nR$1DXybO%B_R0e_eLQ+yWi1V##TBC$SnR4o6SI`+ZGWvg
zNuhnz+};0;^Vt_AS&u8ji&>V#i>GC-Mdl7hd`hIb`dG8gnbgMyzSFW5p4_lZGDcuJ
z)cDW$Tc>}F2heY!RPIb=UQ^e6tlnfG%(No2<@+jAy^{umKSiJawTWVxZW-0F&|GR^
z#pAj=X3#PsjZaAWZ_ALX*Q2-X9x(Wd4D7d9EREkf#Q4j!Uhj8x+vbtZF7cy#3&r_N
zMrmYmTWnrZ#f9Ry#+sTG4$G07jZGX5CG%!d*6lM|kEZ(qDU`~Ck*aF)G76C`dn6AB
zh*R7L@hjz0Dk--;cDheRGR(v0k{d%;G}o-9hO&#q%Za9rewwBe2<EG!avYenVF=Uz
zF{8BsgB4ctj8oj<6vtWJ0>w_XCXEcsGaeCpB^R^{OzcB09dA0)n?)y(=ejFWZl#0B
zhi9AE0*+77R<ReNG!LMR2zbr@aN;gVJu@t)SKt1$+@$HfuiDg$=#!=?*0zf0wRARa
zjy+Yd)p;VA;t(97P7Q2(Abaj~!aK3xS9^*c<e%r(5qwuoZI8=b$R=f5fbmg;B=c0-
z9zFIb9?x}g@plL1?wM3en=e}?&^aPBF2t8htyocLeN#)W-ZwyZHbubUX3d+xE5>w^
z-!-`-A5B-ua7R{-Mq0Rt1+UJ(*P9f#d8lF;r7+-RD#O=L>`V?_dniG_s_2v)ZjUuc
zMtUbv?Q)_vpzfYV&22q;Ixo>boPkd4@7|<kOPC~AH;TQgYjN0yfx^S)b%BoHhW{P7
zGPH#=?BS+On~Xo)+w@%M>PtsGyBJrO@lEVym;E+wXwkX6!9v}}CRHYdym81$y0ro?
zgGKFoeto~S$wlLhqwa3ozvrc=6Y__R_MA|faGOuPP*^=vu=iKD&rebD=LAO$ZDhC*
z_E1Gl&E#iSTbGBEYt5ZEA?K_`;(l1{tLZq*QMGAUy~j_Eci84H{B>BvRpx`0r;k`P
zM-MT|X<iua{i8}P+j)wbm)a#PFGHSB>?ZvdBhEIL$>8Ino18>GZne#<tPBaV6lWt~
zZD=*Q+!l`=k4lDY?jJt%Y<Et}FRWZNT$%rd-fkANygjt`MBowgs(YKbRa8_WwpnBz
zvv0i~bzrkxcyN(P8W#`G*=<M66waT2`q7soX<>f$sOicu&5||7Z%-D_Dpp76ja&EY
z2`O?(D9+|92M7&x>Wn?G=?gQ}lX|eRT8`B*qvluc#|kD~)<M+Qmr^j<>ZDoVn)j<L
zFiKtb`Y9P-c2xrdgYb~c-{0LyZt9nnldEpdvi5)WEG9uK;|RIdKYda|C#4`eDLue?
zn^Rf-$e+gbQ9eJOYLw%RrOvXm(WN(-Id(?|j7=6S>g?FD!!P3J&nx0{Lq)QuPa8e9
z4idYrG+3~hf9UXG{Ws?~{`u!i(SghpN%=FeGZXzy2ewl7o|v&o+t@6fz3<%TnQ#0%
zo4MkDz4DEq-q>L%(Y%<;wIahVH;K!BW5EiUh`AF=p5@*g9AuCU*m>N2*~PhXS`z>M
zd$plf)Vbit4<CM?)7H?4Nlg_nZ_ZS_apMN7gmb3!{Vh9E8f=FgCkI+IQw+S>UB(mx
zCgx`+xepyu0vat(kB;7_Z)jL*kW$NiLa^xIl`S5nhJ|eZwxuSN*i3W;&lKEf*s0+<
zW2&olOs8MrEz3+(fk+n1_Gr(X&u(#b3==kh&+a1Eb14IL34uo}TjOy{WQ`BCzs;VX
zXxg-KBWKaU$9{e<Z!BA7(vYm5++e1e>+ohXk7h*4qrExXfp8nl1Zghp!q-j)i{`Am
zlV(yY>!+fqc(rbAlxEho$X{QcbyRKMvgO+rw~a$er#e0LGt{YmpSc*zv*+srx9jnB
z^WD*DcPP+T?O!nz#96=7p{qq{PNCjBglcDY`P=RytK0%(zYdA-+2{NHxwsa;$7|5p
zZ(V6UNJo(>{q@z!otuj*)wG_M;B3c*X&Uh6p=V_Vkr5G}4zBHQ%8;L)o@PB__VmDJ
zFM4|V$B!RJ*!I?ySY*`Cin+Qb+AhTG$}lXoinZ%++|SEfKAn92z<v7J%nuwB^@q2;
z9RK2BG*4@i41Pm7*jGg}jSN#$n%!4Yq1SWuj*3GK{~_(BiKx~N)0z?We8<3^0uPTJ
zX=^MaEZ1S;r|t}Qt+ukVdi}*9McS90r&|7m&YCrA9D5ToRJ66@b92QoxJF}LU&~!$
z8O!h8V_*|=s1)vOu*kredb0}3`UzYpq?`>ra+XcRrt;w~8SQc{O--&xB@ylI?I{LE
z)vvs*3*O#ZWs0dPuUNKfor=D`{)V}d_JFScVlS_kRIm6uuye3Jc0rV4J2d;ey8=~3
zf7FPs`ORe+s?W<hn?@$zg@pwMW@hzbAGYxyUVI`=%-e;+x~rqOg@n$x<k;88s>X3r
zH8eDq7OYbre?Z4c85!=qdu5IUmQ9}z$ClE)diCm|XJ-t{C<0IV2N&iR##En6>T&Cc
z*4che(VC<LMw}C8v*a$_UVOBEbieM=N!{onHJzReeyf%KgJq8u)rt@2Vmvh|zUOYN
z!=1F{yI_%VHJBa2e+k-IO1D-rDJdvC$hIF$#4wt-WUG2yj808e;maMmx##S2m5%dK
zeD_p&4jee+GBft|vp=`~77qqS#x$d^J9qEdQ^&oma!b5wmCsqfk^|!D#~tSqLiN_O
zP&Y*?wc0a@%D6BX;Od0|*XM)__wV1&#m%jJ>C($RXP*A$_2K30V1b_Bl{e>8t-c?3
z8UDb+df4FpK+d29)7C>Lm#tiD)w{hqN?!fq#TS46`6t;$IfMCH;!YFs>MO(e`<l8l
zB{L?aO0G4&znNcf)mhy3_LTM=J1gge{*X)Ln)T1ts$CT^#N5{4LP^QVhorq+YElO6
zRFB=g+-28Wu)W$hiQVN%uvnLs&)C;(RZky>HoamhxyT^Eda82~W7pN99JcJ$qUx(k
z1^MDa4XgPNG4*<uAI4d{@AM`)i8(SK9FkLVP~j5q?%s$iQhJV~ZTWE|N6y^}Id|0{
zovTEt;=>#K!d;Kmw{6+lIi=<Y`_c+3tM`|iExJYiy+uBXyiDxYdUO(-o`Lgsr$_B0
zxGkU5yi?PcNR^rC5%2bi5({2VQDBL?7+qC9a=@@C^bdFIQ@GCNM>eHJRXr=J%9*hQ
zdKM3~z8aerRjMy#o7~qn5?ym|+=UYML36r;-+R%{H|lb#!=%4IcXRa)io<SnuJtR@
zo_s8SFhZ!7>{9Eu3$*E`$Q-ly@oE*t=z+Za=9u>erpX<P84;-^K1`G_%F5>_bI$pw
z%MHc3hMc2N3bt>s>#q%|IwW6Ap#<#$fTK`24n->8E*~A8w7KswT_+zqWV4#m`Ztgw
zrO;DsV=cAi)-7?i@M9q@y;1ad0nJ^MehEoP+#r8Lp|o!xv>spoC;#AWMKzwmo%Z<2
zlb#T#fv=`Xd39aI>6!_o`Fb#I+qil2tKp8~IJd=x@&#szDXCFfg?)ozbLLfSfHUES
zsLccc7Vh2Z@#5vnosq9zeKtrDED_$!s~!FN^})!<$e5QekCUx|eT)9eO@+c!`$%3Q
z#)~gMlUmM9DeQhq&qQ%&mX()R$94nA`Nml+d3_|J)osxQyY};?H~Tm_Pue$M^{lF{
z4sQDS^XI-phmy%lrOdYg%oeWl=8)9wDE6&y&2t*3NiJqdyXoskR#jD{UhR-xv2vx!
z)X;ZU{!6#N7}U1~54PrMnbyUB4fmDksmga-bUh_4&Gpw`g$<iGZHnycGef1(IIy|D
zAw{OUySp)2-|Obhn>J168OEO;>_BbfEf#T|%XFO^mZ2F2mZ|Lmm>WL^O6blUm+(Mf
z?FrNkRw1jfCp=nv0?S@~@M1Cj`tsCsvFq2o%}oQ3yZrpL-gcn5Au>TVPR%a@H4Hr~
zCO-ZQT7P}sWNX}`z4ARJdqY)pbwf~h+LPM~7UwO?o*gcZENg3949eG|xLbDAV<#4V
z$ak5292%N*vB;wYBlOo_QdBBcGs7ad&(0~&w}VETak48q7&N$yo_=hi$uiD<uobrt
z^{93kvsl>CHs=)SV=W0<M_bUQ$^!}t+#a=LtffrpUtW&Io%yM>RQcGk8|Y^JW^uf_
z=)5*V?Z;@34Rcy}iw+8y)a;{HxzjBZFmDvF?W#N`<Foy$u!XU4%$YrpZr-|ODsGu$
z*B?Y%y=KkT^y1=TZV8Fj>FY%vOv#!F-b2Ck6lW%a)2-ICvnTqniS~S8(X&Zs6S9&b
z>jZt2)OQRJ)h{_YIn|(yRea*xW!~byeSFl;o%0!W)uYZV%noj1VcClo31rw)cC@Xg
zuOTIP_Gf43egT2ha{8?lcaKX90=UN4JSTfq6&0i2x&)wGtVhT1cUT0Bt0GvN>e9BT
z0?W>fb%*<NtA9SYcGKp~CYUbSGiM%i>HtAD=GX%k_Hbiu|M+n!Z|a8^eii>_^=1mi
z`_B3cC;{&-y)gn<C7Y#MUNWIBL5o{hSYxRl2$6>|F^NS#s7cfjZE=}w>1jxb;@rP)
z-{}g;#a!L%otw&@3(Aq6Z=TWe>e#K7f*&%LrwVx<mHV*vRIp;4Pu2AR2ay%7#-z>+
zUDV03v%P-AGVO?YqtWNb`=1}u|Gah5*V{XRe?Ci7O3HnYe2Dht&MuSn>(|e8a4dGY
zc0GwmR?>X9BQH-=#7@Oz8&~j1*IwT>s~<S(?b+S7L#Ht1iPp>9ipWR)A?Lm(+jgCx
zyZ=#!sH|4Jv9S>{?0$|5vNOdyV{}AIYTS2-7xA)*@44eC$K>)wS>xH*_t~mc$<y*>
z4@G2-_Bpt(3i`&aA9<u-INSH(oYv8P;mYJCW)b9-p)SW_`*0wy;9@a<UzL`ntlVZ{
z2K5vTR)O=5R?ll@w!0l-UiZXfL7IW_(dP0;A<x=g-*;c7vPo3;Y)Sde5eZLHZuu^z
zG^9FBTDXkYs(&soSG#)k4HkO>ipY5uJ?A%AF}#Z&KHFa|NVFbkS-eo>JNi-9c-Tv6
zJRBRMd;Y9a>w4q2ZIK=a80<QeBBv{4H+3pqP$)`y`E|@yfq_|g){dv4o-yLyJL|TB
z&uTHUrdgUR(#>-f3N^@6RK6oiqc!*X^TUQdnC#&GW5<qh@$uQaPE1bL*Tg6by3E+z
zxpU|G&REO+yw}dCTdWQ&;~#OFvG$IOjfl65@E)XI&D$+0SzM|#U;AUrehICsW{0RR
zUMOAVUtz^~dE(KR)`i*FmAv0VhP$=%hCf(44t>mGbuKyojMY&pjGZh1qZfU4IZqB~
zyqq1#kp~c}ic89mm2jCsF|6v;$O4oXvFm%fFx$3hQsTGUq%A+spdut8Mmg%x&gL4{
znNP1c)(LN}iuEqrBR(8(c+ky6vCr`8#*e(_u2tVTA|A_ZU-3HhlzpnG{EHhmMn*>;
z6%-Vd#-*ocHwIqP^7`=Z@@wb6Yb|Q;R$YFc)nk`1eeKNx?NveV%Lx`Ik<G5(f8XJc
zjt)Vi%D-F|=O?q<Z!-2Y^Bz2?5G-P=1S%wO<<lRXG@9w#n=5)+^KwA7>YZ<V&mS$1
z%c(GXa*6LrV}||xKbqowlGNhDh0+_fn_`>H)nZ=0`)Zjc>%U29LYH<?(%8c?Zby%@
zROmj*eMOwlczgmr3`R}5`behtRFo%eE*Yy<Pid>rlW(bOu!zY@Y8ZSxq3LtAhsIib
zFz)TH$O4rDj-kCw*;7Z#bQaz7Ke@<Ly&O&ZRGGUI=?-t&UiewXh?Uq2sS2RMB2h74
z`Ww@#<26#oKJD@qIL1}eG!z&ya?N-qG~z~-Slg6kN~D|R1GW6!cS853ne6s|T+`fU
z5n#c-XWs5xh53r&Nd9oPL5{@~o0h^9+S5adj+fFqwAJGBbNl|Zofv&oYodxT?*P&M
zvuWn${hghi<>@2$0C)ZFubgKwIbtAJMlX_9E6A{RwZ+is>?rB}ZKe?t**fu?@4Kl7
z=9ksX*>hC&YW?bJv8)>R+!k8P#M1UL)kr{uz#Glz?0ap%k^YY#N4v~U7&aFWg0Y@)
zTveQF_DgkD?R;dRl#J7Ow0YV(n%wTJG>(q3S-ZBjvC;g}!#(O0^}~UY?vymBIJDqH
zcUrfJ&swN&dhEQnXkg1&deio}X|;k34L1|f{*<zwsl1ylj+#z9#jiu_?0fjv&Keh|
z*1l<zr-`~~$>gAR{$RaRaM3|V)$+~xEBw=?6@7eO4f3v=i=AloB8@x-lEw9!jczUd
zd-{dvA}6*_%la1hhb0Dv*4%758m-ftZCFSdSV9?cSM>Gwqrf($n?K^DF@JvZ<_%y&
zeP@8)K#4=NoN>%p<Du>WYTD%N{8-2N&mBsw%vICtl?qI^`yLeWzj(T|Ix_u5Rp)Tk
z>^8Pt@4g#8m7ZrbJS698K-G1Kn@!yjx-VBXfVJN)rzLz>|A3j(h?RdVSKO7u@yU&`
zSJKpD;*(FRIr*E}$pn_^?#{U=86_DW#9?n#op{vNv~8AkrkF`c5#^$N#X2^Th|El(
z7P}^M$GPER+ftqkp;>8f*2t+Tht8irqvZmgODYr<{kiwFacRWE@Q4Yvp4Ne$xX2@h
z4%@%HxK-sV6Y*0tqLkNuV<Cs@U-K3!;W8eyr)35}GhDA7KT0u9bZW=^c;Aqbvdylc
zPgQc$do>L=N4@!4gY&BT36(NVi!opL77{yl%WV9~@lr8BzHV^ljM<Fka)~=j)pY3n
z2Kjx0f~qAYC4Sqhnd8br#I^Dq$E)sdJxnBvn7<{yPqG9ham2-PQu(wFh-icc`|UYx
znYTlH^k{dCu&w3MYv1I#_XJ;>ciVo{TefP(O*p^x;Q4!fT@IncGa8KTf2C-Yu<fdu
znOjX&%H)t-cyayNS%YUhH%2mQbZab|GN=`KldtV)<^do2qM8yNqzX6=Dqq}Tm|wO`
zh-Hp(HN~Cf)0Zzb@9r@6_Hlq|y*l%(>i%8lY7S8u;kJVALG=Q9p}(XYfAV+px%nk8
zXieJL<#ed3w*_)X)+I!ifAuNm(`nJ0xsr1EN~W?(|16b(KX9DZRo?UM&1IJ=`m#Nf
z=bwZuR&b0Dx7u@3>x+Hr@{Wo{GFJ2se7`can^=H_j8oIo!B?+tPiZ*N8^g%`LO*N3
zF-A_k*xE%@rar%{Cr!!7sNW^WURdq%lZYLXJkF}~ZL=-oRl)X0O|-*e8klm}HTaKv
zGF-DUj?(MB@wQ^$z&lehHb#j;=ZUj`+)>_aqH10&e2FQNyLO!bIEqW*U^h5KJ!~NB
z>GnXPNon@Tuft*ko&B|v1sNgf_iTP`^U=Rc$9L(-v+>FY(~cug{55T8dp8ys9l5`>
z<_yY%V!FDF%v*Cc>|1OD%#H-6Tec+vHS8A`*D<Y6RJE{R^Q}5oS{WH9VDx${bkbxn
zsh_?5#;pj?#eL@I2fMTuIZ75E@Alo>x?g|Lf1>Z9<HsX9c42F@tAq6#Urh7U1Lm@<
zW@aWhTU9|JY@21SqFs*8@bECJq-(C-U~2-KsNKFjdp=h4Up?aX^p}I3ym{Y5sEmD_
z-jTqu&LhRE2P@k$Hy8K`Tubff6}3ESZUOD&0s)+S04*=Zd+Xy~zgAeaZo2~T`_5gv
zHf`P7-E9<IR$#KV+q_z>aq5L>(~n3)$+}l%X&G%|&BOGR8b*R7b@vDesJyV{d2-tz
z#f^3@Ci*l3<D|{w!=gn$j4AD=znhbvpI@_X!!sJ?XsCOklxp4hx0jpWt7%67PdSLJ
zq@=F?UHMn6SRvLJy+ZE#+|~Jh$wjs4!M%<mvMmca%4`*qq1ZjHV9pdN#_@8|K?f1~
z^7e-@%Hxemv)va0kHoU_%QCDTvxx}Z>XF*dct`xGUg@F3;=wB^nrlf>cHI%;ws5`0
zzRe}X=lfv-V9YsOqGNo;UuC$3a>kYMR&QnRY|_3rGQvrnx5!oGi`NJ*<<0q0be9s=
z2~tZ5jStkNP=03t3gyfil1~w%_rB(4K@l(}B_*YO?tvN(4w+B#re2Qy@FG5X^eS@Y
z+(OLO3;xR=KV5qB394dO#`yz|6Agyq4}kXu6^kEz1Cx{Ud;UBYoFy(cR@QSJTh%lz
zln>ftMiac>a9*A4t|wX+!!~EoTm*i<r(vketB45BXJ_`<_I*>F2xTw0eCSzCq#R4$
zc%7ExkGp%tg8rnqKk;6wITzA2I|l~`=bWLQ#Kp&d={y`Ky^LaIK!)An23bs6>1J15
z^MBmqPIZ~JqtQT*h}IYK_e%J}<mn3;#irVC^|Tcfj7qfOS&V*udZ?N2oJHJy;^RYf
zV?VbBYT-w&O7fn)i8<%c$d!P?Ig3{hAzmP8(JVx*I>n+BF*ar+FTEV9!xG`)yE;E!
z-&?D$=l}Ek92W{BM6hZI+~mPg%N>MvP$)dxmX_Mf++1yt32WmfkV8$h{YND~U#Q@6
z%)iIvYq(4;M70*@(K6ota2a2<*@=EZ@T9FC-(Mim;qcFND9ejS-oCvY?Kv|w0OY%f
zS+YiQaax`>W<;SVh;DvMkH@`YU;q_L4jMLZzpJY&>%j}|5;I-9m{<i)Z`gBYNZ~l9
zLPDCb1Nyz^f)wx-JkO$Sbpv~`<^Ekr(h|Q1H_#%RiVBdB;)dtS9{J22$iJRKY4zN{
zG)xUChH?0<<IvkxWPCtSHK0L%E-kHwh!wZ>utClF9jqXQOD&kurn~A?Q-)>zr-!@J
zpoInGhlTwG+HcZcc!OPb6~*0tX{O}Hx~paI<jQt->DR7ZgVXT}lJ@vOix%EG#Ag*n
z%8O79cNYF<X9yf50KVv8s%4u5o4BJ2AmfM)*qf%<(m2TF&W)7byMO;NCz#ld6CP-v
zy!Y<itEzl)LRC@m`*);hQq6upX>FW3_wL=N4jX)V!09sG8Q?gWH<=3g4`O<pc~kmx
zQTs&r<CMx|{uOv6zM&drdE?PvW%s<j{W<f7KW^)3&Pv&HE@&TBGx^G0!VDpRIvv`B
z*=OCgv`%64#KNd3ui#dTS5Q*&8?^=)NrR3ce-e(BDParkD=uBSqyhz(iZN8Sut>DX
z?%reQrl|OEbOBpYK{SK0JwVSjOEH}e>cP?LBUg5FavFgPvWnP5L%O!9RfTZQz`!72
z+oMAG`R_%omPG%}AVpXY8o`rV6I)}^aTMY9T#hvqihz&PV4gZH)Gc4yXM2og{<L(Q
zT7n-CJm>zMJC74f=gYIh;!ArmLD%cTidCzy#XnDH&D_)qz<3HA*aXsd5#pvzx6(Yg
z<AVk>H89E)ODy#4u_`(JBi4o+C=^4bBzlMLsz`V<OP)#|R+6pmm0de`+O*i#3+iGy
zwr||HQ6-!Q0{b#5S%2eP?!~?wgU$t)J}{hpdhkWbZw;+-V{-^Msh|R<h=NX`JXgt+
zc9)<0?#V&eyb4y0*%7UTrg*)A{9|vI{=(>t1xsOR0AGFP)ecf(K79O`ec&BN5T=h@
zrybo~K*8K8*i#oG4av{k)vwO7w-ZtN)+%O;dk4tRC<0qEWZbbT;~-Vt@$fiuiXGG8
zg>9&jP<Pm3!ykWyefxG^$EiOpdmy{lfft=6>gw!gq01lFwp^zL$HvE--4lawz8T$$
zsQamo<0fp9uG(bFl}j!<cmDb%(zM&S<D)n6A!a5AE7Djg?$w#%^tt+_oAcnZKwqzK
ze|yWe^*XI~X4|jba!&+grBKjyz-x)x08gqmR+T#-ABCgGzF;=kB5z!+<yt8Z+?eMg
zwy{K3^W<>-cH%Aw;q_)-cQ}4JABv0R{6<Db9tQ-}cfG))$7Na5deCJO6z61>P{@gf
zLTW}v^EEHHBZL=1JyOC_wPL_p^p#x5+<y(PJ3RF((6T&pK^m`l?C;+>*PLx@ik)**
z2757gW!z=V?~0l6-l(#`qxB61h867NeGS#q*P+w*K<%(z+5rO56kcxb>Nl5*S-BZ0
zl<1<l=}|$eb}8}EFNYiRoHEV@2@$2RzpX$osbC>HF}sIresFOiiPIP>i`4dkf<@N~
zkCl|LJHk=)LMVTrFRG_c|1DzE)oup8qZZV08_#yE*uTBJs%1D_BG87^Km~|-fz7yS
z!-nwp?@yi$7E#S;b&TWeh7FiUus1pta}?fJpR9}%p6g%(oy%yk3uJejyGq8mJYO+f
zdcTB(uCTB$5^8ecD_207CsQNhFcgL=_Bb#w4o&IQ$&+0$NYt=7f}8vU0=N$zth`0V
zggnE9ggX3KgL23PmO|`;AeR)w3biENJOL~s{VP`lbY(H!)I)(~gdlJH<Bz}G+}wzS
z2J?ltNnwtjl!quOLjwa7XsosqzFQAV<2t4a`>;~0Va?f;3DyfmN!a()P*F*rAG}C!
zim6jzw(zCQPEVH{ye~3aW{?uEW_s&5_J%t2#h$O>o*rwrzU(Xwh_j}Z$3{n2*6X<?
z)YPbac>msLabe!1J~22d97XX`mN>nbUEg^g-Q3Lbh5Pr{GHp8&e{5+*X`~n!mj?+G
z?}rFlNqUkxZu8?aMH~xZm_<#u`CdIi)4Bul#b~LgteuuSzzw|T<5LH3*S22AzQiEK
z82hd<%`^ZaZkk=cCicpaQ)=+#>tNHQS+?bi%yKxi(~;)GxcPulz5-h1RxHk;A*=qD
z9Cft2O7Eks`>6MPeFHdQ?41?0>r*&$=1jQ9a!e8wzE2fN1EuL7E-9i>9^QlQIh(0N
zkyAQ(@^)R4UeddF$BDq`@9&R3qH^WRmCH03<9;5@9EpIPy;%2&X^SZp64hz-x=mkV
zP`Y?&Din1_CMH!Qqv*9omDn&gU6ng*duqCad?$g7Os7XW34+@8XWYuT+E`-netUiF
z;>C*u3}f#P_8tAVZhGmi5WIT?1~dk3%<<<#83BVYTWtFqRX2N`lAIg5wJ@IK*8Nb1
z{p%_Y=l1TW=xs(8d^dMf`3mNetyAj{t|gX}PPUDrippb;b!@R%kr}=knwqh~*1tYk
zWQ58^9M;UVB5DyNZ~^-PZG9{ztfBG+`0et&^-1-)*GI=bb0>9meGUKo<%`)E_EJzB
zr~_%31W!T6f!_5Bz&P5lBILP*OExC07IXg3^yW}>^BmKFHp!x<R-u>0khk{Q%%g<D
z-I#7p3Q#uN&n#MqGW7PA)<Ls#A!0F;ZEkvZ@7{&+%CKh5S2p|R7XV>YzcPbrIljCI
z<LX!Yq4Q%=%bq(|*Has}O&sNeX*~-|RCBX-KV*A6Lp4muRNG!P7^+=$T9)HBD94&&
z*f5ggHOh^-*HbQCzPy`inLF$uVm}ancf(#;KO&hyCNn}aAD8%rmCpm!7~rw%4vVf;
zH-IQQ^>Y%fC@hp>;*ikZC4K)vL!V~Of59jqPz3}C-JpEoT(HRgUAw}i%b^qpi(Rjp
zZaBF1*_l_4ZlP#9k4vs7AsJ8;F2hXX8il=BnuE<*)txVM&=+^_-YqPdVcw)RJ2{v%
zISkN{cHu3(_~P6ssXKC^7q`e#sb33kEF+FL{0VH&^BC)@pCyk>%}i4oOtvoqj1lA_
zr-i=HDXdl_hV*8(ZSN}-YHp-9PMtc%D(3LoqF~+*^d_q>0~>J3Rp_sIhYs}-W;thE
zZBLy<+B`iAg(8I8D}Mi;IIED?)nG<kv$nRaK6_Mn*6ZHAW=*D<$N*h(`AyrmYlma+
zx_qvvsDFL2=*ppIsCmSCCk6PqK+MN|Ryyb0%x+)YQa%ZVlI(v{@X515jVJY=tRl-+
zv4HihjC{EYCp^|(b#2YOdihav3eX`Z<{9_0jN+c1r9+Vq5z_#~e=cC`N4T!LJ4MH~
zR@HAiOAlINX6xh*)E7CZums?m_E-FWvM}&u2T~ef6Um5rwE!RxLc_(&s|G9;G)r1S
z_DoM)a45KHbZl&_qe~L-ac{7F>#)so&^$#n0sv24^R|4QT{1q88_enmwx6DzmAfX!
zA;iTci#`=s|Av?gK&53GsUe#0d9^bYzJLD?i6UIrdH57f0&iXk7>!$rA85%j1u*c7
zh>t&zpn;Tu_{7sv;g{jzASk;6%YHe>f5q-fg5>lM7gOq=9947ELlHfG>eL<jWy>nA
zF9IU4W{jeTmp1FJXoPcthD%It4UK26vrTOxv*{Ksv9Q4IczV|Np1)k|Q~O-dG!+JX
zf@Yd5K(g|gGxyaJG$YAGArPRBH^wWtiVk82#j@x+9DB6qZ0xjm4L_U6|BVAL>-Vmx
zs5Zk=kG$iNz_a*|TRpKR?AWcT7j%?%MHuEp^HJ5-?yGafq9yg!wq$pJfZ`zlkZNJ^
zRp<TS$6ZNju<858Z>XiMt(`Ymnjm<;p)a}M!=<ly?KsRNqc2bSvj_8MiR8?*<zQh#
z$GK1m9jpulfD~k9?+xV7S`g&Y5DKFn8Yx^xlXth*pfF`GF3eU;&jRD|JJ0rK0CGGy
ze&U2JP*ykmomvzWlX%IC8BMlhU5YMK-#xQwAx^|1y?K+~1%J;jyAJI0=fp;%68;t;
zVc`bh+N_!UW`N4bsHonNQa;J3gZ%vb&n4aR<+QMr1oS^|Jp1fW=l$7_A3xSTUC`!%
zj|Y68_~x38proWEZcIex4v@tfxN75L)$*o2H8GB3Uu8%!Hq?WvPU1_bIBXKm-x^$L
z!S*KT=b%O-C$FF|3ZQ@fS8<Lye9#^=ziP>aVdnmCZ%&BX57ZExfX+#hK4fYH>$St7
z{nXE>Ah+M%4$S=@WW=A^=Vko}v_x+({&;_@nCBl9cj-D*Ay#=#VDE9-1XIqlXU|xV
zT4)E)U;cRiEBVz%ax?-J=$9Q`AB{%KrDfDDdtL^gkck1;0JOhsokL+4wTs`(r>kw~
zCN1p&Q@Sn;=m1{=M=}QKLWHefVI!nnA2k38(E(6y%y)tP*cgfKlAx2LhP?XOhafn(
zj5RK^9j0TR(<GVjY!JTPROnZJ5f&}k#^A>H?%p-7c|iUaMY~BkY&YRgq-A2}Q>|&|
zuvyQN9z8M85{Ks$w37ycX)Wa|Xx`Vfw6S>IANlzO#YuU3c=S}s@zp|_PW<%glQc7!
zF<(l9S$RFmrhb~HPUATjk_!6q{%(j5<RmZLnYNPB5=e`Vj^00lyA55wYE@}Pg**Ta
zo(e|`>|G`133VV|^gpkf5HAFZP(|A{`116{^eC;=Od?972KRE{39k;KXuMR?PPvnE
zm1wp{@h|{mI@TXSRY?z$oc{R;$lVvtzc7zh`Wa9q@IEQ;{zHcjvF`qhLdiFon;I^k
z1lO&+y=H^h)0Gg5Xm)f#{{Fjxd0w3Gl)ZN5F7N*R#$y+7NyEY1QJ<(?_}4fJU>2dm
zTRB{(UbH%op8Y4ymR$z^5Gai4Xf5iepHcxnMel(BTtEmE7Fz6^YjhNM<pX;8+Tc;j
z0lYdJw{6qv?U%v3ekqTkv}W7Jee~WIg*p{5VVamF9TO7+gewz%8+dM=+jOa3>cyg!
zYL3A9bj`Z&uG7kcv!N*2eAuLJ)RJx6m}MQ6zc6D;3J}_baz%)E8ZgeQPYgJ4LTqJ0
zhoK83BvKehz%X<hSjJiLXZBN(ZDJL^7J+pbFFD^WuleTEsqi}ldye-vp%k7X2QzFo
zqD(z$+Ep0pJaQK}xu>rH8Wd<^&&oUW6iOnQBhkTJWx&v-BQEvVoR&|i7tHvh40qDh
zXA@BnA3l7@v<~C!n*N_TV~JVfv{N5gqr+cp7{Rv~!55zFRWKmO-!byl?7PWSRH?T(
zquyd&A@05T4+`Zps$e4E+-~ai>(>D+My+<@4t<+e3?bCv54CR-x7xlPJB|ftU%Pw!
z_;I4Y@L=Rgj+FiS&-H|}@y?Ryr1xgXC3>U7#T-r|coN-;etIG3ECwkW$f7xR`m39a
z&Cbr6F3e2}Hj>OfzCwO3E$OO_!q~>*c;uLfh?B_aO)ibny{0%R%vi;z&!3e5AU2_n
zL1+_dx_tR^8q`1LJy@cmztxm~%oN+fRvo~bYZ`QpHPMRT0rk*PX>V@>AUj*g-@uTS
z4c^@wq5!qe@6n^()XiQW8jd<mn2q<;%0UlOK<$Xy36}HYw~$Tz1xN`!Oa20k`}+SN
zn9-s3ZuL+!ENC!;)MEbeB>E@^1D1GhAq<2xhv6S0vxg5I5-|Jr+ThG`iu<H-rFfy3
z2Nv!xe(PUnPRYoaqF-iX?$N4<m`RWc@+6u$b_sw|;%plz?wk9Aq}{Kt#Tb-vi-~DF
zO%A9cabb}=EZxv=2TE^>ok#exUEjACPoHl27xezc55?9&`X64iVedI3EOB7&K-a17
zs|Zzsk2CTwTs*o3<08gQ<nb3T0J%WIs<CacVxp`?)R5R1$fB|5O*EQ%2PvWI{^!&W
z?>z2%1{9j7=%5WmQj%`~M)wXudG6l+&l7|UqqmXlsvm-?glj@9-UpO}=A(+Ok>`sX
z9$^5dyf%Xg3Q}9|FtU8M$RN-Zj1-~YR{#<{eWB=!1pPvV;?YPx3E_xPmFJw`0ad~q
z{~6#6h%TV=Mrbj95#U5h09qj=rHYD>1`v%mOqd)7Vex<ZG}?RH(IeDWlB$`XY$Nd^
z$RAkeHpOiA`{0U@`05L0$vFl1VDP+^FI)&4$n03>0?Q8WzNH_eZdA#oF?`-e6kFiw
zUKELd|ELv4!($<qI}}4|p6^lDIR~Qj*lsSvl9xSuRuxx;n#7B~sA-pz_t7rLez2}>
zVIqUIA9!E=%9V&xn_vHm`|g*N)Pvj$@$cHis+B8c{g9|BeRI#t3mU-=S}D@3-=K!i
zSk1CuU@yYH8f`*Ba~$jXK9~DHF$w$G%qU(1jk1Gl$!I~H)l4^g22+R>FwBGT$j=hN
z^aOK;aRg3lvaAgQ(t+)$*X}Q1!6`y#v+rK4a9aq6JSfEY-30n@ZIH0Fc4Mae7UV;k
z+3A+np+EtarvH;CvP=3C<Pj*%B@KEiDnD~2GhaZciJF^phKg_o+?FsqS=lOdeG|47
ziC&?hy@L(C*>_@-la)%J@Kk><^0dWN)^&jxr4YrGR@&CqmH=+50zKjU%ryU6AUF}L
zc6Ss@I71CaEU-e_P4FCmb(Im+M*`S-E0!<sij7P3Bg~--YGPeNMr-3f8WCdubDl|j
zrYqc!^S=~n3N|MtUDtJ5UmO-hc3*O-Nm(^O7{r5$>FEw%Ni|sc;m)7|Z9mW-m-tjH
z6cpNJxLZ=y+2F^dta)!2Rzl5OoTEv?3*jpn_Ugve163p_mT&rhlZL{Cjy}BG=C+sz
zmpy25(BR4yBj6a4kJh<&3A*Y%pcpw>*;ME0Yan2PONu8W@~n|^0njtgpnA{rTe|i3
zHKDI2r=;K}TKCb2B9HT#glR)@uc}hY8+ykG-lc>TTaM>>zjyE6)kMj2kVq~9xbW-<
z^~XBr?4BC#a0C!0TF^Z`id4!~NI{?(i0stCTNZFzm=oc12D}~y4<*4j+^aHb(r(gT
zy&c!##!wo_K?8Zo`BY+%A-ew?!k+<zN3SOk9_Y?yvc*0QjoUB6vS3~f44T9dz2~v*
z_ur$VoPCGRNOE+c5-wxS))mIcvx5LzC2;d&fR-fJ_6Pp{k$7{O3aw}M8oG{uNJt_9
za{ZT4t!t;vT9wm`tEH7+o%#BBzjAq|o0*xJcXvfC^m&2nBR^TWW#p;|9?ocUZR*bj
zT)f>>l!0`#$hDt;6II4_eunfrJf<L)Y1r^~5u4L^#`0N9`QRIZ2+*O#_rhNpy>J{z
zT>`pv8zkmepU^tX+y6rwE6IBOnDal_^G|O;l-w%NwX*de5fyNBneQ%KVhz!G{s+-;
zsc(f^bGefwH^E9+XsRG_BDOsd?(P)m@y9U!rhWI3VHZQoqlp)!@dpz0UG{V}kX{E?
zG9rfT;*KOjvgEW)|J;D1tyH8y>(%vKBC-3C={fpa$ou!Lk%+DTsE34I=`@1%JI;S+
zcB@2f7r;8<IsZEB>v4KYd+&e2!A^|2j28Y9YCsYeACdA?7H_O#S3nOWu*2vi7&H;m
zk)B-gr~Ni^aPU!kaOB~iKFYiY3S}oqMPhXHgHXMe9Hc#80f|&D-??*Vs_VQXl&>CW
zE!Z582_xV_i?iXxN)p(PnDdlXf?k0x!jsks^Pw&i&l1r&SFS<}m&6wHL!Jj11AK84
zD{D{pX>@6}43@Xx&`)sQ<Hh608;Kl%8U`JqE=1fZ6(SGu8t?~d%2zO}hA;z-pekbL
zjr<Eqe*OA&2F+FQ()-oo^JCS#F6a&<DqG+>&&u-`h4KI}pOE&lr~D@{D*+A8=j0Co
zmPTQS<dn3vgMc^E><4v8>I6w#(DWC8qml0J?jOq$v#4+yDcM_#y(~W2Y)v}EN@jNL
zUcsZgb`>q_hEBN-1`Y>Kr_7^BSYNk<y9*6#9Ks=W8DQXksPk%Zs%Z26KcD`oqonlX
zy9ypVz;RO{w(89Mpj#5koA_*B%3@n<p64lzy>lFkYm1Hzxkto%#XP;?nZB>6NB!Qf
zC+QWhu6Jpymw)=4+Ms6f+_S9Z%eDT9s$Z8cIs7f9^SLNoS}P_;v4VZ0oR`3+ipbqr
zf2>+%km=_3VArFtzm`qjA9dBK*Dh;rbNaP#CSZL0@FcZDZ1l@+<~-6!<)k3FW7U@Q
zVNpBeX>MwVpaTiHY+8hq^QYe+|JsfbWnec}EM;!=Uo@(jh6>kLug(TOXIq20SBRs}
z#bjZ}#zjY;Mk!B%Ubru=wz0AS#eEY|qK_ZHxo*2qdvY#n5}JNzDkp`)`$BESx3)Gt
z05U&L_|q>xIuR+$;|+`x;x~{|cI{enV}@lCG+e~Lj4}Bz9=y;uW!9{uwD0V3-qeNF
zRy~~yDel9E4<;?yX<8?Uu8+F(S6g)-9&G7z0TiKO!usn@7TdE&ny@&i93ieV#?XpB
zMgSo@<d<gOy#Ol`P~pXc2Y*1~5e=}1<_xRDf^ijv!sUW>08E{9)fu=Gr?OmjEotaA
zS(H$rtH&W)5H%mYTRU;3IAj&LZ$L;D2|M6D+k#g?MAGo?Sm${HVyO30@ldAGi$Q!Q
zUI_9nxF@@;Ep<eBhHh(I7q4N#w1(oI<b;sG7)*X38@+=DTRcKsCN#vyDo^a;1parb
zncm+1<;$7fd-v80Ntk^*h&o@(4UT0s4migRn;uDSMQ7(+*I_R`5)$`h<{-*3Jeeo}
z+#aZu*RHoAY;LTCRwyPP%=o&XVDZ%8a4(If-k($r_%L!7Z3P4yEvFX8-gvh#uU=gV
z8Hw-wIYi6bb457l#j))Hp=yLe9zJ|Qw`^t2(j7pfibp(IL=sf-x^3$+F)$4hB^MRd
zLKHoq=(_^4%0woB3PB<{u+&Wcjt4EGP_)xUSlSVeCl(m=@SgglP|p2Fj;Q%<L<Ake
zS{$S|GCKH@a;eS2HilFLKQ0m~6B0J0XW#q#@Ae76Vto19YOI*rjx%S_uhDkTz=0U}
zyqx^Ty{UG3=Xt-!!NCciKAldpXwigg@Rc;J_Vx?J4+Oxrx0fb=O6f@`U*S-nr~{-~
z4agJ=Tw@AuJOj-s4S<_U5}Y;=y)srU$ueOQJ<hAgNLmbC6T2y(^JuF>IEbb{#%OPd
z!*O)w^d=N__fb#^GFD(%f#l;Jl9XuTg{UYGq2s7^IAHS9%^H-FAn3Fohf+h#DggA2
z?Cb;Dt!bv}C|ggTu0=0WpwTzyyEq~BT`x3XB{X#keDtX3)gQOvwQpo&OCV>xDhB5^
zO8oc5`F?5PW2|BPiNhZ_w6^0WKho=vo1smY@O;PkM#8Bed5^7Iw-SXsUM<er3^`xo
z?vmcLVFUWow2AGyp=AA1kcAyou6_F|{dUWxnl~L0A1Yi1txX9!=sI>WsO19WU=Xx=
z<H|~O)aU=x$`!n{69iH|fnn$}0zdeG30kkaI|bXmyE`g0|3VA?4`JG{^lr{7;j$fb
z^Yd}yGd<hUnH%~iq?d?hOzn<J6aqltxN7vB%nF2B_#AJO#<Ozmmdd~H64Tu_pXpBu
z#mJdhkLY25b%J0e!rRwjG>zbEkjx}8=0GV@&`#UG?khxzxbeSOl6%lO3r7{vWrzMm
zHC0qodxGjpgjf(AqLX3k8^~bWax0L(p@)V2q{Tx0UoFwZjuX#+E_iSm=95(q32ykJ
z3DAe$_QBkA;9)ojUEu;sVD@*9ZDjCa>OEycZAwx}IuR!KuVs|O%?w-lBP5O}$l@fl
zQws~x7Jv{&?9k0|Xjf-WlORfFym;ZiUs9RGK8_}Ku5#45Zm3^h`!m`SiE{%F?Itc6
ztr0x=-wNMMnWbxZ+v36`fiJ<AVNocWnZ=VChv$<eh<ndBtA2IXKn9hd7u&TWSX3R|
zj~o!=EHOzFDaKIK#3b*Yc#S#PLe#_j#RVr2m3^T7=rsm1Fe|xhrvK<73x<=kW~tPf
zHzY^uzQ9}F!3l|1fD5r8ZNwW$%%3qKF;ISf#VGmEBcg~fuqTx}xzEqf>u-_v?jUMm
zQ~x-TPqop|h~-OMbu!ivQjq=XhRvxAkpyw^XqXexIB8)Z1Nd>d_Bu1kNcsEuDRnsf
zUQh~TV-^zThcp?V2~vtgcOipCWGv!BNX+&fCJmk>;~V%f??(R{#c(P<&9uIU<g$Du
z!NqziU)%zduW`V04;`}QEkrGAEO0CMbHfHAcvF=y?wtFfD}X|)0lBuYVwvk1rZoH<
z)EiATPt)`d4Q=~p3-A{D)x0{p4Ye{REKG{f4A>licpKNR|2_Z)B2@*DhlLx7<V;56
zu-n3DXb)=dt|&}Kj<y<|#NOYU$-rme9Fv2F<=bKXQn6fW;d9yar`3NpN$crNDySnF
zDbj2#-DU%S4?uvbI|!Kz;CK&!lA!K-5$smox1>=(+#?<sIXGArD%qxg08I<}3ORs#
zb}2j+ZB3^O6vpm6bPO$)%sJCTbOM!K=<8@9Rmh~-;v^1SKzQc5N1Lr~!P_ZAp31#A
z&YI4Z9E%0IAl^LFL#*RDqsCNYa^gihMC-t2Vy+{<bVQhlhf(E|bmg$^pqbbIp+RsA
z7A=9!C~3qUChp&n!2XxmYziqNgLL^&v#?`I*)L%Co(J9m82IJP_UE71N+xS;%n{?U
zdgdur{NY14*^vnG#D|52nbdpTzaMWe1eLtu<_boD@g&fe7L0jg=GAai+Xfst9MZ&8
zlN=*lnuOc8Z+j&c^Pzy&_(By1R^3g-;jNS2lN2afX#`b3lqG3Ds5)v;*>HxNbtg>A
z=pVSIjRYSLkLvHkA?Cm9Z7|~yf>DWmeSNIVE_e&3T}O|q6Z@e&Saf%=WMeF(4N?&C
zYBK{dzy#gkB>EPd8cIBbDB-UEVv9kBuJ@}9hbm;&6F=!Fl-j3AnkgwOlYWS!yE)ox
zP5)C{AuSXK#c=fQ6_kQ;1m++UT2F792|_b1;-Ld1eT=j3IPk_=?8&qZNy~_Tnt(N3
z>u=cy{pi2}72vw2B~*wI2}@ql&Ct+LoC5n7fOCVK#q<!}Li_#i?fp`Q{dz>>FbQ~U
zIT-d3*tcD$M;k?3sv}U`DIJ{0AD96Oap`xfeK_Dtv`I(DY;tHEQ2!a!ONy^r9K9IA
zg3u}<rGK8DQ-XS=I%owRs)RFD6N7C@uz-Og<o%%Eb5pUXs!**-GV-6Es_H)3-$Xl*
zry2@O{1FB8-|p3mb^>S`#}gQ=YDZm#S3yj%#wAhzV!<hKwD+(m7M^Z6)-u^PT6GwD
z+eiOTfamD%f(c~)-Aaf9jwL>u$X-Kj{OYeSTZy!jw$5sx`3DRIQ;5uj5PkmqnRk}4
zHr4wdsT{}bpk~-@MY=;P!=m;Cvv|b%UD8HS7_JFzw<sf+tq*E1xPCS~<1{cza+omG
zbtVQSn1>1j?~xC?xXCr-fW)*C*4<dQZk^io>sh0=Qc{Jd>Id5j+!B@@fECQ6nOZs0
zS=wnOMDmLu$x`y-<0=V~9ZFEQYm@cmU`&u$4T3f%RS|bbT|aU-58WV#CduDoB{(9@
zK|VVHM-DRdF)tFteE=Foi^)7t0PAZLb|cn+`0)D;M6)?73qDATQ>D{I1OyzElS+08
zK=DtS`h0l@NrT9TN~ASrBCvsWXivN~Tn4^WJ($1X#dpht#huiM>)%-)t$~x!Bz}&T
zrjej_=k{%*zHe`kmh}e#+2L9A0?wiiskzYof6V#FOH6aF*g|RNC+Sw8<*bjxxNC^F
ziR(9Dak`kOy5oppw{t#%Z16<0^*i7jAw+U@e>M(c`Eh!E*hwW$Hdqkxu#re*+AgF3
zM9{Wkg)`iR#@82bfzAC&o<mRq*jWKWueAn|k=hICfQxW8&mOrba_T3j1G+{UFa+->
zj7eK|&}Gbk$=nWLWP5U`3IMtG<tZ<dVzzaj(5G5H#h`%krg0Uj5{7BM%@SJPXmg~s
z#N5z{@xeAd+#h6jauyKxk#?<97MCM}EJQI9WGJ0oTNFK!+k(atgT2LD`uvDFSIjl$
zF?btMzg1zPk`vTGTx-<5A2-D-ksheV(YLPTJ|)jvaM{*jQ9Aa-M5Q_mU!wWGr!9So
z*e{)fN71{~QCyiKA+-VGVpgjVG(JETah`e&!CUY`AqDyeW{CD78L`*%y?IFs3yFHw
z2iMZ9aefs(X-|5SMKw+Z?<qRC#M}_0(R_X3MycK*QvZ`DMR)DG<)u$YVf_Mx92E5z
zF*PBE6GMU>c`m5`<N)w0W_A*@$)Ebch=+d~gneA`YyFBTaI>#?-B?rGg4)p*n@85u
z_^b(AY(*wEk(#VN?dxd`X5_%_=mU}l|DFCl?FXroIOia-y)ZI(o4E0rVF-TV<S1-%
zV4f&A93rI_;1)1T*&utyTp#^aK8d&+Ih`;)gPg`g?E29H2s>Qla4C!S?L&Ng^3WW|
zR`jlfgiMrU7@n+$0oipwjL`?7i~}9?!XQ`1WlGPZ#*v&8tT>?n!pROBmaE_w#E7<N
z5uqxfBp)&R7KZS2rn59KIN;(w&!Qc1wOmvYWOG1T5R|D&H`k*2f0&8#e0<g6Fs1Me
zkUfEs?e{MAYziVMk$h%|^{+2kVF+^K6KA0ILdM#G#09Kka<~}wF-yEw_4i7+w<`EB
z5t6LKD1j*Q=wz#eOMCmNXe2v<^Mf7<59;UQ*Tf{Z$YK!0f~FK6eww{%3aY0HS|$jY
z-P`l1^B6?`pdj-3AiyYk$R$@l<-fd<g+)E8|4WdtGGGEgpgOEL0ql9(?y3*M$XD4A
ze}i1G)}%Jcb)5Tbh*(x@8^8k|NB9g+A;}NGx8=(|hn@|g3(YLTE3qE<k8z!w#9o{e
zB<Ydio~I}&tb!(wt&<Kez4mj_Su+&%rBn9c4cKgJHeMj#-Q~lG!K@E{X*)A!oN^f#
zvRxuFcaax^M0JArD9p_Eyn~HtaY%2TIRod!e(6OkNC*|$B{AwCUW|;2O5R~cstzGi
zI0c9KvD-r`YMTB+2X1`!yO9ebq({ZW+4{f(k|)5+pzL`h%4Z5Z#rENEAT>*8mQ=uA
zy*mJutI^t7f5+^X4$chWqX92&5%eJ`O^b}9OW=eVMUTRF@P85*1b+XHMjmekNmPKn
z!#*d5AX?1ll9I0oVGy~RxKYT!8x(_>g)b2^`u_!DK6{4L_;B*hyMYYD5ho%>3F<c~
zpYRlckiYm~oYa30?r$;E#112$yKd=bhhcwlA3m&%41B`yy-Fhc1R=vQ_*ONHkUls$
z%7~z(Au9ol3Vf}KFbzfUGBajVw;i#`dQchSBmNcxJ`gmSE9L5%pO`ZsvNVUazsYFM
zK(-vnn?W&FY24M-KmkiHSBIjh>oTrNj#q$v#@gkPbj3N`@#OEB&7c0YEm*=u6G1T+
zoupTBCxCmFnB_%0FM#B29FvX*BP+f1=IN5ScLF+3A)_;pGunFMaHS!Dtg-fTh%0=R
zK!JFQ|9D&f9p_~=C=%$poqTu!%Qu__LD~{$=U!khMB~ua$Y{=xeIEgGw{5^~SN&$U
zkg18n!5Q=V^#dn*<gd5>ep*DZzb&bg;^nv<d`&H%Ad~b0TV^k`9dI<^P`%GcKeF!F
z3h;iHEGbVYhoLEBYA%34yh}@oS#>;DkGw<Jf)K$<e6Nuh%!s$<CkBbQqfMSfIQev<
z{{hKZJnl#6cmugV_Rs$w69uHPq2!!D9}Uw8f0D`L2H6t5gdkUhKQ)l*X#}vlWywNu
ze{MzQik%A|8bc1W)`Q@(=oH}NBLdQbNgxN|HNP4hsV$wk9}4VR1FaJvWk1p+kR{2H
zIAZa)R%Gr-XQUK<SW*%c;6{?L8K*IFUC3SBnVXB1bcn-wR%BNy^vw-%42VQC2<ViY
z`_5dOsEp$#B#?^_h(Z4)AIJd4mCri59iLqUm4adIS{q)#{$NP9e=SvI&A5J95=6!H
z1}8BUJ@{}bD}IoHGGtC_jaUs(LvTes4r9H4|9)kQ?pgi1gKLR!2v>w?4?qayBVRB&
zKw{L|d>XnE;j*R403EWi2U1aq0-6VeNswr%ZO5Cu!5pRkGe^eNd#cl2riLz}b(t2r
z)4_ki9oB$5rXd1CWAEvin$1}#x7hMg-&_(p3-sh`5)g;@vUNaB-;ph3rMOo<q~_US
zLjjvE1zNJTejRu|2<~&hd22J2ICAg+ELB?~3Z@jy*tc)r+BUjE3#!erx1{wV-JBh?
zjR~KqLCQ$sDYz6dFpJ0`80!W^`IoHKRjYbEwpmmX5x8~3)~zwnbVVvA{UadsaP8US
zF=|godG-AdBXT!zEl3dS&1KKEMG92vU1B;TLRB{vf~pt`K0%TY_!T@^;$I+lFL>!1
z2+A9>YPBE3Nx=#u@nyFKXCi6>i7IMqH=k-d3Xhja%1G&hcT}NA<19#kv@8>Z60((~
z{0KMa5*N>Fe1lyMWB_PHO<Tx?I=U1{<vvw!kOSdtgQpZK19&MtDQoB*8_3J?g%6tj
zX9tokf$FoDEJA$YV_yDE1Hv!<|M(Vf+xjR&XBPL(9ZPa^sp{LpK1l2AoRe9O;c!^9
zFyLd-=^s*i?@`-xwxSO4OZNM|^!-7`(Buw%U*v+Red4-c@umx*dPepF-weq-xT6s2
zat-!rxTq?b*xh$D{%+|XdcdGKnnlvfo(3g@;=E?XmRAkyox)$<x>e?*$d@!Hc-rt@
z**P*dVMj8_hkg7XG+8$b8tB@kr=ubhIAp!_{QoSbLJPSYvFBKocS~fs&X;pQ?v%n}
zMfaqQT1TkHaS^(Ckzsq3jG_%Uw)<T&{$2|2v#MsRpQPJQ^4w2v9l5sq|Elf0|C+kn
zJ|3u)3W|3X5k;Y`2o(yrKm~z{f&xW$!lsC98H$1sNKjPjV#yF>h!9pJK_Ie;iZyIx
z#>fbYVJ`^*GDz<0^nN|#hv(PlA2`WL&N<)j^%?K8elcNX)xCh8ZrbU5ew(F^UJ`MA
zli!>6-cnX?jb!Hq<ad634i33#b~H)ll$|3^Kk08vv&`QP#HYNEPZylv?A?kOT;0NN
zCp98A@OBP%&z8z{me;l&(n|xa255C<wfIOAx=ZLI#ud2DiM<uSl(pb4EY1m(b~2={
z+4t!zS7#cEc#bb-M)RswRw7ISosWokx>8H2V;`-%4QV<j9Q+wJYU>HIVW|wi5DB<H
ztrzupYOgqCPI}yO|1Tdyg$=2UZvUat>VazdNE<JtvMy}?AL)4b%jJlBiV>CMpadRe
zG-ob@y%@}_bJp6%e9?mLpf)`ek2-K@OY4jT*1inkM$F09yw$i7kwgk+V8*cqo?$7{
zn|Q~2^yvw9l`0zA5xwKQKe1_yBsh@=6p>TGQ@z!J*yvMOq3aTzMmcnoOHYqD9Uva_
ziO=Lb)n2+=mmMHnBkTBn8=O`<$j`fJQ(L!xhK>NMJG}!^=EYDSJ{TC3JI2;wS1E1_
zfAJ_X_;lh`Nj6z4AH~0QlbXyMIT*<QaGfj7R!h(2HZ5}sHusK17t}|ZQy&tehDVBc
zYdPGes}WQ^0yQvTiHzzW#7hbD%M+J$8&&5Y?^>BqW$&ccJQy8mG}~#?tXT4Al5yV8
z9<UJmf0zfnsKfoP!DwONWjV+G2Gb{IZBOvgyW@W6i`O?Uf9s8!j&}og8tW=8p}qT{
zTJl~Q`dSXw)uTp}*)%V}Vtmd=^K||QJ67{uhE1jVRLeE&OXUZI3yY2?elb_C$nBfK
zYktUjbT1M+dC{U))saqVq<9#aRrDX;537B*^n}iYyH!jyJnB~_%>3TIlcsGsqkFxA
zx5nj~_VkcuNf*+cZkrvanGIE=7k-@hf_9sZ%TLbln3G-bJfYdxWqD_l4!xNwc)U>W
zcPrU&J==iipZ~sSv=AG-msrrm?k{C@k0{B)75u1cMf@^Ul40+GlulsqL6&FeeUeQ^
zOgt|w{gA(%P?g#{_Os1w?%wCiBjF@$Eb+(|zpKfz+ike*In6&uua)&T)K>X~vvkL|
z6TCasXx$qXXil!&gB8wZzcx=OPF{NHYt{EJ?Uzp3)OU`*rAaG#Ymn!$tJyp>E6j8t
z%FEFO|HF07KI6e~k_)SmvXQ0MuK7x6NsGtxeI(E>2OgA0RV!htird14trexS#I_^m
zsaE!MiXGP|Ef;qoHBfMkZ-3N&1U<kZ;=Es$St%_t3&vs!2XJx5i<UifADI%}e5?V1
zIlNpPwvYSt{^dKL1PrO@C-vwvo=WImtz0_3artPrhWcPVAj&&$8EHpOKhlljnWuOA
zZvSYfuPd~n)=%0~f<|9`nm?c1XxA2DDz&vFD!MP2@cP3Zj3n%;A7RJk<_$TH6i>Xo
zpm3o{(=9f6XqNxnmdB(~(d~(YUEgq8cWMKL5`SpDhJ+kYArYz;L>iIpfVN-nS^dZr
z->Sf4Cf|n#e3>j#F&wp|tF+dv&!n2Dnk@>9Bc~;%Ewu)5o_DPyYO3aG8mVbH?ry8&
z#+#(13YM4I?Ot}mzC&Gc>gtc1xbHK<sA}dJ9`d_ReYhbkl<+<mQg297y6(8>oV1aJ
zzGOIAF$WKTUeUK?lFY1fT)xZd{i~Bt^4~5jJC&Z2^yiO;SIxFCtETS-e_h0|H@2y_
z5$-FkHL+=msE`Adn4>;9o`M_mk7g$9O@ngrpNC*5g>4QJ7Dq^+6elEt3R@kBAlnD2
z=ca#xi^=zgTux0Zx54xAPy0$bpNfA+C^=f-laqFp8nRD%8+rJsbD!Iop;i)U+AcyN
zAIrxq%#W-Gv2yAi`v{g51b7l-s)<+F82!uw{Y5X>QGF7rVO$TBN)A1JDR`8y09@vm
zY}Vf)8b`XA;@3W}*B;L1(=2J4;Tz_x_14KPO@0}f-%}26FukJFl$b#+ZvHKac@Oao
z&+}fIN-b?me%R*K#qpY&-{<_yKM3ZWt;jn9$OoxY1tuc~78c@0r4L2xj3ZUNT<Pnt
z8XV*tFSS?|EzJ(pvwzy)#+H<JVX5umql@KL4G#I~@2B2noZ8pPEV}TSj^Tr?6xCC2
z+Z4*EcUJ14*6b~m8w0oKSJ;07#I(x`q-vy&_%-CjC5n^%|Fz#X;Bsr-UPu0q{Dv+H
z>{n?%2{r~IkFc+U>e*g{nF=lHpLX(FIBFBu!fw})E9vYe?4YSH(y<F)Flb`Rq0$P0
zS#MKgFQYQGn=iWYnHWmok{x@$+cCVSdd7p6Q#^1J#ohTd`#BC;cf|6-><-u+)GITD
zwp_u;2vHV4H102FxAE=WP*apM*H~4X7hjb1c;X~0{>TQ4NpADjN>biv@G4MTq*`6c
zUev?;hHDE1T_XHzAPL<ZR7ox%X1*$+Rl}IyHtutUY2o_>e+7{=z!FE=wJvE;@3%{a
z+d>waI?xv<LL&+RqVAQ6gLV_rsYdK(R`gwumT!IP^7M-e>3g<;c&RC(Q-uUzpXpzN
zn8-?g%k2A$N#31l?^Vs1QS-WTdtJEQZdZEW#M+#_eJrG0md+RGD$sNB!oJ48Gw@08
z;~x*N8|qkwI~Sd^b4m+Fc1_#my~A!t^0|S@8#JbuAdSADFXFUyyQp|H7h9cekJrUb
zJY%0rr|&RC-_YpVUpGlWFITRYybURS3u&moEiHvJ+LVTt`JV&=!e4p?<#MTHd{$*+
zCdEf-w~slp?Me&9v`UWkJi+V7>P@(tello0i>B6zvNO5W+u6Y*Pew;^PCpTIL=pdM
zo^kjMMjDtutB@8hi~(-tyVkG1vVQt46zX~iI%{$k2?V?K{oI`MqL0$+8g8zHv)Wic
zhbK2Ivs;4;Tu-L9^F|Nob5mB_&@6|+IGf*7L)8_de%CE`hYD4RIsL9kM)A2%zij3H
z-)+vO1`Ij|a7Ep1t-YLcl{<Wvj}`M{d3=n*wZe6035V}op`hLMN;K_bFTXSWw=SGm
z??7|`y39YFuHe8Q_6pmCGPuR1AFT-xHp)&gA9xDO6l|qK0ILCpFRS9ke@{>da|24T
z`H0*|>rS-=#T2q1dfOt9Y?+7ro+Fq3QbtO~5Z0;vs-vmx299<LyfGq+><+mv|219g
z9a`ae=V~((?c*_J#l=x}tdR5dVQ7+wh;?lzD{obFipc$Jp%7h)YR1~@OCIFp`%uQ0
z;v%1mjEt*mO+^MPH@nVPFuFyV{w-FF1jA5g4N{RomY!s5VqSNGC|89vRpO)buqUoh
zN?{&^n6a&zVKO0!cowrU1K*Pf*Dx58NXf9h_7y^3hHM}Nq>*$dD3cJue#gy$EDB;V
zNP$ZM0wkn3RR1A3f>;RU&xg8!<2K`%+u=k;-%!JyD$XLQ-NygQT-&R;nRKRJ`LfuJ
zd!=#i2UfH4B){MmPjw$JjJAy4$UVcGot)kiO|wld(5D$#-?7e1d_5s0b9sX2XqXdw
zDan{x!xNb%o3-H*^WWke<F1LJLI`yqPrmM;`fe!{wu0M2Kp=sj9wKl@L`Q(WHV4{B
z=aWH#^@7m7Crt)1P{G6m#?Iiodpoq&Uu4R}HTA>)u&_0Lj?t`d7o%Icu(zNd|0CIO
zOQh}QY76VZoMRfJx~{uLQN|XrPE+e^lFFu!c?Osf+R551j7z;j3fK`xpWr=o`p9Tx
z6$JI{SC~aX->3g)7Ttx2gvj(Ng!Eg$$8ry(;oDEe<3Zc44OF1)UXv1W6BwH{Tg!(U
z7oU67T(p;$VW_vcD#qV$3+3sF8tO+QeG_nud8)m)A<<Ui+;!Wy6OHmzYicd4#~ioX
zeMX&JjtlCAAvG2GA`h(lcIJkdB2olsDfUpsQ?cl%zdf92olj{B#gwdN7kbqc0PnzJ
z54rDL!XWR9*K5{P<UEJR-cX|7P*Ao!tU0d|T{V7&QkP76D|T$VWd>d>izbi`KE0z>
zyhAu5dx!K#>IJ+{Sjr4z-hXvQTz_?`ImP{0O>~5Nan*~KIr;Q=!3n4s_04Rp1_@fe
zn;7!n5ZIHmc{J7bE3=|L2GOZ?DP}f&ckW_RZgXCuAZlZRE+P~WhmE9se?5KLA`A0i
z#qREw4J)T-Agj7Yf?w5uLXIY}E_Ha0Dso<1nJq%_qf8vXjB!QTG}lq1Zl;v@j{-f@
z-#T|<>BqjJ&JSq`X|r$KLObJa^6`_=&jRvE_E!5xSN1{b?j3SMVk82n6(FOb{XbPr
zAE0kQv~mDCt|sSMYT)-nhZ3J3ylY9B`jmHjVck5-NH}10rrz-RSd6*o_>hCzmbi*F
z5!sDD75-^uG_rCW<@QEl!X4E?T~d@Cb6THu7nK6X(GN^M$U}4)(>noG2#O^VY@N)h
zUuS(1t8oYYo6{o-q%ZmiPJ1#NP6Ix9|H5?i^2%n@-e*`?_neO-I2ILE@l#4{tug&4
z_|F<OQr9$qtWgYM4&-7-8rKjDF{@irO96YMj+N$|0;C#HY^D*WhW2rg2p@8Hp~iw-
zlB?xF#N|K;dKwU!;4V2mnk{p9X#~MPB+rr&C59?ig;R*Be<@j$21`~@ElUoms{rE#
z1Mz7W+)cna*uuRJa_b7z*=8~WNQd=~fF7HGe8Pn}F%r@RgQS9KGD1=Se=GeP0-zWQ
zrXrx%*00j2n5TzC;;7+!;x3Zliy{${qFh`l5C|rh0eOrNH2uLgD-0z&qfo>5k;7`r
zj_D$~h5iE59S9{8YSzg}+Z>{OvzB<?j^l$!!Rw!+zTwfJvzXu=x1<6c7f2=2LF-fh
zk6;P-PBR4Y0I8C%Kc-_PQDbJ!0%=r;q~@*byAdl7B4*i;Z=4iz5bg;A;`F`>=SFhr
zphC^!%fF`DKUDeIB3X@-+<5uJw^~8s{Qf;0qF8Gc<qoD8xZ40?<e4EsBar-%akA#i
zVL02FSaFwa;L*YhZLC4l>;M!~O!uEc36zWGe+f$=GXIYXlSt`i`0m?-2KOQmfxy)G
L!da5hpSS)E(@X|K

diff --git a/docs/articles/vignette-06-locations.html b/docs/articles/vignette-06-locations.html
index b7b284e7a..aad253659 100644
--- a/docs/articles/vignette-06-locations.html
+++ b/docs/articles/vignette-06-locations.html
@@ -33,7 +33,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/articles/vignette-06-locations_files/figure-html/demographic_model-1.png b/docs/articles/vignette-06-locations_files/figure-html/demographic_model-1.png
index 45275daf3778f82a417d4bc2a92a42626734107b..95bb9d905efb79239898d9b029ed4395b88e8553 100644
GIT binary patch
literal 19842
zcmeIacT`i|zAhX<KtaR;@`@A-pn!l%?<i6N(wlUp7b((vKm`;8G?Xa4w*ZmeiGqND
zv_Ob}H0gv6p$EPh-@W%a`<#2nxcB?MG44N?p<|6CYt1$1Tyy@)^E|)xR#RPxj)s{A
z0)fy$l^<zCAmn`z2-)OmD)7y_R-+<tpt4X^dIXN(Z(qJ11`cOjl?~k?kc%Fq|77Df
z>^>04H3;<416`lw<uP9$T`aMEqxXkZ$|*SH0>$;W%bBO{1{-N<k}+v!=wEwDMoC*U
znB&QPyV$t2()kHY;Y)DDnH#4wuhnEK6eqAMB*;BEd(o;Z6>76jym$Wo@OL$?PiGp;
zdj&OCf4R@~>>k_>5)<ELiS3$oI|F(nSNt*1l+?Svxbx{`;7D;Ai~$l%cFGPMzCg%@
zz~L!`j2Rs621Dq;;V-#h(g4q1lf(6E$P%-6Rz~rq?QC0~+~8ZgN2eAfc(}Q>Y;1DR
zvCAyu!K^^~-u!PgN&zmQI<;2^4v?DR=OTQjb-G!KQQ9%9W1ihugTp$Gl+ltD=i{+8
z1}3I?y!H_lsdHcGm&Y58MbEJnHSyi>a5_BYNlQ1lf5kY^I8Jq>d<s^#l-6tC8LP1r
zea8EsicR(bX$<T4@&*bnh>qvy`FRcyRxSmYs&iiQ_ZYy!h)BQfI%Oyn+Qi>J|B(U=
zd(v_OcG76nv51&5gd4qnZRen!Y14SW1(&l&Nt&eK7;7u5h35Rz<_N8<)&AnRQp<2u
z*E;Q=d1QC5H7Wb}xQRbY@p3s-J#PSK{r6R1YUwYKLm)yI!f9x*Qxoo-P?uk>Q9+|Q
zSxWxluwHD&%Vr+g`uSDK!TUnGfpfWzqaU)^oSub*FrQ~fGEwgsJob6g!T_VXE=LiZ
z*45QDP`}1MxVLf1V@22if3zhDri*^Q?VJ7HdJ?uk=axBwJbB}?ORwhQTYCn9Aas*<
zaFt|=^$ND2u=hohnyU9Q#^A8LVs~|eMo18v!3BE1Ziyc%9_U6lszaF=@{MnCi);!~
zGQ#HYWt}x9m}t+-bU_5evBPj#{8w3--*|YiV1g=!b^>J(*>~QQaY8+^YGp9v<-?Rq
zHpPRYhfQxIPM<5>@S{F<%>%Qj45v|1#riwf-H(~KYwx$LOV{fPs|gh(J<}uIXU{U6
zYxn(#8$0G$?yorP;J1wH2D5fcTd)BID3jxtV6rCu(_%NRS)@ydwZ4g)Nis|9Oh|=M
zFNcn3)5M=h|D2{dBdHj>dnwG;=k6vkT9hiojf{Dz-F$A4rpu@)pq76@cV>QSm`~^l
zbD(93O~a8Wxlj*_La3WKOVp-b$#08Yl!wFJlkhYSXm#?0;k-f(qQU>T-cyfi!QDKS
z)4f(t+U&2N=H5FEeOVewS5{@tAym~7sy4Au0ea?i{h_|4NLPuO&R*n-6E>-XnQMQ_
z<04otdj%<y)ksr(+abGNU<wm2pGo`_{yqf7{fSs;Vt9CUww^`SId*wxY1EXT=s3ou
zKt+vo?7Dm($~>|m75MQ(foZ$jU`8`D2HY;%E##M0Yr{fBU@||>p8cAKwJma{prEn=
zc|cotqV>sQL}+9edzmz&ab1hs?I8$<Pu&(#Ptrq*WIq<a#Arjcw_r^?P-}YfCgQY4
z;?UZ~m}A8`xq6O31R;{Dp;37n0n0RBOIsjb6>Aa7kWB=3qDQ@4<Ve%Y#ON7g6s&lR
z!!~r09bA*dL2z;};owGPKk_=h5n%LAjzaE|zEx~OQC!m=%47ZWU|r7OiN)faV@|}t
zUs2of*_O51ov&7>mAIfpKhYy1B){nAYTqare5)0vpZXT>-nW(@xW`5V<%q|Efz$V6
zJ?MO@&ivYKS@XKNPe0%P6+#uv#fy@wn|`)foTIH@HOQlKFl)f=^hs#vI`=J<bD+jS
z?+HhJ3(;s>YkG_cqJi99K@7}dKGOE?p6-PUE#BCrc#U#7$S90qzKCo1#9@ysohCF5
zeTu$d2eHdE^D&i3<ux5OPYggZFmS|=<cO~+cge&b%jg`8%C{Y=thPcf!#ITL5q@5=
zO$X!BH(0U1M0aVDyGxwX(tn4q6_>5-e53X7|4nJVwmU9;T!9<C8{`o`@BNtBAn%|@
z5$rI#mk~}Fn5Z&WtOxEd?)7hJb0K7ywb&d+bLFd5DxnIi;fD^RyED4c6vLQ|o0wuz
zmA9s2WC!?4OdnER$KY8)lZl5}%a~Z-n>u#%!)Fm<fl2~mx=*rMm12ZDy)%x1@$Eas
z+Ut4pc-fEVe45ogRVLHhfoQ0N+hdxuMS&BsWCsJNG1)`&J1JLJWP-d~7bz;btPQk|
zQza&tCN+|yRn;dtREOB5vU4Q8guz-;c-vACJR+6ibNT5xA&Ag}TkJKWdRr%Dn_UxE
z7I}evtJuE8{e!16^21B=996$srEGeZLh~_5;m-G*M(nLQ6Le7U8a|Ca-`|7R95Kc8
z=OHf4q@GC%a+955(p;^*5w2rY-x~PqMyGVA$!HV2!ac@6;tj``Pu^i3@te)_v_QTw
zk%oi1JC~70?Q|tufn>)V*LbygoRimp2~kxrmfyzg;yqy#kuxsS4YI7_h4X0Ry!b4K
zJ&MUY4}-y)xy|vUt|o7LOwoRa+`d_Pcl&I;yq{&MzkGS>SoY^jI``jgn;y3otd$!4
zZ1q{Zh57gl8)HKq`Ur{BQ*D5?1k=YIttXq*MwMU6;OJ)ci*i13jUOLP&6xOBrS-mb
z_Y$SqQP8Ur{QGovm~`eaj$ZfSDUFM&LFd!aN1LjR+Za{0d3bEBu*J92Q23ZL?r49j
z(gDk(ps47uz0kjeLzx>+ggH|~V1!8@`=hK?Jt6&o>#c3dvg{tz7LNJb-ycl}s;Z1_
zrFGlbFFK~W=dK+`R`{pIl^QI?_2pO7Bs1Q4`D%N-ThtCem31(MYjG0`l+Cc5*-sHJ
zG-firfUCk>^s^OAqh9#&<FWMC55c9OV&nJ~L7N|%R#w?_y=ekT-MQR4xwknu9yR*!
zI>&iI)(yH_azs<zWq(y4NZPw3FpM=5CnvsNUlsqMc<6l4A#LlYiSDFl8F#JR-ZZgq
zby#eA5#@>aGMs0Oyu3727>eq8&%tTl*NWrZ;{-Q0t@m2YO>!<Q;}aD%KpLQiG8H1N
zN}pI&8VNi95^LO_OIjMM_kLAP5sWSmQ?YdMLO1kRYh-Ts|0*0UX>2Bv2bZVxQfx=F
zO2#E5<hr1V`q!^tubWLU7V=qt`sp#(%2>U@y?gg)XlQuw>{qWM-@Q94&Js=KQoW9(
zfyl(w+Qw`~AJox6<P>D+`Zzf`d3boV(LqEHoT)cA=uxlK*1L0TQu(Ep?L3l_#xGvH
zU`sngOB>5E{JaV@VpPoL)DlLgqpz<&d<gn5L>MBQtdKcerlqXO9)->0RUGmL`6*LP
zJeVOpPh|od1aIF~?n>a(iD8qry8nsc!i5a~wMK(TCV_;Osi1AiYhMJX>rGNGEnTf^
zY*0^;C@`*e2G1nv;>C-)Ccl&&M<HUo-O1uQPNR<w>fRY0zQ}`0``;tdefiWouOr6e
zFkF_yv3uIp!=s|Cyj*&OmX_8U+=ZEWWO9X3diR&Rb$gPClXh>aG~N>h(TokDpkC^Y
zsTyLMm|g6QWX(GXrYE{7tDJnUl`$gJpj63%spbbDjb1B8d}a+#2J$onB)!)sA1FL}
zbX!PBcVlB?q|y=59xxMlxWj$#o)%YZ!Pil;FNiyrBNI@24@K$G`e{Ms=B&R}tVM9n
zhP-)=nuYDDHa4nntR#r;*CGuV3SD|_Mqqa}%WcqlD-kLXSFR2F;HLFvTEcR9ufud(
z)KY&Ih>2)~+~*!gT@h5|iiLjo@L>syL|o=K$=AwKBt4^Yn{EsRz0@@*-S|03)8$K-
zvQX%;{HUv<4u?A{ZW1Y#2ftz_)ZvXnsGf<@>l+$0F;XWL#KdZ|;w0xfsDXK4<->;$
zmyQBK^!w37Sna^+)bQ|dqNq!1)8VS5f|3$G!PvPUYhZ3FLT;k1ojEwD8+g1wPnzqt
z81}W&Zzl592J<rv(Qhx^a+v$^$?WJBIfSsg|I#j=fxTpF{?M%6h_7d5tu-PzH_xig
z-OXBfT>W+aorPm1>j?X4{C%Dc-(crLV4sh!kTK`Wl2@$lI|lC%x$q$NmAeG&6gYx7
z`%g&xzaO6e-^>16#`1rBN`e^E(I?0L(tDGy*yfkL>gT$X@@8hN)KaAk9fykS2fe@(
zPj7&jGr4Mm=r`4WRlL$^bZGkM-4#KGD_08kx8}QS%I!PjxM|Lu(OMgCyqd!buDIRn
zqL^;(JIhE|=x>=eads}zN0}AhvuyirQTpAX(&1Ol3l&vWM_|N3>o?G+`+zNgcUWQG
zT(~LKv-Tf$Qj+OE*-6xGSy}Vi+S<<UZoLN&LVnnw_u8pJqN6yo#m&;%+UmDIiy$l^
zQA_h_upPUuo*sP@ljQc=&!6u?;oUZl>t0CgQoo4bww<-Lb$p}CL{npiKP6b#1M^!D
z;)o;49+M5Fp`$A<lfL$){1KuHo4+y(%rLi%j2S!*fyj%IJ+2vP0zDQJ+!vB_8Ld(i
z7Z+a|t#*lD>F@9F?CIGNGkEgkc5Fv>Utb@U3(9V!!v2qy!!-DAz5B%vuIzyT(|%Rd
zd9cnCFBGun3WX0W#&ykN)~8y8E!)nvwzoUtN=)6vSjwK<foK1lEdv9M2smK)1x(HM
z@!^_d2b|NW%<8RKz=i@R2ZtT_=zuFpxpDI*7HJ^BXHuidk5C2uOB8lYXnS|%<{yg$
zf#}X`iTBNZ>Yh{KHGcf*zCuAk0kgPUwo%pdf-{%6@>LQ1QA6iIN8@bWu-k(yg_ZUL
zT!iIeolCdW2Y-AL3^Y9l@!1P_P@xh*kK5l0M41*}p5NKEP>>tYxj!vlTefpp_paQi
zj~LU*Fwlmx)@^>+23q@H%ijNw?}dMZ7=ouD(zZP;sBIDmk(g;jX&2t2wUa1h4+p!X
zy>=62*&e;vBf0i`p^kJ$+Ub0?yuLoo_=JRnM*6d7i%L%>zrMm7FoNGE^Bfq{=X+9k
z^9Hz~Do|)YSPxbr2GTxbb)JrdwTTITMv!ttg0l6={6Mf}-M-ea72OuiI{K87)b#O>
zb@|uqL3_{S2T31=&Qvmolq?`oO_<-Uk?|isTn8PqPK)O?7((6!zkgxBa;}f|@*Pb#
zx3ZCH7gUL9J<sji`PY|7_c(%2L5SZjY1m;4^@=J$bHU~CDe*qwod&*yP+ZNtOln{a
zGyq~OeKRvN`9acEd(tgreoKP|b;O-fJ#B5wc>?$n64brUvOo%RNKu5YZXDG)mi)_n
zM!K~Hop=;_HJ+T*TbPBI9PDVLT}4F&&cQ0>qe)HAw(X|{ehEh3XGupNAK=6CNl)JH
z;@RF1l7H#3A<*<GC@4q-?0v3wAhmI1mO{P_!a;1kw~Jx1KJbHxl(msc<Pgeyt@1ZG
z0RpoWle4w9CVBo82~P~1Q&mcT@-LZy8aNzY0is3JWfHcuN83VwL58?JprWRxr=k+U
z(P2><SE-gJV+!Ir++cxDd}``yBM<48vd?@@`SoWisHm%#R0V!}bHS+8LZLC@&lJW_
zL0He9zlP}2BT{zT46p3?F#<CJ+2aM1G?=fIDDCf6xw5#p$SmPu_peLD`v!8=bOHhz
zhKdZcHm2LVrd;ia2ZWI-rz8vpQxUN5W^H35@~9Xz6x<8uUrS4Cq}DwTtODeNS5#2#
z;h}L>Nv}a*p5x!YpTyqyGq=yg<W#i*8dB*tt1>$?V|Q{)Btcz|AHTD(uxK!42aQb{
zEHR{OM7(*U`RLJW6Mn?b@^DsB(H)WzH?AZ#5|pPdd7<Z%VBYhJ;aNBP^~tXez%p7`
zT6%U1<G$OJ`&KH0v%Z-Bb*>Ls2W(c1Oksgeo(f##`OkEKw&{Y9M_W-b3<N19!0wLY
zLrHUV{AHaX+BPjvANUY0V9_|CYfXnGzkmPsm{0Lmft$gIo`fZ=s!s0X4d)zielYu8
zDe^|^uFIaQziO9>l4}iFqt5lk%Yj>S-Id;JW^1~@&_njFw2(<$zy2UJG!&lwikhBB
zP*4YF=I><aZ)V1Zd;!jwd}l2tr;#WGy{QyEc({f}IhNEaK6vn;Ztr&_C$v49b%3}#
zftnz$dYcI_fb-X>!{+5CJaMK*=%B_hx~p8}!$tHmdvYNmA@E;#`S!Lp2e3!~b*@r6
z-{r`;o)lze<_$(h#>E3XO2E;$-mCcj{re1zjI;lEELwS<>73jkQ8^bbA5PECFNyg0
z@pgG_58@QKC0NXu)&Jeate6biDap>}SQ&2|MJoK6;$~nKzyqkz*3pTF!*6s*-#W-%
z9LPiZ@2)0Fcot#?1{w`GNpo`RY>b?MQH63xM+del=C6AXdG(9(4gR$SDagrZ+-eKa
z>BD#G$IMWf;1UQ7Ok)hQuu8dYZ%LK3w6v*w__^YLULr;92RvIuObiY<7ongdA90p*
z#pY<U(JCh$BcsFw0rQd{KYkdQ3xF&I)wq*b^?W#UVZqMSd(0DTiTuZ0u^b)F)sDwU
z`$m-x@xWsV1D{O5<50_6bYp)zA2B&41qFLxGMOd4im2$AIiP>+hv)K7tt3$weGnuU
zm*^Q86_PR*oji5ybc}3CbMr&#jc=Dgh>j7)ncRdX-m?rNH4Zxt0I01RaCRUaBV}+|
z6(oBzi?wH60MEg3wliKL=%|L&OHcgJA~1ubjaLU&9Ui|7LJ_Bt3ax+SNS80?UUg~E
z0uDTmV|H%N0jyzWL7R6V$tcoCwnbgdtEi|@v?nc9p8Hbdful~<V4%EU`Yv3!0JdQM
z7Q!2hK%jA3d-(A8Z;Qb7uhiD=`MaxQ$d4cYqNJpZC%lb{(gVAR<($&1YJ_>F2K*e0
z*tPJM>re@gh4<tX6!?HO(7IjXS?YzCFUk1~%N~*TANb#(|DCV;*A0*)i#UzsYoy7*
ze{E$=|Nbp?0Co!${#?ZPRkd-keoR$I>@8kiUgUCMKtMpsD#<;mA*Q%l+~<4vM(?yP
zdiWK;52a;t`1+D^X}qzX&40PHY>SL6^$Pm;ML)VRv+-au3J)RyhX-pPmxC}XY-nh3
zI0OrF@`shrse@0psahzr#_djC)IyZ#cpsQe>0G&FE00?bV>>vI*uwrSC0<TW1rV-9
zL`Um;c~vQ)!Gi<YacEeWh>yPlz0|tg@J6m$(l!c4?CO&SEvGLEvZPl#kMV%zS)~HD
za|n3GOWU&~bCe+;zI`}x(zKim0;vuw?Wa%Ee@{<WI^y_a_P+zKrek4|`It)wB<*0J
zjKHn}KbnDSI)bb6_c!j2Bd^Pc<vWd5@$mBxC5yXzOh<`!T9n?te_y-NkAP_or4@D<
zyiI!2z-n9H{QCFu)U^_mLwL`8zrZT)_6S%CiTPyrPPQ+|V)tHYgBa<Zk7hyp)&P9J
zEg@k9_6_#^b@|SN|7ox#Q~``hcTiweuFTHIP4~^g`6bmrQ#i+NMPc;Dis#zvde06{
z#9kfCm$ks%G4I5Cw$YfmA6n1Dqa&c}G1T2HEzQEu`{am0&8I?NzXs<`zI#fF-Asgs
zN=%U;lm|WWNk%W1b;Ttn7WSk_`WMT={(D%Ln4hH-Q@9qaN>H7@$hdPRJ~}~#wJ7i8
zsiiA)#60_rxBK?@6#9i;nBO3d@oR7|(<5Z|o@<Ki#O?^z;e|6yK;OgWusu>>$pzJ|
z-IH&ea}-dZYEaOuoiC%Kc@2+1Vr}NV0?`Mwz_ahyWpZWxNk3P1J+IhSrXRA$bLU2(
zy7IrRpWaU3Gak&-VCTGb%Mk>wq@XrgJ<n@xytvkVz7Ra~k*2^y!wZ=q2Z?)MOh<6N
z8JMa0gSxoF%4(iGm+a%?b|xXKb%M&{$J?F?!4vya;q2Sh1}7Tgs3YUT-((&z&7aa*
z{!EQo9k8RU%h|aHbu8%}f(`G+UeU|sgLH<Mk53ICE-CyC>km1(xEy|de&%;V+!b~l
zx+8tOH%+I->#F!djK!$ZEndaNZWgcYaM?j?wa-Q+&_5N$OLl~5s1TYR7y9ASvIWg)
z2|_aR>c=^qrb=H0Th^s%UlDkwqN>27y$z!}4|@<E^iZ|Oup7ud`phtS!rN5SL#J;u
zWM+;a+sk=JJe=KZF~4VU@%(Moa+~F65TB4K2r1!5W@l%Qc!jrQeEj$k1Ru%5N3xRj
zpEwXqqUz*5Qfkw(HnANh(_zmb4(+bNL6afW9$*Ar&eaAEfcNoRD##Oh@%s^q;g3(f
z3SuHQX+K?%3HSl$EaWz$Z0h=zvdpM>q}(><@|`E?z+~bx!r3sgC;J*Py~ITg*rrkc
z7sIJkKFn%!;V9eUZGy|93|pysXSM1Y;lsmDVv5f))gtcV;@G(M&$n2}b9MXS2aD6@
zIp)u`AAc6gt`w^9jQyxWE@a;xt#UI6FB7yZpnbwCJ?c`U&%>XPE5&c>ODNbn`N$tw
zred_0%7D$>!BO-*1v99`9;j<mZU`0;2sq4gcD{U(BVm{EJ(tqSUa`QJu}6hi!_|f{
zJx_E*T5qN5B@t-5>KNL|nf1<!K*F(0OvUE=ZPe;(yY%bD#}6g;M@K0a<0(*Mt3kxA
zHH^yv^y}Bvt#gM3y;o}4f~i1MiSrzFhIno}yKrZk`LoqZmk+R}qGXj4ldn-GjgcL)
zU%@Qh#2ixRBq3ozUsv6DCbV=<B>HZf6i!ItZsh6g7Y%9h?u`84_NvXMwAbD`y_m^i
zPnci5Ftd5H^A_V}6_>{AYI}{|CDdQ=4eKy#2wmrTEREiG?s$<I<p(y4<69hEZ$HdI
zVz-}Cb8|n8I9s%Fk)3s4zgq^~N!yx!uR$j<YmTsY<e60MpsM^Hw<Mm$D(-b4V<@W{
zUX~O0Lc2HeSI90>ASWrzU{%tlNa%#l{-<k!o(B;H?s&Q{7+r5|I(we@U_$5mUiPz-
zt!0CtgEq3HLtE`{1--2NAGAf1moBkg<9ocRkA-v^?|(R)*Q3aZRUOYkn|e+*Iu&h?
z*MpUDUHZ_3h%!(w=*q#iyEBZRMut_;oYh6Gf2KNN`@%%Jk$YYCNYOO|9ohD`cUeuY
zza2HzW2abW%hNGx(4?V9{Qk`u+acY)pCr8cVWz4ymH6sqcKGof>5ZBSLi>R^Dmfrr
z9!-3fG+6Wr;ZjzHH+p2RU{U?A){!eL3%iHP83{UTE~aa2)(OAp&#J82v+&ol-+0c*
z5g71R_jjE0(S*IgM8}Bw&Cv<xnt;79LLWW+Cb}mLUF=!mc5~FXG_x(<(4?rk97gcG
z;yhYiYk+&7{q+{xw&ec$L_2TJ=q=sj$0uakhvzT@8a486fnXMqq``sP+v$XwcO-T-
z#+-wXgsL1yX79_Yb~ph<e1(6pvN09ur<;WoR@d;%pUU9ci|;)V)#J?Zr0<)yG(xu4
zzA!7we-Y@nTyt`2!FtwfzPKi_Ez>*ag|~7p%}GYmW6YpEX0H?@K)nQ$hL|Y?QyrP=
zxu+jid!Q(tz73e>NQy#Th<6%OI9t<B55hg33EI>1K8KhYXx}I{;{Mc)>Fmw^;zo}j
z_pOgB)fp+3U&2&w+Kzg^FJZeKxKlt{evV57gF)bmSY-l!8dpN{BL+O3@uOS+)B-pZ
z&&LWVvCi?fpHj7Fg0}R;-=N4YvRD-5%dV9kyIKHQZ=%C<Aav^HBziydLR-|IzucTE
zTatxMyh_dLAT+W;Q27I=Nm(2SR!M;t)Ximz4+=1M5XU9**Dc&w(>tA9d_30ZMI6VQ
z6885SdD0x&u8qY{2F2qfp41h_e^`PHwzVQu%!>-@oW9eo;-Kt-MztE{bC`K^?1~?R
z`Q6lo1(4eSt6d0!LJ~UiV@?P0?`dH2TM$WlJXfkhbwM=Ec@30h*LFP_A<Q4ZwRvel
zM!*b$4E+O#;hntHgEX1IIuHn!fP{xfT6)Z1kK8u>{rmIg2u=T1bwl~$rV{dXjiI}{
zzh-jNYm{jUeFr!odA9%K%&~_Gq;wJJ^j8(VC+y?2=C%Ws2>MOWE5We3wFmZVY~}_p
z{QU3dtlMb2UOMQKoXi8(E;u;2OuSdmtnYb3<L@#FNa6O<gt`6}jOG2gkpNQ_tcms)
zC8a?L!TdI~7=>*mvIOq6`d-Fof*718doatiSb$cUHjKEqPSv6#rWnPXGhA+qU#(v&
zDf?$g@BQq{@YC<_ud7sc51PUnly**+KYm75CaxN{wmz0$0RpPX;5muSZZlyQoOYle
zC-=%gH3JLFa(N;Iefj)t;xy%A&U_A~{b~xNn+~n#RUEKuzS~<>ipj3ei4@pylKyg~
z*)=#-#3_*!mDzMNN3qMAfvgAw`?xBnQ4&uw+lAm26fD2)K-r@#da_f^Ec5*1eL+Dz
zP0jS2@FtA8SND#g^iZ&7;J9r<(EcnFVP(_+%%lvG((e9^n>lD1&e>Tgnr+ilIPV=Z
zpMnDP#f#@9<USujlFrDOlOpNuu>7->S?Ig;t^3^EQ21b}rIOQlgW;JoXUzC6(&4)V
z(1U2(?g2^rEe1rN)_m0lan<s{*9S}9_UvOP;q}-$!Y8z1>F{&?@N=vjP{;BeDF1}?
zXWKKUNulM%OP2!mDf+C09p{3`#IpWVO3~TT(SJ-l1ZYVni0xayee<c85&CQo3p8@M
zvnQ8ArG^cjFwe^OlD67>rA4D;|3Uv``|UVjv;&X&59<K!5NLW4<i*(%rxrjatmaj3
z3J`_A)A&d7gQ1G+Adz)k9bH>2lnuJ6ntmc3KbC2#q0qXE{cha87_N2_+c9&?orVnZ
z#jtx_(-H*DBv%fC^{{;EsD-F~T90wY-9u|1nZ;lHM7^2QLqIhXXbPl5e{Kmvtpulp
z-yR2xZqoVjMA?(fb^S9<{L>Z9#+(ZP*J|Pi@wA*aYp$yliK+u6h#3HT8#mgnGK;yU
zzkK<UhhTZDIX8p*apc`kKZHx9L0Z@M5v=H9<LXZ5YJ8$&$%mO#jFzV6B%v8Z<`5W&
zsbPCl;dENmM}5>9yZ!ydq{8R*7R|;Is9t5hfc%sQw{qdleF~a0E_y@9JM5WO)r`0I
zbIchTmV){`7R~){&m+>LN-mve$^BSjun0Fc{~%nZ)-rY$j0iDkf2BTlb92*UdqInm
zii!sW<{)k#FQg6nviP}c@pf@px2HwDQC|6yjJlG2&Q=M}K>`CCZk({TI2kP<ytHp_
zRkgl$9^eHoo-}=jyK6WS3fnC@;ghTZ;JmF~S(JXUk=_<mNTEZ3xOA`1#JsvG;L+D-
zd>l|O<B)`7VJ%<us(#Pcl?Vr8-$q+R_HXTiFLRX_j}p6A@nF*B)WF<;$S@zIVa!rK
zpGnWErbZ0ptswh^l5CQ)a^HBPKeo)eD=eQ$W^yG;@3Bo+fu*Y4@O{6GZshoj&wX6o
zdyo9Q(FR@X#*!@@8)@_a-01@G-)6@x!Y}vUMF8G!h{-<v0vYq32=v=%Wg^S~{n0GK
ze8=cRe&Z`~OvbMfLN{-ML-{k01B!WFn*M5Hy{3)(m<=Rn42Nkf3mdqO=;Ds;C{k5I
z7rwXY_pD<J?lfvU8`sE@*ONw1Gw1RGqRHBF#%`)P6a<eNl`FgB{wRR@obHp^B&gf_
zVta{tWxH=Yg7^+<lh4)8+|^g6>*~>KNbQ=PA$$0x%WvD3TMJ@$JB*^D+eSxDj=*02
z<c<W0NE;(w!V2uqdB{D@=jhPV_E^g9;#9IP<>*~xz7^jc;#KQmw_9x_L%bhI@QY|W
zLMF)1=e_6-g;#E3X4Km9S}}<M<^}-mDg+VTglh|(QLS4}*x4)E)`ws!@_0*XGRH%|
z!Ulz?50ax1+-t76UNA)NDo*+aRH6cZwbn$PGE7M4sjJT{KIJTyD>ie~{05YVuP1iE
z9<^2RezT-W&OvYAG3lLoIfoMgxD!aagl8hfSW}?C^md;-^F52#Ab-CW*_uf~eQp6j
zXc&2V>1+Y?QQ38e&=_4|>VpQEo1U$&t3AF>Qdi>9Ta{iLfFz%0Vqh3CsSUr+0fn}t
ziLX>jmvlM50*#-!%aT3~4C=EMsxhDGr~{AzulCzi8UgS*S5=^K(2=rvw=j`n#iI^q
z<ah<A+LHP+GpZ|9b7&|g9=2U$<XvYnh6+?(@%yET#D`bEGdTXx*eM;C@%`h-k4}KW
zuBS*4o1_4k`svdr5*XsOGQtUv1#l<;=n@T$LvyUg)Ko6|dA=Sxzlv$8vtuo_YGVBs
ztZ4Iy4>n<$eFtpmv_dR7nqKzEHOCTAA|Ua^*WpVBK0j-Ti;H{Y<7{ysZLTRY2J(AI
zr4AC30+LgZh;svqN1$Q-Wkf`TRr#<(z}FMQmk(#%Bmzy_?Bho^8P78;0a+3R!hv<8
zQ33#WkL=0dFI^;I8{`f?*M6*-!jB@NmrZ$1WyWziy2$L?N=iE=2D+oGP5dOPN1Kmg
zJ5r!2|HDls9&ji5u(SPOM_HBXnoTgNsfMd6;`4e>a#qZna(fJZEfV<Z@kp_g&rONa
zicDu4G|0X_YvKn<6R8;wIl15i0{#}Xjxc15G^Joxi{!lf2q#seGSL(mQx&l79a^pO
z{^;5x_Oa>*RWGw>953X71y!^~{c9Ev2P5eG%Pe%dxO$vNk{|H=z&;yQ7WyG1_R#Nb
z<`+7lw0cHy$mV2<CVnOOjK|N}iqXML(_S9$`v5~+&ynD>toFp~=VWKw0elj!-tD~#
z@NfVmd-N+X`Jx^QMSSU-i#YAH`e(9^Ql}<x{|=8^iKNwi1FTbgTwFHD@BIG>^0I{l
z9W3U>gNG3p#{k{*`RIDEy+{Ira@A7`x3@jk^!@<pXIgIdMPBC9c5`#9T)C<Et}lX-
zcS(<gs7|8(j$4><J|aooWuq$u9+Dn)-5`u&8yHO?|D9!r(Sd;hVULBrf8plt>3M2N
zxqEv)#fIgI)9o>cmZ?54xd5aV$;rvF-`||={PE-Y8zT~4od&?5tjx>`_j&awR!RL=
zRCI6-z|ukC5?;&s-ASUn*ROy52SS>uj+X<tkrA&!iLSbO^v>D@3J$*reDQ5AuC`Ia
zAM`s%o$P?4eyk)t{aR>vI4=)Rr?K##AG!d%3a<bm3n1l?dPw~xOim=Ad!ztPwFFcm
zm|oeF<rQYqsb~PhyZrq80AU_fd$6;D%uz)+07_(}=0z5;<JX`hJRNw~dBQ`|*|4WR
zH0^08N48*OdLxf>lw#OO3><)K312Aa?KXb5n(+e94_RjlQ%XPCN;|Ow>;u4US;XNf
zr$Ga*ZGbizt#r%-Ue;lh^;|drQT6@<wbcN`ZH&~Wa{Q=oy$L^w+lb{-+f*Te<H6_x
zEX#tp``q_M4;46gbR;Q5An*euS_cc_BJmXUm61x-t@&OY;5wg7>ra5I<aG3rkidz+
z-2wn^E(ZdDjfBOtv%?y+CqVW+2Eg;gyU$+k7XAUZ9~F~(sJW+teg>CAQ$(~$Bks}v
zH)8D{+TAbf`s3p^)uSYvr6IHb2+)-jRA)5-EdvJDIWpgyh61zB4WKvxOz8t1jFim8
z`xR1$d}ZkS0EP-DQ4Szb>6~(108VU(w6B%ed>_Gu0Fz46EdhYJZ1{QNed`E-7Z28Z
zul)_pb|P^M(SIj8H9cw7!_N-@XB-ETlHgznB{j7p;HD1%xT$AiQuOPWNwHa@k!9Pv
z53Lp-unyRo7t<hS<@zU?s+$LC1ly3XI%ceo)B_~2^?be8N+H0qaR(Ag(JbH+J!NI(
z;o8Z`r%)2V4O}?mrAs;@De`c@u>8&4^#ehS(?pZ0an<vFfPGbvuuBr9=<0P>D)c5V
zuj<>kZ%Hbm?Vpx0i*rdX{Ek=uL2Q9DD_tglg3n<fHwrAZ$rr2Ty)pwJK4WBIDOz5B
z4p3I#ig1Acy2Y4@dembb2zLG;$cm4T*V~a15!tEOJq;d|U?kuKw+A$Mg@yIFxw+vS
zk*(T8B7p}Uo@?W#Ad%AZ^Q+6iC&b0w21xe){ys3{_3Dg&O&Vr)&2Bb?P&o^%g<W;l
z1Urml5tx)jL0k38K+*tgz)EG}Gs?aqXrntS1C(VzRWRHqjNYHS+WkpT3AWe5J_s0w
zvaNs6=L^RFMxQsoKJzzy4hH48AOl+L+Xe_O85x;yG$!#6^9D!;&?i6%3;{aSh`XZg
zUj(}K&9VOu{|3Er{jUUFafv4g3i*ty9)m?_6@RnoKSVvZ5nVaS$t>N8!m3~e*#ASU
z04(4C0+TniWF$-6jwR`ZEWW)_ak96!2hsv;9JdY$g$3@X(7G#uv~TJ>7PW^^8KYoy
z!JEKz=I7sD`Bkd}$YpH$gIB=uL!t0&Am?%bOc)3P^Rbou^~?7`e~a2=0I@~lqq)=)
zXJY@>X-KR*`UyDYVv|~J!2OS)VQ7%;G5{G8$U$6ifAGF|Kc`t%AohCm=Cliz<bC%V
zB`6n2I}jLj48T?kO+Q!yPK3Y%LTAXjQkcQiM?u?YDtfjekhOQ%lotT-v6Bh>=E?;~
z4q`Q$s76}4p-g5!|HTWR-UdS)1i^k-K8YG#_XpM=$d-!1HzJStz&F7Y$N`3EkbIe-
z8a5+H=|9+d2Ezg}GEL*HU@joXZLYz`U|W0atv>@$FzzR+4l7Q&F!u)tX~81)#1~gD
zfsj@UkRL$5#+*sgS9}ow3q<y4{T0CLbyf!Hz*m!xZT7A=n>G4@t_%O;-oS;B91e*+
z0#=w#3q;D0N!(eDsVAitp+IYh0-g@=^5&(brT3F~=ht!OurXdr(qtWfa}74Dc24oz
zUFCxUA&RvcsVpH`4?H;(`nc};68Hp3Jk{b(s$B>!H!N06U#pouH@xCS(hpj<=WT%<
zJX9|R&iXPj$RkHmq1+Kd$o?U|MJj3fX6K~^sY%5Rwt}FM`5&LA<7L1nIUQXjWW`mi
z(+rq%pwHYDGdKop*u62|`4l=+QE<lOrOL$S3(U+*rCU$IUlb7#?I!3#@C!s%4+)X>
zUNr&s9jXlsEFh5$W>;UqfePjSjr?Hw5wIHnsp^76tx)}6HQ0Zfl0CIx{yTRDovO$5
z7aM@27N|brl9P)70p|#8O4P_7s*`CFPY!5#kemX;zyp*V1%ZI|h<NvI5GZi1%ZG)m
zI~f44lkdGY9)<aH#r)QVg^CIxk_IsicI=x|^*`}aL%08UO9dJB5P<JjD8{fA12kkw
z0Ja~gH2SwA%1Td9FLMX0yz~3__bpR)>t2zqzqNXSloEKqJ5UgG={|Y#(cr`oJmD`V
z`%-dgK(@kmw)daXs@M!H*i=4$I~G7I#QkwBTwJ+ByTH~23%;5-wXk&RluyEfTP<^C
zV*`o&6A;jVdjRETBfvO>b-{>&1wWWmF93qr+Sb-TnfJ~e5|3wRXZNbwfMh$%rqz-}
z7C}7^a*Yq{!1T8p`|_`a-vag-2wSA|=YLkaqyZQPh)bCpNYb<X?<finAQ|=}Ead<q
zvvbys`_?TbU?eaf?^!bcQKx}Z@?4cM|3hPpvHLs52)dLm>9L?Wd(ibCR#ta*p1JOC
zc^hZ|+zWCX_FpQH{_9@YyM$2fJTN@4N9vj^;~gNnUtJPPDT0?+#9ZZIhh2hxo;@c}
z_;2I({hAxqKt5#Zi;Yj*{HEL5)&_{F<TE=~DjJX7PlP^kL@FfS)ndNaNn{H)2MWw}
zN06H_|6Fq*X0)V?h}y(HB!9^8{M@FeAw0VQ6TsKKr%V)RddO>l(#}&)2{65wSoSht
zF~mRTm&vX_LbYr(M;rz1PK4s<2~L5KPOyFVDi67iA82^_KWe!YwpSOYlKJwcr~kZ_
zLU_es<E7R9P631cIu0RWeuPyyR3Q{|sG9wfykl{D*|J)-z#V|#?4agn-5A8;Brydh
z{3H#fTG0Q#=7FExqPd=~E>Oj3nj>C?{q|64RnXSjH+a^)Moa6sYGAT_u0PLx=CiK$
z4a)s?7xUn9Rx;+`iO#Mr9bLQZ$D3=lX<VNMd5S`#*=5y(cP{N<w7CUPiR(s@lg3^p
zzkHL5k3%U>sbplV;>5diH1P$yfchA#=ZOL0nDV*TwqIgMqZ5YF_udkYmmo%U&1yrT
zZ#TELy3Bliec9IVn`OQ?m#L`JXsY>)QYHBP4K=ooXKG{T<$saYNub%_&Z;FWgA^1w
zNic(xL$XO>5YEaK*O>4F2?+Kt6a2-X6)*orN1$-l@$so4Dey{6C5Po7ZoeAPlh2n{
zTc<nYi3;xjnA!UQ=!?ukTWn%~*z8tBA*ocdbeO9zLxe%~j~zT=#jqf*_ZBa|lzx(8
zfSw3>!`!4dbDc$;gp${wH=0<pe3X&dvXnfo1Q*)vgxT-;rqXj(`EHRjj?i*KuT+=l
zA>Sv!v+2UC7dov)Rp8E#Yt&uLW+%2vHGuLMNWktgSUFw_GOhfCGNu4Jm-Tg5kn!M?
z-Fvh^q|`Mo3bZ=QKfiZ5RCYe!`86~-VA(EgHxNT8#NP7TjlX;cF5;5l7OiU9i@4ac
z<*5hZ0jtms9@a*$|FGO=<;QRi>Rdq9u@iqR2XP;*Q8;Li>p~E_Esa|JKd~1MF0w5K
z*l+z%Asbn(Kn~g?hOLGYZW?=9k9KiF9nT#bfEuN92zf7CqLK>IqIwiBZJlAU$9LyW
zk+&%b_^dNTL`14aAdVb+VXt3*|K`^wOiPR4Cq|)=5KRhDzoenwQJeP-r>6{0Zdc*3
z#okP{i8U-gMCX%rIS;a<Q*uwnjtYV#s^szXE<-uRhYOgYz|MoZ8H{R@ZJ~l|T*>oM
zjYw};y1YWTr@9?)jHA)x%clKvq6bG`63ZZs0c*FI`{wQkCJ^E|3ceW@zQmBvA};5D
zG0Sg&2QdDm^=P=OwAQ$rN={CGHhjWDU|EFe;>F+h%8Zll`qxcf{U+}S(c}c(q~}l4
zY3*eJus@t5Y2~9yD^}ymg1fo^-MbZ(2JBdSg53zkBvPAa(BQW(I|Em3^-7{yraYKh
z5LxN%#W<xX<vZ%$;u|N1upJ&>Aw;j{kk?C|MTn`+xQcuH@G?O;rQ}tHi}0X@%)vPI
ztu$YAEWT>wWsDzb8u44Loli9y!*%S}0X2%lXjSnPH!b5U7hvGY$@ivNAk4I|-{kA!
zd0hn${Xftwhpv7Y`!mX=r&?8Rs&2xt_WI6hCf$3R#`)T|1sscs^sqebJk#q5=LxVA
zDliGN9=`n4J44Z-*6l89CfU{A{Drz*Z7s$eP;=s@+Aza=Ncmw*r%g!rU}Nh(4eePS
zpQj(2$%|IgLY}FNmMh1ecvJ&jk&fkhu|pF-B;6jEle^!*YUwlcxbeO2mGl1e=Jo13
zXGdvuHA~FRpbv(>yF9SZpjj5~#ii_FU*#3!XO8B+Po5U0cgS6bAe$*#7%$slZyloF
z4_o!G)#u!ny8Bplx2UhmQj+j^qCf)W+W^iH`aQkD<HdkKoCwNnTBF0<5^cz9^gZ03
zUrQZ2m<TNKa^ZB4OX`Xfan+BXIN2{fS%{Wg)dUi|4Ol)Tm@5<#r8w;-VZ|#(LPVSR
zXP2LOLtD=keCKAAP-kTl|AqSP)~=vl@H6B0jabFeHT|GxO`DN4ok5qA_YAFg3%w1$
zhJ2-|4&#=(VNzVwPeV`Vji#qF@Ui>Y+>*#GCG^;)T)_o>Qgxi{gKnm1R)2S7#CX}&
z;*=PW#vG|mT_>1IX(B3Ci+i1){N6A<k*s*YzH+GRCcZr~<O^31T>-veEY39rNFCi`
zXw3UEk-wWQD|h-7ox~cTqMRn}%)Jl@{V6-42unlR1-vNNcdN361ErIjv2V?vA?Dp`
z0SXtFmWFCMq=)*`Xi%EIBKw&&<{Z#bstr}GELv{FI(TZx+Vz{<o5E^n<gdK4fA;Jd
zDB^l1_|`3&MLbi_ya=;bV(GAq(1G8A(%d%)T5#&As#YH}4%HJfBTjj$H7-2(#Cvq;
zqtoo^C$L4~Bwc6fPSs$71YnOXJ-HBbT#NI-9DQ2<9OJ;O#|R_N@#8XBJ!b*5g<1MB
z>sIjja~~d$mtcuLV6=fVSbTWct-S>)@chhRk?@f3mf5Xn`A*jGM2&`OW&KEo-)33f
z?O$50@pvO<f59^XJYGNJs@zNOa76~OR?B})&>=g-aN-=cKCoAAe5bF=*UN05r>bat
zj5@k&3?D@5?J5>udEdoNJSvY%x|xD|P-YZpdLfwRbf4YbVM!dAzg69yuTC^kVg`ul
zAiWJHVwIMos~Ag|Xsnp&9v@KQ?%G_u9GUYxTLd6bmali-a~ez_o=aZv;yv{=X5Bw?
z%<s3@_d6~>gzi#5>HX;2V9vH@q>j+xPEqoEZ3|8O=~T3O+}oluwo}EwG=C-CzoJ1C
z?r&IM_3U0JcZX9&;){gQM%4rSH0C&PHRZG#*JAV;?ajY-??KRV!C|=g%d-wEzm+PT
z2xv}dia=VrO!blcgQ0w8*?Et+PUpjuomCl$!h@daJ=;0;Ip1q%zMXnK1tn*oG_>I3
zbSTwE@Ku^dQqk97XiSHwPrZ0+Q9NW>NsXsj(r2R8&q>d$uN!eHxTB&~5K*VaQ210h
z^X2CI?~l0zKYgl)oeDd3@qR4S?a~}kVm`HYcTTi+V#ZZ0?brVJgg{Z4oiZJPnlUD9
zP?OyuAS#Rj<r3gjH}U*OhWk}#T5fIJUja8iEFXHNsN3@VqI&h!3zabB0kkZ|{9L7*
z1M}M>{^;*@t0wn3&g=T@+-gsjg-zD-r6RHoZ}RsST|i_<qdyXGgF&(`FUofrYc`FE
zGIHT(LbDxFAI`BIyyEk<yZk9!U%AmwRY%}+i3p_oe0qO|Ou))L;;gs9ZDyHk2y_fp
z^e3LYyy_|`Px<oqquOHX-?9wNCtUrk)@BV*!R7#JSux`)wny_vLZ94XIfqX%m9`YO
z(HRcp31NwUtU|cJeepx8uTdc_`<b5b)@h83!TIzM=hu*^bFuH*D;7l&JiZGLitX&u
zyc#xLk<(iyLE^uHY8xmw8PDpbjtYc#p`;Ke`NUwB)waO)d*+g5Yp(lN<yVcguM-H~
zSdFm_#lvtrvvrHElC{$MJM}yL5fOep@AxVS)U&(0lQ17UW2y1KJZ@e9lWB#nboX?M
ziOlt#emWt;eK~gT72Cq)(|rqfy~1_O(}JqUk5FkmmkZ*jT)FqWW&74$%(%$In1P&_
zj)wGJ0FAZd?&gxI?Yozhczc$mjcj+=oIJ0Ly=<S>+TVNYXTOW&5fC3(ax)@~Fu4@E
zyfCLECqt@+Ftcw*DP(CoIZ-{R?s#Ep+BH8-H_SBerpgvGv!N0i$ZRg1+UC^jY~N1Z
zJ|2>gEho?{*cU~56pAVS@hV*cxSc5ag_V<9d6@soixt0ldI^Fgt?)s^zO>&C-IM86
z>V!gzrPey{qZNs~lW*#2vA(eF;3MB4@f9bHO54uC++I<9Cm+s{Am}JjS#n9w-=-2e
zA2P9QTUp9_rmE2&%E_q|O;0^()6SQ&B?9xTzsi30d3RGI=EtITVBpr4uP;VnD3FJ8
z^2sSX53!cT{)G-&UB4X`Cu3O5xS@C6xZj*jV|CCc^(5^At=N*dmf7;U$J^&YxryH<
zy+(oC_>J!I204BuW<IVL<)-dW!>2t*%wck526Fo$N7K>7g{~M*=)C&J?DjPOhv2F!
z=BPX<S8O-vrmbI_M^*Gk?A*FLEr_sbPLVsf*6YP9v)vl3x&U4iP=J)>#19zMtspSH
zupyt_(Ywr}B3x+`Vzku3fHI0wkDcFi{Nc^yIFrr0y_$Bcwk87uYuS(+fzr&e?XIem
z+P(&A+y>G_q)`A(8A@sWIV~K4KCgawxJDD08=B;TsnZboZr?jlIZ<zfJ6W}D4~(%-
zBvq=guSI8u{@(Bza*t(i#9~=t9Rv@d6s*nUmk{7mMn{T7@brUrZhE~wkjhUR))-3-
z72}By4vyI$va&-SAd%#*4CF!*DFaEU7qT^aSv^U4D+1#v9j>ETEN4bIU=Zvn^#{Jk
zc8~ACT%dvlv!iVN-KZ1ojIhsK3~3oKS;x1E=66L_uoug%6LT#}pM>*wUZM$QP}`lV
zm|pjhDtPa=T_mz{rs;=6EEme)z;AbDpIR<hk;(1waA~_h)X<db8g#^LE2Ly^3%!ID
z7=JI@i^{p=f}6c*%TiwCt&?ldbMdXm_m@&?)jm?=OTq)5V%e7&7+BnGtEIo<=C=q7
zdbi2a1uvwN&BVR_)Yoj_jOkoK#^1ff)6Jgit|4Q*hFk87PrA&_g3{R|_+h30a=N!?
zGm3Jo9@Y6;SyS@_RBd+ESWhiX;x~NioOsVbl1Z<paO*rUy-Jv@QqpM$4GbO%OB%D_
zUBwtqp923dO~_IGb?I@QI7o162!!rhUH}qSo13;7<G0yj%c#+&H?x{`C1(KeiQG#o
z%baN~$nklGg7tQa1so-$;!W=nWQGATH5lPCmgU5lBy2mFQ}`C^usC{>AuiB^8$BG?
zgPFgjAyU4!g~$Pc6U0p$d`?hrFg@|^PxG{2Y&{tL>OfowMS9M`f|N8|d`0kPwvaDx
zBmeY;`Pk&DEbvRsX;3&7HPuo``5Go^KxE<fvkBY|1-o5aAOcxeH}YGB3^K+j4O)Mw
zR-u1VwkSc|Rf5H}tZGVK!O5QL9oTeV)eqq{Z}F!ekdIv*@U%o>@cYKXc74~7s#FjU
zonIkT?=J<Fy$CseLPW!}eac>p&l2iI9ENBOItD*@MfHc?6!ynhCk&p3EL^FfpqqTn
z0H%g1=Jw$Ep`Aqf?Zc({m@Ai5MeK+rQGRHfJjIzbiwE&wO5Z@}zPQoR(OnVH3;b2Q
zlYXQ0prT~uz6n?-WMl`Y1i!CfDk%<)O%>Iz&Jv7F$@}cV9ARt{b2Izx+En$bD_bo}
ziyXK78Vg!gRHSZoFcH4buk%c9^QJytsExr@?lnERk(<%D;4RaCcGt%P_xZX$f^71X
zLa-tN$~7Iq_KR0z?IKdO<4mr7VM$3z_j;oJx*EqtLo7gqw0XA(ivvo4`0BJ9UP=%_
z;doo_YUAD%7ia~@FMMBnn9C)5DBpNIYg}+_3w8U+%8c6;isJw46Snsh<U1cQU2g-i
zE{n4hX8iRn1#4y}WzW5`RzzptKf00ODweGE-idcFtYxZb{=UfBn~YBn0h_Tpi%eqk
zmrCU~?;PPuVk7Pl>#T?@jT_eSCSf}Lj~-jnpmX&mg5<oe_UceTFiOFSc`Dy-_~D1s
z9PxGL9gn$$X5P`XT6%W6Qy1|Uk0tKlt85*)_Y`k@D8hr>wBSASS*n31fB%N=SvODL
zpDTBW4Xq>Vjd_H*@#@N~9eFyu@T2L-wu^vQx^h!4YapPBNZGjgU1J-|Bss$;W_Zcw
z$mr{Uz(N^5DYxLsouPa5Il^a}So=|6M`>TL1a%Y09ZK|sKw57*X2AaQ=rRwlgsHVc
zo-q)-q*H?yTUwXSEf6Xrw~uotl;Mi%bq}qX#xtZeP(u?|i4mfY82%NR3~ItY55H(Q
z?!#xi;7c*~c31YCYU{fi({7BmNjw)m*xcMam;&N60hV+1YKNX=5V_um!HVG{wfkB`
zYXUYx!D03s!{Ut2iJ3@!h5>uh#GM&7);Cu(uXA?=nI?7fhmQ(qG@0K@@qmY&V@=ef
zmA`iWef#u>u^y!(fbpINmh$m^wt?L6AP&J{F;#iuSAk0Rtt`Ii%28Ig8j4Q$!&%<H
zu91R;YV-zn7SXa|kSf_%16y!Z+f9m+tZ<pB<pi%);mK|~E;;G|z2m<QM*ovlJ<xrQ
zII=lNhzJ8hVKGLQ-N5)%&)NV=iX=+&`d2~Vl{|fd8^b~px{Jf^x-W3H2W{^TXEvFZ
zviKY;r*N`^;iwY*RLN@y)cbY78~6!<jwVmJ+7D*)3a><`NK9u+TF=z#5f07moMzAK
zNqZ+TFVP#B-T=sUeSdR^`n0Q`=h8mT{+swo(e9AIz>p^zlcX`h;SvO+6F4cw&Q6`O
z3FziDHanb0RnbrDT!hF0_3DC$?`m+HI<Wu*2sy`fFKWZoY6h3!t}g}K{H)#)o|sZW
z{@zqC{AfK!;#81b0JHl}11t`>FBRCb|Gytt^qHPk`bYsRM&BMV^PRoDI8Z68Yi8!`
zNdi$9a@}O)CP1A4>4!o+Btw*9rO$>%?U*y*<P1$oxPM(nvh&m)pd`0{M+%UNxbL5p
zH$(lYEC)KPVtx{*F{Z0N5dy^MF<6wjtawV#U@aL1l^rN=QsKCG|7_Oq9^F-u+ZW#|
z!wZo}#?z<jD$Kw|PXzKPwo|438O~BArEKj7z_aaycu=^}Wk*0(4*>0Bq69xb|LV~V
zL7}oVtMa<jk-jNTF=UW*zP2k9@5L?Ru+}qG&VJ?h^C&7f#96vwl)3g*5?-KSizd7U
zo=tjt0;o=3{DB#<-tLIH3MM-G-qanWs+?&209jprzA~Wx?Y)3CQx{26V~d&pxB<$)
zpeqJ?diJ2$B*p9|Y4C_-z|r-kKOm`2F1~%VWa8;rNh)7W;=^6N4o>KUC!gyB93-=-
zb8@QB?=zor3r0qa_{^J31Z{gN>V&{49y;P<&DWveMFvBi@w|FkS{X8d2X>(DjU>at
zS_57tJ*jgavf?2j3^<YQ6iM8YHmPNb84p?#L-wAV^!g;od$NBXJeVI33<9+xhPfV;
z60S_3(XqS$i4N8~-Dm$1NE<=vJqqw=+1;gpqZ<GW=)k#Opn(NL((h8fWj8*&;o;?t
zk#q(DmJuixBT;EUs=;4Kg7Zz^LQ=v2CvW%T<2|Etn<z8C#hlCS92d`+x_1l1vjI!k
zara%xoeQK>Uqe!;TJBZ=#t@1D>_Gfuu9z$37OgaTdU^$!nGa`YXH7{v3Q~8POhOIt
zL>R!y<%^E{$dgn!<tXgn;6mF~m$=cDAy6QBrx57bAf_RZln_vErvvDB(O<5wnx|Zs
zj({qq5Y&GbmX&FLecIOyM1+vO;nRIr{sProq?a8?`|m7#rDl&C^x54X_g}+;lAsE^
z{u`0JKms6VTmC86-pKTk#zmS?T~MM$YA7S)!PlpnSS*$_!d#6sXHQ>n?**$<3tm1m
zACx?Xtr0oVx73sK0n`s@I>RTVvZOk4vN~3HM>eSh67Ys9a8iX9sGKx5`TJMtzLlrd
zqm^Jbu7rIs${~%WHbk<|@MoicHIQNOhPOQX^9Hbm8-L^n*C;An!6}d{f4{!q|Lb87
YCZhV}x_KQ-J*iQsg8HM9hfiPrKm5yC@c;k-

literal 19722
zcmeIa2UL^Y_bwPfK}0}QK&grd(naYV0qGzu)PT}E2%&ciAP7j8-lY@i9aKc5cS1*`
z*HA(yWKR5jzgy<c|IYo-nzim)vxeojk}rAR^PaNLv!DI!ee(Xfnmp-MnyU~9gj7*M
zRucljAAmsceqSL3pL}RHC;=~orpofN;1&Fx$TS9k7h-1xeOCzN1|0Vv-mE2~Cj@dA
zq9`k^?U}kc<K?Nngu!BwoJr;^LCjKb?=t@)Q@VAtyyl@c9`*Aq=-tbBB-%#SqhooQ
z=$uj0jmQJttG5S+E>Vzh-`%#BZZ*KelZ%7kKO?TN=vT2k!SFsZXJt+5RQa|43g+4?
zywbALb2e}lkvhH7qAHQm&x-39sVSKU9(a{fej14D`n9`K1mK16dM55uf-7K9kU+dk
z*5KtU1fLJQynx`*fR{&skQoINzk~P&4^-LNb|cSp9{4o#4n-zy=aP#}f4aB`eD!&q
zGH6EX3Df_2qokmf+b>A}c525}B+a1OHvVT6Bf^#F+Q3pGjA}b(q|i<mcGPXc{@}rb
z={(tMxL!$J8gKTgExW_}^vz`ZM%ebu$Q(nSfBjUOVVwG@`#u)FW)@{T_J({NIs7@K
zHvNh;Zm?ILAdto7NTVD{@=Zi>g9nL+F#j~%F#vWLfX#O$N=Km4Xy?S6?jM417gwrF
zVzlWOBYy30DsGvY@`;M8lB6e5Z)-B)<0j;8kBh6To0_R_U3UR8Fqf<gvz+3bNI~&;
z2461kpP2I|eTO}3=Cq7$GIpx#4!!cvq%tsFf<UbAktM_xbx2Yv$E3HOx?r-^``Nd|
zFUs|Zl$Go-o3*tU23uS1urEULqGXJPLsYbsd0B-p^x~I$N|E8Ds=N6VU>0QSva^|*
z(L*k(X(Fn+M*c~s2OIR;F-5C!soA_{Wm*X1=I^kp&XYbT8%8>GjL&u#^HXduS=S<I
ztr5x&ev!gy1_!;|esz-SK2B*x51+ibwJsA#7<i@T)xtB_Q8-7uzx(^sqnHNfCJw`s
z#(8dzuPd=F6V`*FR*Zu^a~B#O7Z>TyiI-Aqd}Gp&;jxnwW^fx@-O}-LEiekVXC*n$
z=FaD4a*wSIn^(>*YKKR;S)RpcYieYDXcwMzEs>g|6E3P0ts-xKkSYMC?Ru7yva%b-
zO?T}`@S?zY<~vN)PmIRPCdg9qf;uh3KP#Awo@V=X;IQU)wArMIUar1mvy_ZxgLQjk
zn!!mg_Y9x2R7A@%p;(kPZ(38gJ%0+NS-C=th=?tri?$qYVAb9SwrMlPoZs6jGsF#&
zomI1z#WB?Qd}@pqd3#=By?Z(>doEZ6e7zhUj$l?q9&MCX{=sO?ieGg8G$4TI=TrYO
zvVlgi29pk=_xlQm^7q987^C`G5&f){ow@lT<T^wl&D%^(3@2gXG~W8EmM%Y(D1=hq
zLP#MLwkv&z(Q31OjLfLEm0NT|YZC4D<G{>UT)xnjCQR=hES<+g!O>%8OCsZ<<yH8N
z?3~eBKhY%Sh;^?7<Nlc(m^{fy&LrIsLxy^~QqZ(kcirl`Uz5Zf8{=K{(n|Ni$^>as
z(dWt;e}qbBTFTZLO&IQ}T`LFA{zN2-B=wp`>fMzUp#P<-DEE%JaLcS;N%Ob#*MNYN
z_~mOCit!iEeoCD&`^6l+3BK@o_qoj9P;%B^Ylr7I-|*`TTd*Ij)2Qq(n7+U?9H=5%
z18U@EZQCS2d<YUyq#3D?WiI@+6qCaER{g!`0A%2Gyl`H;UEUEk1OEHtNO?2g;?uPT
zj@<RO+Y#va+?;nqxuTe<qn?k1R&Ou7AtTz`hpTfY+aKG_hoVKDpy=HY4D;Ie2i;Xm
zzoRBou{lGcX@je%0x{!;i|Fx>bp)gkE~v(P^hRS|Kxxe0ULM)|A#zEj34~HCD&}e!
z4f5zM_Cnpo=Nv)l#5d=LS71BSg4sLuP4-l4WuMoNu}nn;1n$&xnT#B2$Y@U1L^E%R
zc$ajph|IhqqJbC`u%^dGQYz_Qi*~$-H?LHrbhH@^A1X#E1f=m`mbe%E9$?KOJgxVt
zBVzkDc>InFv5|1osU0JZfn?ZVb4nK~UEgjh!p3UjXH_M^8Q56fUHcB{B3WY#b3LDi
z3Qi5~gPmI?wWPL&bx2TQSxb1YIkMZ@n^Q96^f-?nrJPc$oM&+naz-<G^y7IXYCrEy
zNS+Doe%`?W(JE<%Vh*>s94yc4^hLXVR-<#~9a16p_okEq>ygsVU#e8Vj+*40d-wUi
z6oC{8ykr2IK6Ye&SK^y#LZ77kDLfp0TyTqWbf<+CA1z9Ij*{Vy6M1k_YhvKK4%ynT
z*MXip<RC~wnT`9^)n?cf1n7kdbmlf~YfyIA5<58TWYX2_FW<(^|G;~qZB@5(L#WrO
z=zYHjog4AcX0CRT+_KIXE~V2>DJ=aOid>-75mMc_l6?*Kahffg@%Plu%n{AUk$ck?
zvX{31cF?r<Etkm&<6b4yyX@COyu{S<oMYl@Yxnd(OH9?iA@(|jVuNJ^cE-<_x5h(I
zLGVLiBX@XvRwUbf8b6p6bNTWEg{9ton3ab$f?57U`YAltkrCd)v?YPk4T)OaRTr?`
zI`+Q6x0VCwGJxMMt4y8F>bF;4<($OlkQ(3_Jau0jTM-~#xqqr$-VtJcd#&*l6Faw1
zRui^cRGarcXQL&>AzX3jtFkU3_I1!BXTRM7-*>&0H?o+y7~JX|eKD$-l?@|Im9o_Y
z=RD)X9e}TH7Ez9;dmU{Co}-W4!dDbc+3@kvM2x<f`RU$8UFAJJDiQRe(3nN{#%lPQ
zX7kB%ijsoDz+2**e4abMrRJX35=iAXxEApp7-KnU>}C@in3d<~cZf_wU=1rQ3w|)l
zi`{<9%BuAYoeq_`Q&yWJ2{sSJNU>q1G5Y}f5BVu{#zxuLp^QAbJt|~zH#k75qt3sX
zza@YX7UN<W!G<=d{6e%U?sK?0AHoR5USRATG6bE{Z``<11-c`grq-LxuguHK+x6?0
zb%Hx2RNG-8cZKmla=Lc=rj1j2+(;kx_gr1{1#i!@GslxINlWkJ-^t{3#(qg?w5d(6
z$eM>%@&<d*Sbji=`}$lKtgibbvqkz0>+}QJT<~C6qx;&V9ip_l^vjp$8)K!zVU!%^
z<$8<GLVOlox0-jplYj?b@e+l|iiE(Debca%A6lq$6`eTCnzzD~&v_u$Ti+QnTThm0
zEv&3I7NR7fDk>k!S}(jKZu356iB<GD+SYM#sjwF?%nV56hQ+0%a71<7+;b;sErMKs
z=F@_K%E^C+U4~>fd?B<>?9QvGcudM5rdK=d5`LA0<Y1?beB)~nL9&=v#oqpYd`yhy
z(v$M?@)+j);X8ZoA#ECY4(QdvuI_G<&7;kU%JM`qNIrVI5!JmO>v(Qa$6@p7YV9pI
z(;*v(P{g}~UA2_n&$=dlIxFcuWhyMOe-;+_U6)mw&-c3bJkO*24NG;NwDjw`N%<8_
zacqnhC!QXm5Eq!UV#8WTb{?nC^$w6KrTvwDWK!Bqz$D)I)eq<H6FxpC{kq1pDI-*5
zH`mh4jHUmbo6EekvZ8Cl!^IWfX&y7<Nbcq17A0|N{lcd^flCC{w4V$wg*?xA*%rJO
zb>u#BmT_xsh1y#U6ZGhGJ~M9L+pu|e88gp)8S<371n;!6Ut+kMr?sTsb%ijJUdUN$
zcXzj9YD-B~b+ioDSoA@Dw$>?QXlTg%anjRSGRQnwzbfzEvj)c=O7E2%%zCuuy~_HR
z!v<b6d73xJ_?{S=G7lGK5O{g1780f*pPiA*+I;&SFe;pL1CpxtohGwic!f@U@7Mj)
z^ZkCe?fR9X++11QsDHe_Y4JEI?&HVIAOa%Wlf5MxN&hCWO4TpE;m^G4ppUt==MK42
zySQlXaD43QcX6_Edb)vFI`F1N*YilG-}8{u$Z4&WpJcq1IDm@7Ou8H8+aa2vi7@vh
zQ&UqKj2~nm=9BLM>8S(Z$;pX9h537llf|2WW3GW_kQRV?))8#F)W^lg=N--ZsBcZx
zB1a0fidA^2*AE*~Z-0L^tSV@8aeg9weiJY7vyYR0jL}jE4dz5zw?wN^P3E;8#lT+O
zvN*em@qVdnXtI6n9EX<6igTid%s2sD8Ss(oomU2=hAA@Jcr*#%$JF{nfC(F1|37i=
zzw}t@KTXyD|J(kD=;eRkl!O*QFl+f4>wTiA5|^GWce~D*2VFfqBfWxiB?lWwK|4{f
z^3Rj_r!S{IH!vuYXY|vyO82siU)W#n)q48$bpkhxJ%6}RztWP8h2`0uJ9m1A$iN-i
zRJ<OU8D7hG(tNh-bs{OLsBC8%+>(V}Wph%6(~*UIR!Pd<+}yOZ!5LOjcqpXVqMIs7
zg#|dEny%eupCfi*VSRxVF6R>EMDDKM-uR9MVt^d&?d|NGoZ2s6W+^EtnG4AOU}OaM
zCu@v{IL9R>Zh(P|0f;C@wVahH8T$LTXliN}(S=7u#4`)0^?j@rrHf>DxXGgO+}POo
zMJoX0;4(^ZKbOfk0a_A!qTc(7SH``udCa}I+V;If210*|Ls?y2*VeX(xv-!>!KwKG
zn&t<BNN4Y>c7e0yW+7LvUgfhM$TXOWjfsiN$S_&bv#}`zaJuR4^J}U&<%yNrU!22j
z;OOqlAZ)d{3s7voUlRE<f|lR1=eA-DawEQfW%v`X8K2XvSl;K)w#YKro~Kxa9%oFh
z-`OD>7ndqaEZgE1=%<w1OTVj*=f}I=OYyo@E=x+KM)jKN5+}J0ZX2KDIdqHckYzZJ
z4ctVv?}^KauLR|-Th`xRU&hTl^SdT`&_)6GrdcM(layY5%bHL8b_o>~6?~4<Le)-l
z#>1agwK639l1u+y*RxoRDij{E`3I)q{~*ddHxZ=B^FIcZ*$d~y!J-MrR^88wee>Ko
zge;E%LH|#M^8fob!+)<gTq9bQ3#YoAM-KWwaD)G$l#f{YSG%7d?zlA0=tS!H(~Em+
z59h0syUpO%Mo*7+sME2w7=*oF7%3@roUj!4pYfHlgYWJ>1vm2;H|ZH8bYu8zhM2kZ
zD`YKQw*(QQ?wbYoWP&dtp5iWh&G%4B{nMj+Uod~dd2II<JMVE^0iTDO2QxDHVh)fp
z8o1puz{Iyo26tohKh84tKe1R^UhZ1h9htP+nygl3zlOVBHkbfiVbL9zlvKDq(^%{@
z*HSTIpH)DD`%VFTM?fx?DLCgFn<Rj1j{Ob;L9^4G1Szcm=dEh+3525GIk4~7ujd*X
z8V_x8(hF1iJOuDE`x(!L2$y=JtO8s=Lal>4qyo-&JG>7@b>lyMDgZvb5Rj;9?xMYE
z*Z#|SQsAPyF}KMEz`jWs8XkV2NJK_A037CP<6GdX>xE!~!9(9>;{YR5?vWTFPRp_G
zO}>@c-PHwO9mskyiVr@KZF+Wj6}UK#`}bwYLSz&b6@wPQ&KUiyn(`V-hil0OwA9oj
zmY0`TsEaT*sCTi5zas<|DhG{*Z_PG~7Rce+iT=JL^Ojs3{SjyLU=<W&+D(b8decO8
z2?(zZ0L~O28+)^f9Sk_|%FAOiG~j<=3iYkd#>>sBEqjw$D8UYf$g;QLP1;s>t?ZBb
zJ+0JttZh`3!_852mvbNkU;$LzMtW90YHE?sB4|c)v%xok3KpK%(5b?%ammR=08{PO
zM+(6XpRBTpb>AG%+T64+HEuTWIou?cD+QNJjo%>~c>n(WWW8%Xi+Xy=d|QaJ<_jBC
zy10)mprz|$rH0SLflLEodN&W@QluAkADOJS(|P_p6LWUdz2w3Mkh8P1^II9tZbKhN
zTpeKj(b2&nAz>_$p+4t-R=T&h2VYLIA}q!&|J^@)whzAJ?0j3`%vS=(Rc9JKbOAwy
zpDZPIufYkc6>yhnf^{s0mJT%^%@$8oS}Df}IL(%L%zDX_P;owYbS#^!aVYL!3=ms=
z_jjtl?p%+QXYi8UKq7SKTKpa;@-(Ak02=f8ox&A|T>HiIal~n?ME_eU&K5!;!6+%G
zsTx&){A8Q0$A7{b=7VYR&m5*|MrWIS-e}<<>H15s%#Je+g;iDBG4S7C@wO%^<#ml-
z$z?Vd6ogI)flHhxAJb)8^?b^DweWK-Ue|bVDba)*FkKR!rspwm^s2^;|Jl~*5>LP#
z^i$kAE4%<J!&N(-t(qcW1L*VPlwR+xYUC|S$_Jg2l5j8!RiLFzYMitxi{Ivv$wd`T
z#baCoxOseh{8~|Q+0`n)!wD<kp>zOo`)6I`a+ZS<azA~#?Ip%QF9Lt)%n3l;(ltk&
z7}reGv%H!AtTl`Fnf?5r;qBevj+&((Vfx$km&f<MhcI|6=<Dkz^V!5OV?V9lT<%LN
z#v>rKKHQw>>g*&ojt38vdi%(i$h$p^61eq}O25tVa)U}s)z~oHIx)r1Lma9NmiAB9
zC9)<sYo})ZZLva9QH+wtB-Gp`mEx?d@;T2UY&A7-t>r!zBlPY|6~=LHpy~0ut!sl}
zI!sn60x)~-?p^~$7?%RD;5+sO)b(^l5?csXk&TT_QP<cj4{VADv7O9{074BK;Dxik
z`>GF^nXSLSC8iPcbpE}c<%ok8_9t|iH*emoc3n|ZO&8N|3n5Qr&dbkNj`2R+EG#vw
zEdo&psfeze+&jy@)Iwm68y9CNUEq2dJb&X&*bBh?PnKeGz4n*c?%k6H1R{QTZKOoM
z^486pIUhcJ_!sG5w;K@~_-FIDZI1JTPO8!hI*uI81t8KyJ#4=Q;*m=91LGU5ceTQ?
zn8!&fVG$9d{6M4am&6*yXzS?}xNVHt{Q2=7IBORny5z?eVHK5?DS3bWXWtI}Dy?Ik
zv}3KIu&{7gRFtl-uWwFklU0AZ_j+MI{@K#f63*7}%?ggTXNp0nFkxRKmB9^bcqk~S
z%Miu7xD}tj8k>u1#Tqk2$;ZHVeoM>A$?=~~g4G{*3@`$O0|4sss5td91so<Xm44hn
zv>L1BYk}U6U8Xj=+bJm_^58+9QtSA*o?^`K+BL7nw)uh~_x)Z$4jvw!FB!?nlLdAF
z)1<mBJg;>8X%8m|&+}}a&7`aV?19U$Mg@n9en&IZLXOqY-s92y0TMSK3z-nIQ5{`a
zUU6B25;Y7S9Tpb0z6pFw;Bf4`=m5Fo`4>Vmx&lBFpMhy=3MQq6T3q=-{7jdLFG&SA
zOOIZ@W>HPf-P>~k17MSsY*lI<8qx&ZZ({54W38@72ffFonp*pb>{cuW6<kIE?13v|
z9@6Z0=3YOu{gv|$xQ(5S&1NJKVNk6^M9whc<iDTtl^22Ji0#ZRD0uh}iPHIbA2n60
zZewEuR_>k?&aR4pU6}-64l?~t)cyBv_NEGJ-ItFXNwP{W1l{2>s8T>0xkdnn3N!~k
z=igFXCM`w_9Z|Y~Z-QCAUU!pj-5oTODbj}5PywuRV`sh%41eg3RVsxi1d%1;ww?>p
z59&ajuL25ud|X_vM$R*v#b41u5|x&{nowv0;C=arTT^xb{nr6`sH`T)f<mF6-}Q5i
zY@Tgvy^@aRuG`FZPGs(sEGe?G_2^A?FZSPy&3M}gJRMVTr|z7VtK0ErWk1flPft&?
zcLW?(X0UN^DC3~hASl}Qe<W;NBhmNe`K(vvPP2f3M?Ic*pHh>&qfm03&Nt>Tk?kT6
zf945>46{pq6x_&kduO(SxxR~>|8)ONnO?qoTJWWErp%>Ry*}DN3kHl%=N}!n4fc%m
zAJnEcat?a@<Ox4uux?wEI)LUhdZ5ODL5=3CGx`DMQpzMtpz1bEUu!-0B8vSd{Hh0A
z^~Uz|Z<&`c2S@7Gi=oAF2?<o&VIZ|3+J<)uf{~G-JZ1iu-`}!#3=a+Qd84hljOw0)
zz1eE)yW5^gNt_xf#t=1YboQA&LH79Ly+*lrIg<}J=9JOelQ-wf2qA%Ao`{j7LAa6&
z0DN>p0>D(ML3I)E)urFc%{t5@q7q7?B^SaGT2+UOTcjR$05V?!%WBf<4;YSHjgJ_^
z16EeS!iSloq34^MRx!TPMnc4pz=R&>z?OwbQTW+bZTFTt(t7{I_jEr4Tax0~!1l2H
zFBAB=b9)uo)kCl-JT1rBxFrWsi6P)FIQN^tW%wDm*p0Hr?IE)$ap;n3z;PSnYbf6U
z`*e-!B+S5PN?bhe-WF!cOi-!aaZQq1fj8ctCA~@ibnroQ+`aZw>Z>us9L?TR-W&33
zMEP-aToIcOBi6^6f@6S_BnD2>2i2@=S*dSfVF4l!#fU?&DZ~#(wD==$X5zgOi<x%o
zdAwE%pWZ5NHaY00ZpgDS%r&FD$huFTlZROv!XA}Y9cH4B@ZN$<8_ieeK*!CB83@f<
z-Zi!sMLUKQ@up$MbcBdZLw(9EheM*x)+MD4GGzqa$FkzHlgqXJ#ir`|!y8hSJDoeX
zusU&G&FqJsueC{O+8hPCP5*=#+`Vym-oyfO&*E6|oxv?_g%lFz6#Znc1askwsESeM
znF$%CsKguD>*DRSA(GIPBW}{Cyw+qq&|3u#Hzfj=#<nJ@%t#ExA8qIC55qO41m?L3
zq_}`wk=y8R)_WZ@nHhWU`0XIJcoMO<6M;&*I6@!CW{57vusBFeu=cTYsgpxTA<3)5
z*?*`ohJj=n8Z@`5arusqfc{7KZN$<%eq-c^5$kA!!>gIM4@X-_?`}HI)r(F);OmUv
zx>3zi8c^2_dHNbcD%%ho8RWl&-{~k^k1bZcXzYo>21$2HerRLG%DknwA2O7THZopI
z#-{C|o9STuWoHq=-SVD7Wd8TO!e}ukv?45pkFxBpbJ!aDbafw~&KcWagn`4i@MuEe
zr1S(Z%**?(c^KxDkx}j~p4bF`R*UpCM(pOHB6ix)4H_kacFSGu7xy3DX`q0jFitf_
zlo8|uTCt)<u>82Y(2=HDVebzR(!1uk$tiZYhW>bXQJ)#%S%Bd43r8bRxdAJRg&hRd
z%xS~VP803`I%plrB0gsV@CbnX<zgyG_3o)!-+}L+C7U0fON-WQXTCOYR<>{KOYzE*
z4y;@f{jDDXHM0KrY$j;7#jg>E>k+hqN_#x>T#A6A>i{~?Cy71(t#hRcioHGX#UI;Q
zfH@ut;*YAXR;q>Xx3X{C=7LRrQ*eVostf>{tSl|zBHm{2gQ2c?4hw+{v33v|$EBy2
z%`_Sw#<oehI@cSCoT~bR6#$vxJ1qp50U2@}{Kc}U#xU!1vSyrSh<cO}Ub~$McmW41
zE32mtety=cPoJC<2b`CTg*PLeb@3olvi|`oGI;ajhiqQ@jLanU2|23W0lpp<9um?B
z;wq-sf4LHYS-PG0zqwN2*q@13iPD;YGh>n8*8r&PD~<`06)!LtHcJW-Kn9X;GkoXu
zKlj9eSM98~UFjC!0yO-#+OZ@A25$f=vv&)d)Sh1m)GzctZZ}v$H)wAX&%tbfBXb|C
z5jh_;uHKq7roN`4dnk(4gu~(Ew)p>wsa%(PG=MHKhA=@CYUP6n&>DpAlR%OH3<Q@-
z1l*YQRyBTp-iH4!NyHs-TzsLSsQ4aAfBrmRvigzzfl!Hiv&T1l3=o!S_^dzPxOuY;
z3fy?~FK0qRY#^Zge`^5()l^hQ`qD)Cz4pvW8Mhg4CZ(ryE3Qve=GQpR6ahAVxmw{j
z+)5y2G)JCXamdQ+jM?OvVZ2h$EIuLOb5&K<mr+14nExBH^&dtmATlKtoW5H-RhpVg
z*ZwzTk+|iCLsW{+9ZotN-Knup0{-@?1}?O@J_wMmJ)D||N_p42IjaCiu0!>LL!8CV
zciL!#oHNsb{3MP&-=`6<{{&R`2>%jU5KadUNB^5lIR52{=Q1KdOnPm68`tK0&?eF(
zK>QXkNWeRu=6o&}%>vSod?cNQwsu@hObmz(C`%f^mw}-p1(4<dtbYU`%8Y(!jrav%
zgF`_0FC(KDxs)eVC1#A!hEi;^8wli!PE&$APC*bALe2<9%{KF!{<wmpSb!M2h@?$c
zScHS9bfnZ!aBHSf7la4oa?PR6XORlzEjK^|HEHEB$Nfa$nHEA)lE=T_c%uj9-KHFB
zb%Z5AWdmm0qjfk|U7`>qbDjh80UrpHZKHqqk&G%Bb++GjJOfd)%@9*@)RoSWbP8Y2
z7lIUogGm4^&kwjUx0-%Hgg{3wtv;IgqGizhU|I)=*RCOt;vYiZhRs8^;35{)V;J7>
z_H00Ua6XZZTT<R-BsICI-80|Fz=Uoa$Q)8IFn7g>76YI{jsA+hZ|8trj8%tsK1k%N
zf0S9yEo!)aoqC&{2Gt>^IMoIc9gqU$oS#x2s9BiDj9cQ*Ncv(1aqnNI<*R3?RwS)c
zQ&!`4gEFJ<j`TxYT;Tn^&pA*H1Rcx0DQ>;|$W=rD#x7nUqzxB9!{CE*(j3>pRiVrg
z$wVRmLV&#w0=uoYLvn-E5ssDsHv6%6>@`saF7nBL;i!<`h-zv4S<`8*2^Pc+Xlk|~
z1XNO2XB7iwC%>zZnUFLfu!9((psr3&U2OMlS69~}Tx0_$0YdU~2ubIE_4H`B`eX3V
z$U+w4GImdVBM`VyG4wg<)d8F^hjaaZ{+LG432P!+PK!9Hlnd-|B`r*9Z}6)|ZAX89
zd-FiC^>h@mo=SidKZIuZt*^l?u&2mY{3+nRab+vdN&ZG^Nw{Ets@6&O*|T6EpaM(8
z?>IZ$Dh843<x*Vidv^#}mzS3pC{(~zSK#!)q+V8aUZ669*vSqA8PZ%gaGkpqOs(m)
zniYb}9w^4LC~C%s5U>0{3miAn7C^WKcn~lOTo*J!SON$$^7}iI4g3Di&cOnWoMi2P
z)SjrI^e0x4np#@b{upmi7NP^fniK(tp-6hsxU{s=(PCXYP^ptu0Cvh>8Ysfct0C&K
zJ+jyt%kO*atf&d%U*<v(KTS5c6$8nW>feE|kd(4!M<ji<`<4Mt@C2DdK;e|s)Sxad
zE?+kO&bLwyKAISWVgf0uuU@@UQc<}ERRxcc`e8ew^cE*l04xQQohjjWx>{1%2NG4t
zYNRnvNah~r$IU-Qn{(i7lUF5o%dsgb4W9x|3#oCKLZ7T;*iBSe#N+~{4J5t%q<Q}8
z8J7SKo$p2kXuG?+N3?7Ai60pO$EXI-zX6a(^j<d)QRzRr3UFa_*b7~=j>ytVzv;SH
z|DaUDJY=TYP8Y-wBNN-Zo+DW@A^%l?I$ch28WTU7*4l8H1*(qgO5f;*J5OF~{!0y`
z;Un}O%$D2pZjQ-){TjbFL(&99ku8o*yUz|bC$w<%+<j{Z5bOgWkBl{r2!inq!~l&F
z3vfgW1eq@++1ZtF6drRlXOiOtfX}ghi3(U67jUMy$;g(*YiQx^$IYPKP{BKG@Fs8!
zSJVOOrT+Vb&~4)`(eY`=THuxr4yT;DSN3gyu!j>YaEVaAwV^z35Rb-#Od9b`7VDkg
z-*hAxyZ#1Q6(A=I64OdJ4G4r1OS_(E^sxFyCu_zC82Nq0<o`)-7{q85ETr>*fe(Xu
zqrrHA6a|P(pW{fvsI3khUWH3Q0P(S)smbWwyLSe)j%h4vsSm^3nE;@$X?=buDry*m
zOCf?(1`@zL2tF;$&40~1KLj}G+yo?55Ws1@>32Qec_~as*8bzirARq>dF2=&{vZAg
z=NG>^lPRtb<%NSl5cK`&tnmzRETv#HHd8eYK(W(}ZvZ$lAOa{tbYfzDRTUo$gr*I^
z<lr-I$jg)e)-NZkl|i^5xl}LWv0Y-(ouDIp2-3Q?KSC+OA|s*yBsr=&e*RS2n`<U>
z`&W#AG~@9%#-GbmPSi5K4~C(}1N^HDxp;0#38!i@pJEKijp94tEGH0;fDeefON)!Q
zMG)Ofnn1e@3lDee=BiEQ95ASs&FI%Xz<)YzJTa<kQVQ%zaR|(5mGxk@$Qdw(-)M6o
zDTH46Q;-lU{wpsvfWiUNwEdhYiJ@${aP-vN1#~6N!}zzSre+LCT^FOd{@Y|Qpe-c!
zP7@V!QDS?cSsOsul<Kw;k}@nE^F|FN_gaFOpD=?5NKioY8O4QB_p3htmn=z;=nTL>
zEdzsO6&9^uQ(!&=*RM&+3jK?6ed#*{Od^7YzesvT#APwm*zZ6`95~PF*<IEDAWdRg
zS%t%kQ9vxkt@`M2QUd1mAX&t%2!xwJWyuBJ0B0M;x`;6#$kM>e0In+mYC<FGZXZzt
zT+zgqJFZRvsB$!dj;WKiP6(Uf&xPt45<&j|8F35Q?|=BD`E<Sb;IMqcKAG1n*s|vn
z1t@#Knf9|%S)fVFm%<vMmoHzQYz=5N;!cc-jg1Gx-8==C1cnaf<~1!WEKo8sGL{(p
z{REgX2JApPpag<mnWQ~&DS<WCD89ovvuIFANEc4UH55v5A1I#BfXA#iDl;$vUwvKs
z1(dAT^Ccc2@uu_%ucSU7!A+4YACN=I86`?URX|8K75L=tAAtV26n2f%97EW@A9|0D
z+QhdU$HmeDE>3{SmQ|bB+S<m3|Hb8qc?hPBToOQn3{IY`HGx%vnpD8NfV%_AlgEZ`
z>uJ1W<CZ|R&G0jj2m3c4bPl`#Ig%=W%&AC-BkqR$CvD^uWOPE$)$*VG*DMCk;ou}f
zKpOu`uVsXQ?3@Gk1v@Y*nc!<0K;<<TK=#b{1CV}x^{=x29<U=Y4PeVF#DKJZ(oLLm
zE(KJm86-KLkKm@P>Pm3O-%8R_{WsfEIFFq$wKWEW=Ii4U@zlniS)Kn913pXRft-CS
z+1ur=DR7&V-EbxZ<~9AqEEBi;lX1xxu$3~=xHb*42`^iNq7YA9A`)a4y2`r{Ir34q
zr|7k;GFouKm2f~hzD#ZDKp8%Sg%#Ce!FnY_aSro(7R{|pY93=wAcHC5;5C6$Uvp#q
zP0R=|6w*77nLR20B6qGdo;OfigSNn`XvDIrlj%$2dd2+Y{Q5lTCPSDA_}@UPp!h4e
z4*kz-ZvQXf{(q%Z_y0?`1%?t2fV2}XTVYG*0Yq((=><xd8}Mm%062Ol|9TKhP(lpc
zH1Mo|LB@0}6b?!NsSGE41OEr;e=<nKg7iULbo3zS7H&B#UbYgX0Z#P(ChIyb<1z^S
zpNq}hn;x00Dxd!PP(bBQhchm;^Be%H1uQ6`*OG*8<G*o4j^%nqBl{Cp8HK?4#Z+1K
zGqnHwNqEqYvyW(=O}x`fmG#TLJLl)VdAYgT!tP8}BL5|M9&oXr5ucbyJ+7{%=D2eJ
zY~(lkIxk)WH`H$QhM$B<WsR=D%kP&!jHDO?c4H=Mx@UHJ7r4w>By1Iu1_aUb^K<qN
zz)?Z{jCt*p3~-A8$_S<mYwQQXhDVMl<2(S#&BKr@8?hi?1`>)oNoi()S%G_EJI$|x
z0xVFF0FxzJef#zjDcJDci9Ct@D_YXhK{zEG%$R`kKbMftf6T=dCIEF{;b(*v@D%(G
zlY9qj!}$|95k0VriACyYbuee_B+)@21>kGeohOzfr}d!F4Ll=W4t*WBdUuEZZ;U!n
z;m^LJ*PNn>Lt<BV8;ms2K(>bu0Cko3+;%j>r-cB)B<66rWcOG;_oe_wDOp^W`8u&u
zSe~e*(z|yTei`a6jd^m|xXG&MG8)R+Q^=)9ZwVlQ&bue)ZG4do(Z=2y>}=6ZvqgT>
zj%;=sp6dk9IY5mfWZ*i8FO6B^b^PPUT1^ZL0$TT60zxI?+d`NVl`AHm2BbmM-UCi1
zv{lE!o|~O5^4*T$wbO6CyahF@VUo&citdeAw`Yc(+d}mLk4z?;i<jxG6>WzTcCN|d
zDt+$S`$}X0zXG@b0U|p)oBb@pOZW|idrppe@Ra+Ox+9Tsq}FDJkozCc50X;IrH(g4
zq)Gg5ytU)tGdwB&;*i(D;NjNSGy3yEXzTsP&g?s4YRl2Z0aUGlHgUxx4end)9x5MN
zc23uctMjk4@VaBK-R@;uySA29f1xjws@7?LNIq2u|HSuFbAJp!(-@Sg3z2#?8v|B+
z`)UO#acfKP2V`l;ggw;ADD}^uKk%P-tO`I~M-T>=5vhr20+CYBa#8<}R#)_=V5gG`
z#p+<WpLcC0uG(xs3x9tasm&PKDsk4moJj*l{5A>nZ|~!~_4!&LTy(nWmySslYyqB}
zhBv0L6Y>6i^!t+!8`QL{F!V4Gkun)TcXYfSplZvUbtyDIUkX<tjyOB;efV=Df#_v*
zJ<y!kJ=|$;SgKR{I*3&fLaM>hgM#lX=%VxaH$X`5^xZ3*-qL>GDBDh%$6okQqr6e}
z0eGjrU$klOl=IyQ;6<#RD6g>~$lZc`%IEoeB%^Pf?CncRpcf)Ufy>|L>OfSyv0pR-
z`+UQrFpNX#{AF!6+Vl7z#5mwsyamOr6J1<q8{lD+iIHVQAQINTGOtAhkAkg-AH)wb
za#m?>(43#EP_mZl2;^l*6(<-_a9y)wd_#9+h<3n0X%kF%G5uO?hs@5dzg$@i`+VJx
z8T4)gRjk}A4X}t|IxhCC?fUfKov43L)_8$lyS^DE?b352?Gn?{_(=J7_A&2sPwkQ$
z+-sRX*kEA8-$9T-3W6XJmJ9pdi=4vQp9B==x$_R3&vIR&w3p+R#93>*#V&Nc^0po$
zw6`_9k`!-F*A~@P>9t!t2s(;cKw0!vuq0B_Q<e(Jm;Cv&F!C)rSI8$dV2&YQ$fx9w
z>D2Mu)~hhr<EVCq0K~N5*ejPg%j(PD%g%B%y=n#hjmoPXO6<wxV-Yt&6%UhCjy_%Q
z*;Cr|43^l=?BZ<<LWRZu9W1GTC3bjC`_-#*vMtGneumzUS81+&)tQ|6nNL^zu}TR+
zct}qc5^yZtSfdoVF}+AMY-(l(YCif+Z%F^1*^2S-QE0`741`f+QfM;NY8jdY=oun=
z;*lHBuCIeGiPRijvjWq3{K){eKzeFm1$p2prpnR7%=VLMqF2+fc!bv=j91ycdl$L8
z^+{>UeSg<4eDR9{?}e~;>zzZF7ydA1R~{C;Dt8G4dd5TZWe^-n{(dePLJ<Oq_}DpD
zZwv-h_L@M{JPMT?&LUPkvL7&5!tW`@&=#(k%6^^@7suztOB~*QqfRR&HPvW|Y)e`>
zOTukSh{3)1@^AL}BhSW#<C@-0@<CPe8j-lv)M@lhyhn~l+-unF#6#3RLe~1RVrx2!
zg&Nw40C#xg5=>4%OTy>GcuVhyTu~Dhp1<~b1Iv<fzkIgeOPtb0R9SnT6DsEwJ>6%j
z6GweEm)7o*cs9r3q`T&NKJG(KSzI#0=-b_uEMbhati1f)0Vr}gy2{1pIKH?)UsckC
zJXkvnNih-(r21T^*RbyaPHF_q`S?;Lwf8GJEaYh*-y#GcI@dEPv>PR1pRq?MLttG+
zt96I#$pT6OK24x=Q94f@9p=Xu+yzd~xC9z|92Q1uR8vJj4aVx|5PKBTf|nR_Z{a{`
z$wE^<3EjnBShZ;GEb(nCAwf(&kdWyL9*y*ZTO)c-PVm=;n{rHQdr17*;J({oR9Ivz
zKa1L6q-;BjyN<R>YRTZMu&~JDnwe)~x3M+V&!##Z?BFFTmf;ryYU|gLM!NLARvl@1
zbDuwuP90744!%5dyqjG=bCGEXWr>f6PL^%xtreaL(|Rua-g()%6muNJ@OP3UCTxc@
zchzl&yUrHbIX<@2l*aAL7bDeKyBX%{mBy>bH+z0)gpe%skG<Qq0-AMZ``(GKi*TBz
zbWl*Hp-pR5)#1YSXT;OO)!UtpiJN4FFduB=%yz&M<3|^du_gj8MTf~7=|NYX20nX!
z(1;c|U=Y(0mH31xwOh9B&*xchbZ}ZtrD#P?v^hErRyt+63H=c(y5tcT%HEs9TRF+M
z7&P~}-_}gzj{U-vQkqEGv?N7f@g~x}N8>=8<tO&V3m6$=3`=BUS8?x42FeF8#+ErH
zyn^?Th|J3IO%b1C`gOlfxZgpan1j#0hJ*i-`s795Fpcw)`1;Ylp+!U_sYj9&rNl{<
zDbnt=t-}?r+zA(dWZ_t+942IpsO>5dAWcdyzlV6yW9#~-{+(KyNFk6cibCHHQbpCK
zKlcgyP3sVwd{%3|C8|I!&LsY*Bj9MtbeDYUwtsy96&Zx$4cI;J!y>gNf9bEE?Hi((
z4Aob=frqf-*w%pCiypLi8q{?Y;2%=2tR0e8R=Z;xE0S_<(U0zR1c7N}=~p$z=on0*
zA0RpUZHIyI$RfFFh%6C6(TRVaH*B$YY3}6~mocb$dTN@Y?X0psrM{kgW4j2VEfe_c
z*909gjnJv|0qvn7ZVw$aS!>oQ<1;H7XXui&cuNPI&fRA(dpIaZ&ofX-ja9XZ5~-cA
z-2uy6ZzK(i%Lf8wU$8_fl)Q0<;Y#7CRke-Ej(Dk6^{!W*;FyA)abnB)_W;hd4?AP<
zr&X#VC|{eQxL<|D*7H{etOa;q(Zgp8&Oq+zuHxh?`+Ucf9d17t=bS&;f6ycS7ukh=
ztiO!<5nFx|8GygOuDGkdl!)_`2dP&IL7@`=_ND;K@z0+>E3W2IsEaMPn)x0v<~?aV
z4l$7p8`4~B_Q;*GQi>91d*DIlAtx0MHq$_nr!NEMBIhi=`o1AUhJ5o)@719vJcEKA
zE^ZqEDCqk}?FU#6-L<DyrQd1`7u<R+UGLU`jjAE;jeaC%p5?t$J+<{1dE`v<bYs7F
zsHJ%*wtocs@kN~w-Gy-OR89QFMHPC-qzy(f@P{H3uZkVzdlW?Fe}MRUA=u%*mSWau
zyJc5x07}ScL16j$n}-|l(Hn;EZg7prZ(#8FZ?w5uP?JZedreAaRL2Kg1bzP9dKXgK
zHPH8lzh&ii={P2N+PF5xbkgoA%49v+QI~yo3|bBi`4hriID=?RZCP-mFp`tgG&QX&
z7UGXp3=7}yx0(5@T33u`P;~|S>JH>J$E4}7tIWz;@he6<?^arPk4|QO&4t!5a;B^_
zq*Z%mT+Nm`QNc4<nrpH4>iHqTm8<xCfzU!%dsgv-%GTBW4apJG<_=r+L5Z2;3l-RL
z28F@Fgq<qG#x;nL-`_ENJ&;^|gA!bAj6pq6oSB4=o+~pTKby4Vmy4H;M&^j}&!e<J
z-Dk@q-LT=l!TUoySr<qm6w<!+^lHG=844<XQCokpARl#Glriw*Jt+cf%1B|7xRXBk
zLEoKRyD>HGj8qagjG^Fq!Tu-aLS0{Su8WhogHE2=q2X9Y9T4Mx;`nJ}HFI`BbUf{7
zqp#L`c+(*Yqf;anRl9QhDrbe@`eVR1O20cv)XCT#ouD@Bn=c;i^?5K_cb(OC=Et`2
z_c<bUhy5jlIr;zaE$=IKSLD^~4h<;h(NEdUYDi>NX`dqZzk%Lmf-^-zT=mV4lnr5M
zzfQd7#3Pf^*`pex*)VoLquHy?`!hstYGy>2RA)VgzWLgqZ?7c}UYIfZKlhn?ADb7@
zdYAr`R|s?nu3|=Xcaxo+{Th!B3{{+|U{SV5=;u-Oz(6!|OFQFj`n?t<zWk9-)-w~;
zUeIZuc7`0Siz_vUq}a2(Hnp-#pZ1;BCVBbKW*_OF_A3vW1;jFVRjsWQGf&!&jXV)$
zW0APfrWDkd47tBJhxe8Qk4EYZsL};$OPjJz#`=;=eX>ysI3@x9O_(PrvP^IaB;nr1
z^s~nM*piQ^MQPPKeadT;xESty!;_RNswO^@$i=QC6Ks?%9B?9;<~F9U+Cq3qou?5N
zb!456RdsI{StlPbMxepjLAoBc!i&O1f06N4i1Mt*MP4gM<wS1d#Y1uD)+$W*=M2P+
zkcpCzJ|!Jf<xjOVp)W33^U)%ZTvmhkCKg2BHO4`|t?Yy1J9OOqsPztxvJ`cjKfVlI
z#vS$mvi=6F837_n?587gOm?>drfL60b(EU+UoQBUEE>hhxEI5yV)M2oNly6At2fZC
zjDkczC6!}F?MJ_o5uj&Bi_Hb@Fb+c3xraG?kgF9L#zmv3QVoz|p*gU6Jv@yR&D~p&
z{2FmwkH!3u334ipcT9!uKamMmn(+ac{+qm0RUDYZUtTv`t52Jt5)XXlFCW2Q=4GOr
zRnI`i0cdr>uQv-{JApdV*p<gMdYrbypCdr*fTKC9OT`?hsV#RIWqp%_eU5|8cLAdX
zaTF0PWU(rzNDJ)qWzpg4-43p>n(;++UEEk%Yl5*rs|Gl^M+a?u>|>OFSmT~7yd}_F
zIX|NnaHX`rxLD0mH$&B|U3n<ZM0pe|35ZkWih>ravZU`EdLQo!J3imjt_AktWhBD@
zwUY!(1~R9T`!;KBv9<;;m&cunt_3&R{UWa8g!p*=dQi$pb)^s}(AB}ZIf&I&YnhjV
z<Ed&O6+2ZFZAQg+V$>NK>5W>~nS@&xyO6jS%THCtrNbR_0Gzcol8tU5rt?9qmGJGO
z_gzNZYrWMm6P%@fx?y1=69SlLY##aOUO~-_7JkX8<`DSAqqP~2*RB!}X_>$a>{R1X
zK|!}))B-9g=w%&Sfx^{4c|VliaEd8<<`jQ!bhjkxcApT2j_F(0E|kKOSK#c8(#cz~
zQ-^&jqBh$@Rc3ecVl1cldNq|F?-E+G2nBxaIX`5~S+#%ujnMzizE>%K`23O2^1LYL
z#c@dpN-T;m88ane)|=|X#x*M~YmdKn-!cZi7NL5$PuiAQRMaKOs3gjEe41k-C??j#
zlA8`{Q4frp#YgWye!Q0|<~_Lvzd_6Jrd_>**^mo5^oRQa*S2PBk4NfdlG6L8U<x$?
zsZ3~F3EKz=h|Ypxfpg>s+b8Gqf3%2Z%cG<m%RI;edRUqKvmjjlx;hse0Va<k!I`u>
zfh12D?>ekLc^h0SbP#@xK-cYZnG)-$eA+A`UacsKVfSR#4Ly!DTUc+X@vVs6d+fC8
zHtKdPML<tph^s%+pAerqe7EotX{kk`?03Q;CRnqqna@|wF*IM<)PN$rNIQBpEze;a
z_1+P$zPQg(g2IOH>cAh}6T7)T-Ckw|MGwDG(91uK3eqbP7;E1DsQwGHB4q?yLo;04
zq1EGrKBd!m+aWtx<6zhTU)xJDxSgGoLrhNIVXMQF6`DFp<ht^j|I(0m60GYEW~^T>
zG&B*v>Nui9r+vM3I%0lzd&n963vUnEqO+n`%>VYfJfB%;xG5;#kHP&eKvv{KCbOHw
z#NkXM6iXNyD!I`6VbA5Oj!w#bSS63{cx={ljo?v{3|C~TYHPK)^0xiSz-EZLqFCA3
zqgAmDMV6O^5YF^638bvT%lKoYn7U=}SMAa<D?S8$CXcP0U01$I>{n+Dr@80~Lqr`-
z)M@4IAOZ~M2|ewzV>z^<q^5m3Pf`{!iu<pY0`W?0Xr0{eSXth#>rS<Oa$}d^vh^;l
zU1%By$}{2srJ7l>c^#UPmYXgZJ$*Fy{=?Sqo4%#@R*YNA_)725BXYFh^8TEiR--Na
zXr~OkuRnP5HMd?x$L>l>I8cVW`zsujb8S`#Vhf>x?p{XPK1O<6&=$;+&*Ph?_#btT
zA276ib*S^J`Fc~>5fpLp{+U>LSpPEA<LHA_^(z9PSv}z7?AcV3O%}&E8Jx4%&xxBi
z*G0|Vx4cd@8kT|Nx$CjabzbGjkQHhDoym|*EZhXNZ+Q@KCr}<K@jmF=QLCd>+qd4_
zjw2@Viyy7Sq}M=S-a}uOPY*45lt}K7wo?+>_1aX9ug*!lnf7D#R!Q7GeT&!ey<P$4
zST|2NWNH@rDVjiDsh2xnpp9T->_#kx$*E~CY+^Jt^=1`5m%kMW@j`YFHq-9sjQyEC
zONyj-f~s<crQP;5ZzJq&YLh25`c7u7K{>ip=|-{R8I(TIJ)K=b<H<z_gl9F1FXNS{
zl6%EeeIl`V$51OKzUgTA#_U}E2Opl6VXM?B=O+biqfIRbsu#sbStADO;v>Y;J9MFM
zg9jy(Te%)5r(xa4{B*BieKUp~N98+rIR^W(pF&Qr(fzXPP8b4}B(WSfgasc;Dd!NE
z%MUD^cC3zJC0b0P+Y**|zRY7-Lql!UkrG%-cQ2R7PlGGM+`jpOwTQaJ?|Y`f?hH?B
zKau$UY{D!&g*b|1Sw!Sm@pkr>k+zqu1JV25yxZMuarIpjCeX!Zs*JH;sHe%IjWd12
zv8?;(#ZD^*PNv>;1EJi(cOuz}^n3D-evYg*wPM{TYXOfMnT12(p-R6mwmr1;jIVv&
zUsFGs_MmmNbWBoJ_91C#HwtW{`uRH&srh?a#(%B9dRf`gO9C=Qj(UL!ium!a_8yx^
z`QThgXeb}!)qx@NO^R}($JSy{!__3BaK5j8)KQl*J$z|a<f;zpYv;dOb{Q?jX~*9<
zdq!<KyV^4Z%YZyWjd!6tv!C8q3{zru!iprI`j$Scy{YP`3=KVg9!M2!$@61<>sh6t
zU4Ya<|CE}2o}Qn5e#C<D%wyXA-J7#&E?W80oRjr$0g0Gjf9oIdyc~4K>goWKOP|Pk
znfYuK>Sg7)5Zpt;BF7ymrfJ}@qEH>{<UgUg;YM)d@v+bDGteotz3Y48`0!ALyISCp
zu47$(Cck;bKCFs-c$9)#NBdFFUH(ctHCuMjFfulqcGaGWmN<Hgru8`mg$=ZdBqWs2
zCoL!TA#HWz*ujRVk<N-+VDyZm8Bm98TX-1io0q*BLp?K)?ZNG-E({bD#1aNURMCM;
zaFG-Deh<q;4IaYai3g#f3O50UM+brw7dc;)^QoO2@!+7yT?JX@+Xe6L|MC*!u32j7
ziZVHmmuYOpyZcz+(P5KA{7CG`y#VI2{nk&guunhcoge)UyC|8g4ZgthO+BQ3XEMm0
zMGwxsSSB(tF{w)7iF~?4SZp~H#Krf)?x(Mmc`<V(N74Ll-NWg_JH=LWos?~rimSW>
zn=wC_C$K4P13`jMdJmY9Zoht5mM6h9Nu%apwx`F7?e)fT>KXW`+fYyl3$jTOmCwv^
z=M-hX-Pnc^CVUln_}!K?3)7Bt>3zevx7$r{=M*ha0>#u5=8TMJQz7)e{|I4uWH~6y
ziMLnVfEw_J5Sq9VpaRbEfFEPgWd=Q!d(RxYHF18g(@*}gk*9=ycn$SH_SOmAY~wp9
zJvV<N9j_uN0MS41yiwVGoq_@neeZiFRNw!$JUn&6A~pKHNMu)qNiVmItgP&%=YeG2
zL3^m3VLLMZFfh$ufL&35KbyczRFHeUQBRs^;!<}aD9~>M>8I7<{HO_w%WofGXToI~
z1xLLQa|N}l(;VR**h&J*JUIY-<nr1n$!AWIx5JAoK1@}RzaV!a{(7dR$u`kCYL))+
zi^3$#ld{<<LrqrY?5EEYF!N0QH_&*e-~h)x7M5Ue*jH0uzcW`c)>()yH!m*<mb@nB
zNX_8kmUeWlZGnEF(C0%347Smr(J@}M*S0zNCG<NZ^znH+M$aM3zsMMjDfiVSw?J+#
zt|f5v&T9VqyK#E~@Q~ZWF3h+(x<H9ehY2EIXktL|l&k+e&I(FMS=Zy)jF|7y1d<w>
zb!p>o8T`+Las7(8S3rOUM~$tKW%S^r;Ie_KDHAyN@O8`Gtn!3m{9{UrvYgyy2*}Dx
zyV@h#|I9(X!Vi9+Mome+QLRJnHpw-a1Vn~{Nj1B&4uNTF%WB4;)_kI)qaoT*=nV>r
zc~_9?4gsayW4<7B>luXl?tU;=`KQr)8GAxrTS}RY-7z@PXS9FpWN*hw#Sz>Q@zV2)
zUP*40=Rus|`nzDM)02Vw)&BeP<NadKWMvms?3=yK;ElgFAHOyX<c1Qu1rlB%-o;Sf
z4t~{D4pNJsOG>T<g3}t0EV|-;a&DEt)psP;7Z6-uVNtI@;~I!)*4EXL){cPm{zqc@
z7)aI(p<ko)=LTCar^FLOXUVYFmkKh;20&2_D2zKX0!0TMrsdZe7>w=CGo|0yraMhm
zy$@mZuN$<xb}4@&h>ZB)B7Itil9ii=aC<!`(5dm~D}M2TVCh|TJN;7?u-E1~>7OWJ
zOq#$Xg=SnnkiebMV3w3*tUdc-Yb`B}Gx^@d!06Af8W>h<sLdfe!wp{i$gq9ni`xuO
z=?C8i%Z;X^JPT^{LN|tJ1J`rpcRY>!OAP*H1P52)p!5ffcX7sJ&f<7y9+aK$f%Gg<
zZ<mYT=qSNVy)G5zS+m^H2E#pUcYR5K6TYE@Ke3M|iBh-+#QCIGJ>!*QhL`&A;XPms
zHy*zBzD)CCC|5BqF79J|d^`B5gxmbKx4=o)Cygb6rk>OMXCg?k$Wp3Wwe?O^#%Y3D
zB6-`ENZ9^!yj|FtOYf(4Bly85?_S2yhWk>gJGfm&Y%bl5Px=A~<>KCaV1Wc0;{siW
z7b6a6s8a^s^qf(rzMAgUYAxPBE$zolf2;uGl2xHZa}0d5@0^>R|JsM>;a>QlR{I3(
z+)xEAl2gjn7YXTu<YzWDQ6#r?ELsODER_FZ00mwZKK;l~@xw>&Xy=Gm$x*@q6LL0#
z<VG%?5B|a*6aGUfWcbV5_A?FD#%3fWB*LH??Y{0Jqu{LP{j@)vzL)@V99^vJ-nF=W
zYXxx+RsF3e_2NcWR+c3Wl$h`f-sKpB10X*^8U7D&=wWRjtIeiE0#so5p04FDE-k$p
zlne4Z&mXn5*qxhQN?I}g9kGUSW?;*^APN#pmmzBE$TNw4MTu$>js=<iB%^3BreDNY
z2M(OU!I=u{K_>8HO#Mw>`)b0DOI`7tdS3#&xf?$$E-tc^>}-NU3>Hb=_nqo0j~j1+
znT@uT=At`r23h`(F)`oIPxd>zyNg{vOl?_$BT*1=(~HVciSIZ=kxiDoyz2dB!rt7<
zieLxS7til{+D0Ua`v`z?H6{*@a8M=2=k^|l2Sc*AGB?<sJc$J*F?$C{lYLN_bviBT
zd;BO$@<NC#<PA`llC-mM#>{v5@+zSx_$dHl5)w0TXb6<wjT&W$xUqvHEau=)FY|*3
zNqYafQUF}JI+RBOekkD~NCNdF^ONk9Nq`!XG*FxjY6`7f-u->}IA?I5qT}`aVMQ00
z$Do<qhK6MDiwodX@+Zjv30$paqu5{MHorPx^%5M~ns!6N_~)$5&A)&WVNR<3qvtrA
zi~gAh@uL!BATKN|w4bQB90(36x`NYxDMs=C>iB_F#^vHsC3klLfs_bP0Radr_~H^p
zsT7tFR8Lje3<rO_PQUsPbm|7@txP%G9MVW}(cwDDM4AcV!o3bb0L#Gr{lD_T4*5o>
X8S7MHcs)qbLKNlHWXqnuc>R9>TTaLC

diff --git a/docs/articles/vignette-07-backends.html b/docs/articles/vignette-07-backends.html
index b037de19e..56e96e80d 100644
--- a/docs/articles/vignette-07-backends.html
+++ b/docs/articles/vignette-07-backends.html
@@ -33,7 +33,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/articles/vignette-07-backends_files/figure-html/unnamed-chunk-5-1.png b/docs/articles/vignette-07-backends_files/figure-html/unnamed-chunk-5-1.png
index 0cd10a097059c9d4020f159d56d63fb7a1f89fb3..e2b2745c17959a1f9ee503a83e4adf9ff7ec88b3 100644
GIT binary patch
literal 16916
zcmeHvX;@R~w&q4u$`a((v5-OqssuqrL`6gh5G_<h1XL!OEtzCi<|zcr5>Kg!f`E*{
zQb0h6GRhQC3y?7~1_%(4AwYmIC6I(9y>?mW+}pRi&%OQh>HgD~=iv$N?EUR;Sl?Rj
zdf)Z=>qQd-iLZBljUb4G;knb75k#~OK{kB)YBOAM43;T@kImPN3{Jx*{BZQE@bIza
z#yM+$1o=w<`YST*Dj$L%-y?>nPnn0L%?^d9^#-O3g>(&d@x#rRwp`k%Z`iBx?HNXS
zzNh!S#7U3jtNY%DuEiA;;_}<>^p@G1so1uZ&86;nd-i(|m(VMcvaN@`jgNX?+J5@*
zR-JG5iHrKL@lTl)oo}zVlcv<sBWs5y8|(<Hfw)PHUgOX~y`0Mi@B`htDrRNH5JCU2
z{g*oU`1WT6hLQO%8xO-j?tC8wU+vo{B90&@fB5PuT>APHB8ngvzuFM8)u&=2@xe}w
z1I4awd8~c#op}^8{Bx-%#dW$f%evUrxWE9O&fM@HHuW8XME&@`_x6AM$S=`0lF*+|
z+h5sN9rq_K#$4bn+>KW>GZmbSQJ#_EN__X|`NMGEVczg+w)ugcur!}<>R^PUx~y)q
zP_m;i@Zq5MIRR&(JjYaDzajB>Z`dojpL&YP<IU63<9T@Dd7o-d;>!F`Rj-1#20A(4
z+Q4j@@ahaXt#A2$>w3>UbWLJcNVwIeMrH+88b7JdX|0#0gryZs?z{?f@+l+a^RHf4
zSii0Dx1To@OnZ+l#dsFxvi1$_ff<kb(YDOXdiZF_#e&AE#m(&b^q5_mFKTLPdY^lJ
zd0ocgnUu=Q`(<TiOXZyMIW!q`2Kz>Rf9@F=NYV129Dj(J|5$fRR6TTQe0NaaPo;8O
zVI7~cA3h9F%%!!<?=8`#j*70`Qz`u8F1Plyl9Cc<?#lFZSB_~zth^(m16tg`N!?YO
zRF!y$JQJH9Kg?p4%H=!?nlj6zyt0n3Vv;a98DSnX3`SRX_meJDtgb)3y?}KcCf>^l
z6M(4MmL2>MEn(#3<U}tU{czW9@pHYqrzdr`W^QiIy5#yLI^~{}%5mOqFsBP`1qB>B
zVUW-vWTuFd);+bjmCSKvQ*2iSDQK)=rD`@m`NmK=rkua(YxRQW;@D;($v<GY^w4+s
zDfG`YPFzq<oam5tX!U2Ls4rs5T^1%BTO5Q(RA_jQP_8|@$1aQ}Y8aPTR))5v#j7O|
zQVAhP!|~oh8TM845^giiI>c&v8Yc3o&}PY;bVGPMl!tdn@{|ufMrEoN4)<-teZA70
zprU^5_rJE4&6RYSKx>JK$22qoeSCd=$xrs19S>)w4zFa-NNz!pl`C_yvYdS1!AAxe
z;d;28GwN?mY3m^z-=(pM+NJ_!!lU51Qnxr~r||IOW&Fj(ld|}}UiMvs%|p7J7rTS`
z!OQ%dL^>zd#K;MUJBy!7S!YiqbrIhj*)=fW*K8~-omMASwk$QU*@HzP?j%lr<C0K<
zrdu<^pxwwWXTG8CYF}<9)$$n1%M;Is9|tpQtW0(mn+j%|J(rgnB(&(6qIo>Yva>-`
zh)G;@bQ`{2b*QMSu<+1!Ws}(m{_opO$D?=oyqxiZZ9!&_+;dz0=uf7<#SN!%l;u*n
zm-l0ZL(RL{F^uI!`N4R{v?CQQys(L3tH7+wk#9ctqrF{5&0q5&R*98$Sr0EH`;Y3b
z)=c{6Q_9E7>FMbrHJa(qxOGfRK~wUM3tAN|>arD-jA@JV>EA~*C3in+;Bq6_!(tm3
z*w;Tk`oX%wZ^D(@TVR`XIrj+($MY$xfzA@vepl9ZBonY$EBnBn<H_3*WGZ<_?HW(9
zNlX}Vfj*SlD7h;|F?8jQbeAAc)QhOs-OJ5WC?L`)YK7<H(vG<X&?8NEs&fNH*P@>3
zYmb$uC08CL=a-iU2fQe)(N#O9nWND3FoZZG%iTHhSi!)xDQ<77J+mr;v-Cg~b9`^P
zoEU-(*QJWFP2=o@Q$);%rBOG6Y22(vCVu98RyY5dYT_#{Za}c`@;No}AU~cr<V`nD
zavfe_>`mv@csH35dBZ|n<CJDf?x>_eK-E!l+x;Zh9&VyylVnDiHW*R=b6Pvbd0h#!
z_RYvGQ8CxSw>K*UT!y-P=Zja?#Y$EVp)Pvdj@aI?W8lLvH$Ki+xe!i?`gakNoTrkA
zt*_IAXLQ1N^}&3LxEH?q`q9T&ZLmj*>9ynw1a9=Pa-WDc(u`5+`v84gTfw#1A=Jny
zSS}+%m^(2|{q)F!L9ueqE^LoiteKfPx~^j7!aOC4Sv@kQwoR2O-x=h5S8qd_BDf9m
zOi#Z0&CkEv-FUZ2Yq?dum`URGh8;wEb<y#_1SxmB|L~e<S9ksoMwWa%Un;IVf+^Za
z@52oba!44{P{(b%wH4J$y*`eH&u4Iyck@zk>HZ6S;xk<s{n`kZ#6pL3G3?5_>TqVw
z(o%Fxzc1xoGEb9P*muD`S=sR6pyz2<S<N5vdx_)>K5v4iT&<1|uOs-@V{_;9bDfNW
z?;Lo`S*E*nzA&s<Uuy9eaOjjU@I9I5=PMb7t;sh}6-Lk<ec}n_Sqx9dipM=xO4aIa
zvTyMgZda;qIt5j%3nzXlpG{|&gqS8_R}=e}m&TcvF9i7X#G0DE(O5JruJ2c|eDyIs
z7W<|x(qVtoeFlYIjO8=i=}SsYnnL;;#p#HWza*6FJBmWgSU77*bV!%-bQs9S`rjbv
z^?T=J+woh3O?$Dp?{VQhZp>Sv#Umy0>K-FedBSycKd`2PU5O$S{8znY?Cu`#+Qv(p
zAWUDpF26n^wp@Q-prPvRO{^I6(@4KrC(s&;n#q`{-V4jtEa<Ndo4+GYb6TC_{;Ey&
zt%S~+*Nwcp?GMwTwQ8`nb1Bc@<&uGlgF|X@<lFBFr14Ih*iO}l=&0@lH@P^rZj#qh
z7Hdt5*AJ$7r^;=aZz3(`8x&Kb1n#DOi&g?%rWTvURH~^#<M_!@jbq2gw@ULp^?H>#
z1iHh==rXoVY`4Phqh8n9GCGIv#>m7+=A>IPZdMxqF>&uk^l*M_8R6Yw59G!YFqe~d
z20c*tTqyFN!y^9}4g62x%YQxc&xKt7HOc=YME<AS*e`P1>lnQSE$%t#?3B5Jf`W?4
z6i>hA{W;?w<OMf&9K3R|BM{)%y=TvQMFRo?#<P+fnjuJZ_V%(<SSp2Y_lSyZYfaPg
z@9gZ11EiN@O3;(RMn;_JMNxiqio8e$CfLUWQ1JS~U2H_#K#GQ^-|A2>z1z&x7>C29
zq@?(9r*quq2P>qkA@F6M@QB<PvNYb5tP|>oBH9RU7h%}TyT8aq=D63Dox65n$|dCW
zbnZ*5nb_LemQxrE2BsW-=wjHlQ*=h7T=+p_W8<E+1bF|;TPb1-Dl;<<;fKQ}niFJ>
zT>IUX8Z_XZqiASkWK>RZlzj`hk<*A@Z)_XJge;5{tb4;)r`*FXAijqBA0`!o0s_nf
z_)>v^ft)D=>;Q-R?5<ohrB=WyqHluEy3KTm=s16T^vbTnuh<)=(J0saJdK2_su6g=
zGw2xTjOgfSyP$p-DM1Xh{W2kiwvk~fqvhwEc+k42f9P3~nq$Trr$$H`0>fniA%C8N
zHgAm7J=mprh~kLHX=`g2g|TMrLMIZmAeT@n+6Go}a|IxCJc_Q9H9ReqOwyj(2lelD
z%LsL44+Yci6`=W)c-BMJA*sBPNiu<87_T(3)S^ZPK<vo432!;T2=Q?_ZpZGoiS1BG
z;Id4|at`IOz8mU&Zh8wGWYl>yd->02ZvS+%{I9+L|6uR`znSD;ibO*uTa%?i(frxV
z$o}h*va+&vqI*|%R%vNqu{Tj*Q|e*SuFZJmL6(49T->LcQ@QN9RLD?+Nb^1U6hegX
zmo3{XHKpKN^D9XkLfEtY?!LaJaTk)0?xooI8^abCjx2t9rtkmzHwNZDKf--4qYiCa
z7;lQl)Dp1x$ZV5zo6=iwIdf+-X-jYBsoSWFGk4m7-kAa&oEs>u0MNG7X{2W1B!^zT
z2Up=ep$uDEij2BzfP*vuEJW8nsZb=clpM(}meKUFBP}Hm$Fag?tpE69gX}kN7B!fu
zn!XNaHg81t&^}C*w5rXM-rnBbxRnmW<LQ5?37Q+YCyA0e3Fbiv&96nZ=PB$}Brk8)
z6Xiv&hMOlKf*Gq?o0y>u2vTRM!uxhgu>xT4QobYq03_!`Yg?c}4!s@uIbvKvewuv)
zp1!!>MQlMrbMR5Tsh$>##e$SLye$m8t6+{MDA@A>8a|oz1d>f~+<rY4RxV==F`<*m
zZcEvkHE6;Qb4iZdYjo73`(?F|&m=ayf<|BWG;bRr@Mqn$ATzyuAfbpiUzJl+qiYC%
zqjyH|sI>Mo()Q+yr=a~z4BQO^T>-?t%!=gAB~4CFwmF2G{{@=THSQOQ=)PnyLE_Cb
zB5)snxa&On{!Tj4*2ZROwwP?`stq^2d{iX5R=wQb^3z?}-q3VrlQJP~yUB?kDE7hj
zk?Ujf^tG{A^6$THSJ6Xxp#X#a;9Gj1|EjR`T%NxV=%VvQ(_bxw(|E&au(F16nzfv^
zWQaU2Ip|9G9N>L*Bd&T(mie)i;*jdx)kbt}OG%&;@OmwYsx9c)!HeJ923a;TWD$gr
z)3lZY%b7(jhahI0KA+XOgNCkBX=fJV<LSv*PdV;wV`FQpNoCIvE%GfVUKO|Ju?OLf
zi;0KqMF-1$X<mbG-z>B8<!`TqDP4H_sU{b`%zR(@ZC$Y&*(5wXJkGiK$$m7{Ubb50
zipAdDCYz%0@WgL^+hNsDZVK2RrEHop$l<SZmem7iy7MGp6iClSw()C}A)|OPrlvl2
zW*I?BL^gLu=i_BbL6~@FYqGk!SKn*J)X%~1f7v3V;b9&vDWCB?R!>`1RaHtBU5m>6
zS%fs5o$B{)leBZg0~y0O$Mw(B9KmWWL4{a-6pgDw0e`t(4RY?=XlH^0rgx55RjqUy
z6<>Y#^8*=mB}2at5}MAx|Mi>nSiMT_Ouo9$;M)rXy_%wU^nT4JxQ%O%4dj|^hnEuD
zDa&*+8L&k!rxGs|SOW*;k_}TC_4R&C)armI@s(})A;Y3;_4f|iR~~EJ?SFyJHBzhT
z4#lBd;mWsQ*Cp;<M<E4_SgoS9br{H}H#9U<&w2YA#l5mAJpd~hX`(&-;+`Z*oeo`M
zYt+%x{7rTovU`f!4#eOJ&SYv=z7<ii3EYEv#M*+++Q@Aw`ESvwcs{F82iqlCH#{gM
zHMMJSFm1J(uAA<O6>`YMA&Z~IcOHBBoBkh@kN|^{!f{tZ+sYtVC)N<a)?3ptn!!VX
zRuG|=9tKw{m}OruM42frz#+dz*<k09x?3m&*oK*X-KIxN{{u(;lrQ@BGHE4CEeqxY
zee=M;z|w>oUMtBlGE|S#WU6FX`1bQX@!cm5`%ku_I3?SZkfDd;I<K!Set}5eHS2Av
zjI*(^@gwW5qITqd{rWlPaE(6zDf)0YvyIDQz>Y{6#H|jYkipE+k&vC8ofgTRMz=&<
zz!#kH(jBQ?c)!)%!((7JU}m@`T-oq^*5yQ1+r*zfL<eSUOIQ4GSCag>{(cvXb?%SO
zKhMrN3BG6^b(P1d?A!9|-)~pCfWEufZ=(6ZcSo)jv(keG_BUleZyd7s_3)S(93CFd
z0QN9|B!EFCNAkIW%bH-!nK@IIh^D5d8#vznK#F~p?#F0}T}OR~tE(ZC)`CS8oVfZ0
zfO)Yis~E)V;G?9h?B4X3iHL3v@D73X3#%^pdgRRg9WF2h>0};m?}5e#GVvAu^vTTI
z+YWf*vHWRMb(qAo;NeiMg*s78xuncdMHQ>U!P%Ps0jqJ|OOrlvxajoNOFP>2GEt{V
zwg7`7>nvYrcqp6N@!Wupp>a%RW(05VZ#UIBA|@YkBO_v|=@1jT9f}&-at|>wuz1Zl
z?(e(~@o&6N*5zC(gJ@}Gl@v@`PEU7yZV+qY@9&?4N26j`8Th{6?jJ&tH97}rH{Nf_
zB@&r-RpW6+OKrZ9>gukoQ=KG-h{gH{W;m^-cD<#xdh)Mcaj2buH*aB490?B}K78d+
zqm9xxDB!;L3Ppjt0R`HBTeV3)TD-{@&!d8aRx#}C>w7A(>v%w6d0Cljl6_DUcx~DZ
z$JOC*U%k}|viFV-;#05`Gzi0O&b_a!Iir&JFBmiJW`iqvVQ!wA*a{?BypnPAM3Q61
zG8qeGA&kX~rk!D%ERlpO>vdW=mletHTFbrfp?kJGLADO4+m<4i7HxUem0Hoa5KH)6
zR?9%CrwV5JxDtfF%qtrg66f}b*s=G2LL0WQvM;S?g)UD{yd8VU=)DLtb<y30wu$?G
zt6J;F#|o%Jg6_ybylZ)XXX|pGQ>+Ul5*4go+v0%fPLEd9DcBawB67#BUqq9M3+dV#
z2C{+XapNPmB~9WW_UMJKjovM$ID}gSoP(RE3Jb!vEucVxH&^b)n&}CJJyi_*L44P7
zlzGlyi9kE?jjdE2@w>xUPi_|f?qaSrT*~w)I}nA3TxGIhmv8G>EH(o`4h2V^E2j+?
zyhCq@YbI@7G1eS<g05X_a4m1BO%9K`b=`i;p-tSvs`+N{n$|I)=*<V3^)^&HH^r^9
zKsnHG1zU{D1CJ@GvrtzJd$#8(WFJUg_AHWd#OZ9daVpTpi5{(JjI!v+@{A(u@n+kr
zbkVekN)rU3NfrPjXK)u6^gnn8`?Gqitb#yfLZ+etB?;Hb!WX5brOPbHZc-GjtAf1c
z!e`&iTtIP(RCg9a0>jAHh3pJ5x9F#!(Xn_Bx<8`4bB#m>!8>p?ya!53%F2Ao=6W3S
zEeo2h-O?@QD&UsPgZA2+Dt&3B{<k-s=nRLNFn-gj>C?l{!p-2@lXn95N4b$R(%Edb
zViWM)er+0*7lrdINjXVb?PCVJZ=Rk|LIe2MQ<o9b44sN)9?G}Pbmv;^`U<*}_iQ%x
z>d?PMQK-@sHTU>`DL4hFM5U&tF0U_>tvtSfzdwBlcvtX*&5wT2*DjBRAh%We(5t?`
zgDZ8iPCWR1*OjRbPJ1vZzOoWK{}~wV)^}p^Y089%qCdlg#uMJH`-dt~Kmtz5EKU)^
zOS)&0ouA<_x3}CC;%zd(!-;HkOeiQ>h|wH=YG$J4a-RXM`1w@LmH9}ZYnxyhglHZ+
z_Tu?-rxwhsd?Qa!&+=%9L{aBeFj30JDNphXD?RQrmM7a<ZZ^tN3ZB6;nVrp~&1?S`
z_ePQE3BzbWDpcQET?nb|Akd_v{-N*nPiS(UMMoq1y?wl$BZ}6wkvMcT;@=(qp2w`k
zZ_}R2L?dtL=pE?;y7H(go$eP|AM_`?02@Gl(<oQf&t9!z;?Q8Qj2AM6{P^Hs(Z#d@
z6s=aX(s?26#1T(Q;gA=vGLV5rzVXRa-*$j=wr@mGL;DLlcy#mqcIJL(cXtWr;m)0Y
zu(|2VSdQT^t1#FF%tmV0pBPA?t`?9kP|O6?LS`A&Cv<(Wu}T9)1?DyYR$GrWj`!O7
z4t?f+ki@3xOX`-`>ZA6Wl>jJuZ<1YH&FV0Qo*wLAqBZdZ;@+N^m>BCf*UroXfNgzx
zM4@~r$X3G@7BZ>;7BUtTO|YldEy`(+fHJvOT#CkX@T#E@P^|yf{cR(PXiAY0Hj{hM
zq56bT;z2_Q5Ga5X-FD)8ctJwOH_*RI@B9rX&`5nYRxTZ4l9k_=bOr1<e7-`N26k2i
z*rlYjv}Qisr<Tw5k@#D_;Rt92`fG;5cK~t?`^nIrmoy-Ea>S&1#n?m4(VyP`@~0oU
z#P*Luxe0|>ouZw^aO26IC$e>8jj?0^Xy!3roE|(1xzV_<knM|l$-h6UQ3tE)NP-1*
z`ILD|Yw4qcI?yZKJvC*|i6rjKJ#lpRCR~IRD3qSA1$Ggyh|=63-W|~U`yw!FRxO>j
zAunG7G}1yXz2A+HuFLLIZ6wj{?FC*q9C23IyO^xQ=>{{)yl1unp<eC>?NoQjDONYp
ziMHIn`-H2p((Vj#n)jI?2CwZ9o>ZX9S;V)I>DW~=haEis@syaXdA++xly{bq$oex$
z+&+{7V-I*@YVisX7QD;vibnZ?bDZEa!-Lzk-Op?r^kuC2_pME?*T0>N7<u7TJk^oY
zSryD|me_)HpHAAaw%91od@yVUkzKcqze*##<M#%rL^{4<zI0ysI5`L-4uu#_<;!kB
zG<vd-NLFAiu+o{zwv_?y(abW^H69iBu?5A-^=vVu+oYc?5~&Mmnl*f%0eRc`*=<M}
zst(VbmzAPA;_Sl*s8p3QERq=nK?sD`<e&83;EWOEOD({M=)L_b(lJ}MD>-++yd<L+
z5%N%vI|bsb`JW*Vut#>PPxB^eyGO8?iiVMk_Z`Q1G+|_6^-?oxihSNGS-N(u{_iN;
z>CqwBkq585{@$tKfnnU>jSmuNqJ-w~icLLMj_EhY8eD*!N=W|$XFmg+9jmi+4ibX$
z08$3z<0hE7l6DHT8(jP^U<qMqMX?AI_Of-kkA2W^DE-xyH+zhd4z(R=+}%|nfTTRo
zK_1rCcp0{hM3b&Cy}M_xfxu<yU9BP4XY@!d%npb3UWPLoDD6!>CDJ#^En6X<3N<Zg
z%)AFS3S5KB=ih(T;>{NOg9Cy@{fKWkl^Umr#;QiFU@<MA(-4w^s5x4LM1IkmiTXUj
z0|^Gf*uX(}=+oR>6Ah=$nJYG=>7{O1YY?7~<Xrcw&nd4am&K^Lec&eLA3K8Rp7{{n
zYZH4i0f5ebG6#Xz6`iqiUep)b^$-^OL=Aa61cIWF{_Hgz^;Xjzx>DBVrN#U4XYR*K
zTkfqA&zrf9oQ#3y|IArP>{4PDNS5Fl{>)i~0$xsG;htFO%~AJk76-GjlDlfBr+nkq
zmpM%XflX17rKu{0K|5dja+!WNyze6W-Zqr7n&NCsPia82#Hp?vvS(S0ET+u%19DQM
z#_yNaTX*Aw`SsaMH@ifgMRUI~yrP6H-?4t48W`)e<aReQH%JtD^pI}p`eBGr!1~Vo
zl7!Gq5~22zd`*Ie-Mb`S%h^2XrYPrIDmTg}&vuvhc8MT8lXfo1*Uz8*z`;xJy(=H*
z)%UcdbgvmL=>qR+fwCx4w#gt$WZx6}%}QP5T-LY9N!1e@#Ot<AR7l4w6AIA_rixJ_
zeb01?uXPz09&q{4FfhD(EAleUXVVSvm>GBh=X~1+WGHCqO_QEN{KJRMv&_-PxEcd<
zS4ZTex3>kNW>c~cJejGPnOU^BOlJiSe75!58$Z=Y!)Bk;I@s@hZmEqKr1mpLJ1%wp
z!uc#`+NFe4N#tdEk?BkrFegWYX8R&WMCCu`=&^E9&V*@)S^Ryw|H3;lH_&rsG@5BU
zke4yB>B$kHL4NJny-sYyu!Kd&swB!@Y^7R%xbsox8%M{|jK2{_yo6w(|0&QAb#!Gq
zJ;KkXXo)kP>lVi@E5w(Ts>)Ov9UWb8?n}+nR6t(O!9r>dhd7~0q-r4n(i094?}Gpk
zGXD}iUpbv`cUS&0#~O(mP=PoN+;qVDulMt2W@b7e5&avC*fxR^#(DXL1E0Coq04@Q
zHyZ;zzt9Xtrng9dK-K-mKcf5JE$!x6c(CZk=zAS@pA(v2`qZ#`A)7DRSqyju7%WVb
zn8eN}${8Vh%yLZHE*Z!sD1Pzyup3R~BG*P|VdMWRIbjm_O#kjAo9c^-zsoCKrfz_{
zQFsCQq2)9OyhY|7aC44FuGU4NZvXO^*yLq_&;UAAEH>;a1}R;2_!};?!R);X4_q{a
zq#+DA$mGL6{U{f)@D8m^fmVq;6l&=j_8Ua0k=8!zK$z;s3;Fp+(45Tt3|w{$8ND-G
zPK*3kz#U1c8cu+u)8bF=ut0kgcj1F1*$Sw@-2tCYO%GMAGDyDTA0H3N0aK40$eKk2
zL(Iq|lx)fYa*7Z^OPt?^5}l<G|AZf~>-u_Y?<6Ogf_Y{q(XsZ)`7DoCv|my{L<xZk
zZ2o!E*0PWpb-1*0+J&~!6Xbe#clX=|(7}<58bFcUk#>0MPuA7krRnpkK?l7m<wDqi
zqk5OSw5MsIi1Xoz&wmvM*(5NKI5(9~BvtMaMQM$_5YZ*|)~-T%4nYyht76<X0EByO
z50DFj@32QJU>?K*O@iU`weC#kHzSg`DSdZu?>1O7Du=%d4V6{8pm`(a=l{FVhalAZ
zkXz&vd4UormqUGCL=*ug!1c^9nx;@MCvA8Op?2AGpu`>88}g#&*IE|XBow!6=MVV9
zEhiz|eH{Q|N-+eJc1leciyO}*zyhvjqd53$#$|+@p^E_l$1rZcS@w8}XQ83}5JO%^
zPCMW_I^Jeu6l6$!%0e<w9cC3}qPla<@{gkt$6UW(^IhdcqHW%hKl2*2<_e)IpdbIf
zF72P9ZtV2XMNV3Q<&C3&6!uE&v10*zM|%kgiGEnT7B0LPql8!37*+zwT-vH6onFbH
zyS3f3*x!?IbZunRp7t)9U7}m1Cyy06aT;UsM*{;*B1DjV2SmhIt}90_yco1!%|Axo
zoqnu0OQC7&#QgX~uN`-S=*lw6U*_OzO1=h!^ji1EzKv6F<U&Sn4VI6;J4PtA?d*%e
z+s_R?O2U#1NRe$8p-{)5qWL|{h7`1XWp&Ws)Iqx>_;Oxddl)6lXSNQqP*JHqnJ?3|
za>bKEG_oa}?`A4_<Klt#y*DQbIoSO#IhaS}_E!liR+9`zVVGR#_+xRPV3?7sRsP^p
zvdjkpx^qfPHGrgPk_{ewj25GS_NO)b`hzx>HuMMQ^ekJX8Z00Uz-=-lyMk$gU_UBY
zy6G(^A3J`0a_V`kNu1rydMTL`H?G1SkB8X<Hc}VJx~8?ou`hbrk?xgmr(9YR|424O
z#CAV|k{^_)YLv%E$c0TkQ*8Q=b;+=eAp{(|DGA8Cb_#<+B)4hO(f~6bYRWz-_op^V
zLm5UCGI@hZI}nLP+UDB57d7%Vm4t5&a{BXLnA$H%-;U4y5HC{WusXMEKy%WTJe=N`
zNGgdDn~?fosrv*9>oy?!UZpsSuHTb{7z10qX}*e?p}$Q_u_>Ez+u*@>Z~A%vM#W4`
zNzsN(zns2I5Ju1$_4lRS0CP)m=c4X0&MiH5+)FM~ydA!`ZLUdGCBEJ>UWJVxiY*E7
zY8|8vJwC|Z_rfN*dQwxXv{xVMGJ2dgtZ^AcQ;4Ikke8d{X3Q#Pi(29oB_vW~`#L(_
z^fTrQYkvdo27=6(E7jpwmyuEde|ndH<JMIxmYdF-&jMAf=vqahG2UcYA-U46bKr3`
z-L1)tk*B>KEQSbrbJpekpwXh#X!eEVI2MEA_P25@rY!(K$ZddmqRz3b@`*$;%H+3G
zfWmXrW50>DSS$ZUwo%P7Y)`pPux}v#<A7USdieq|SCeIu&5R#hzP1z_tHgH{mbf8E
z$Qk1=&=z=}RQRp(w*~4Qng=~K;eScdfz$yUXAmiJ>e<~Qusjc;0#eqiMo}bOd9ps2
z(NvEwbEn2;5a%x77m{npty<xvT9oUm#jW{s<`j~8!^DxyOpl#)DBb2iUqPRQz1{|l
z%42k-V4+%qwK0^F{RgF2N`~c>TXTKI_!FDLt}~7C0<WHItfTP2U2l&ZE~&qiQitlc
zd=kmDzyJzpU%?)s`czGr5xC~0^$9h!E>h;G^EoKle7hT{oOEWg9xe=(JbvA-{0HHe
z?3Pi{|9D`&P?VY*JR7G|qQogzjt?a(x;;GCFq@{s_Ujx>TnS7HMv&uH%brn^Cq|-n
z->4I@we$7zYDGDDV0`jk!P5?bP0H-y71q7-3V#gHZI=I9`+Jfc*w)p=sx254420Kh
zdXFI87urBy?)%N3mbUJSn_!pw=^yPIi-Y=t=5MZM(zPRQEOkl8ORFrLE86QBPb)NK
z4^{AZ8<4A2<6j;A<S#LmkalDbXYIG~Zr!*J%wATOreIKIApX_Bql4ZiBo=l2=Z*=S
z91<ma57``6rVwO!cE#AoA$d?RoBeVwk9N#gJ8b>dP`uh~Qgv={Ps2telfJ#~Bzbi*
zd%Rkc5(a!>ng4%P2m!?U4~w-WO68(HUq^ueZSAap)oHZcO~j|`$7<)jB2yA(mJ%=d
zw$liKmMHOaz!-$J*_Rs+Q%{3m*_L`EAaaO^chK{+DY|w>5MGN_^HrXktHAdcMFe-g
zh?rywp}6yLthw72P(<%4M)fPX0%z*n@v){z&?@Yzk)2c5&Zi035!^vZ`csA2R*)oY
zlYs?_lFjghyGciwm9GfLDvn~P?)2(_LE890=<<PL?3Zy-f9JB=n%Xf|Uii5xiI(_U
zjxm~NBeQGJ!E?DqRrxAeHP4?H@qnZYy1wDh;NNZO>AB=su2dWHx;Gw>0_KFs1#j=Z
zM-hU!aSW}Hr3hGeTf5VvKA+M+MaIB@r|#LGN-H<-f=KuP$#&~|dxX=^Z!$BAXVQ|Z
zdAd!%D(}rwrRgwi$+`qz{y53|gkq|Dcv&y_$dlF^4~I@ol!pE^)&O~ES@pY)_l<JK
z+q8~_QmOH^+Iu)7O@>a)@wa37+R^~_c5irUw=iy2F;-8rTD@D{k=uNCcS<*ZG;U=m
z+MByP;E7htUp%tmX~s`;1K3Dmb7>Y?H#ML{Ics1je-2l*t(`E}ZC*^E#Sd|3O+v35
z7vuoID%@VTBEBj-X3e^G^!kmZ5z*ALIj!Ha7gH6t(~c=B*Sb?u+CP~Ucnw=awcTs`
zZ{Y5jm?oa-xoAn#3<r`sHP_4&1EYDZzqtEqSoDY3b9#$S&TTW!Bbnoa%kj^Zy2B2@
z@``=_>TqdK_mlPO@8lh8d3^(rxgoQkc8j=DLj)BzrZCmtrXjA>>5$uQB1ukst}%HF
z@L5xPL|V&^2p)}(I$HYRSBFy9IRq;oI22L0%@is2UY|>Dav>Nc>h{#|0|FE^aTses
zz>*d7<?qfgzo*`A*+FfH#AigTj5Xz8OQ##G_+pA#mqU4Pg*i?0;flVYMoSfHiV^}Q
z)VK$v|9q9IH8tYQ){K)<i7Qo05^}biTwqhN(-h)TF+gOD4s^!iz}^qh-a~GyGWPvI
zjz#@=^AFtnkIp;%Q*h;%qY$4iq9-n{hA#sgn)Gz!+4#I2jgDB~PI%cD$vB_&V7Jy$
zR9yun;6Rz<)$J(CweN<)esv+4Ho?P^ur*?_?RiPZ$GO5o7V}~OC2oAx@`=G%fj&tC
z0=a-)ZskLZ0!SBCSUL5XJVC#o>ffy_e<8P!<bkEv=irmKn`$44)mfpT1a3atKl8dB
zT9Io13G_eV+5p9WM6B?o$GTF}h_JSX7h6<WxO_rTsV>?ozUTXE-~aq_zv1Iw*Yb;=
ztS402T;IQ=^!()KO4<85Wz+7Md~XA*;7dA{9vpQmD|JI^H~ad%M?206v<vs`+4udP
z7vfW`b?a|=2C7@P{L=0+-?zGW;x2Z6l{c{*O2eO6ZCVw4M>36$Nqw@*m)miCNbKHO
zu-H1wD;0h&Grajn!xH=|m)ctjZ1Q$fG9$4jb00dTUU%Bt+n;z;F}MGsQCid4(b3_}
zJN3=Ko>w`kX;ZV#X(f%SxLem$4?I6Piheh>ZNkrnQk{60$oHEhlMZ>S=I85v^X)^z
zsSQZfA&*EWcgx!2SR8j-bDe#IqG2MLkl^QL5HvU!)y?ww$Z`y1Zpnk{)svEk>Gk6#
z%;;a_=GrvlddrLnjGjL1s$-<xfDL7Q+`XmPwf_B}e~0qXS$zZg$KrV)gwwbm{Nlzt
zgeTOO<*JKzDT@hWk<C$ZnYnwd7c5J2kg^w9`=_S{S1KHEWi4y9JmC3!4JNHLBQ^39
zbyEi%SXwJ9w(*zUpL>?Ge7_P+RYn!L4MB3a(?OW4*LJjX$cjxl0zu)5&+boI8uUZw
zl^ktvjUInZ9T~?oS_Bz~5TQ=@ljw@?Irfz8-2M?YgXK!+{=9iw@Oa-CMQmYYXxo%P
zBue!!=v}*Chc9V5sOZ=06d&VIc$LBAKvm=9z@h!QjG?FOz_!qZyzTz2TIw|Rj_a9W
z5mHbYXI}RKyLN4WqCx*_@qCk<m~rzGr+Df8b4`ozgI7gSKV5mfW02syXoe+!s=+4w
z@LVtXHtDc#R?0vtDbmj-#i4*5QYRm}^%(5wj*onkD?)Vx!_KOkS4}Fiaq$$nfTfk)
z@H@X#qAv<l2cAo+dB0na3fVkdJag+8n_%~9c6Ia)c=G~-O0vwCQr}%PL|<I?V{JP7
z98EUb&ECkYe74+g)M*;KB05wg{p}re&lFrd(Bs^#Y(n<+(N&pff9PFF*=3Z`{Ier1
zO?R816Sc{Pq^lxelso-fvE!QM5vswCuN}|M^7>bNrW`w11On^BY)Rug>*LI45lbt!
zmurU}O4-ZbGx@`k^Kx@r-o*=a(RH*Uq$aZieJuj`myNxHWc)8|(+Lfm3DdL;@IHy&
zj5&UFYFOrQ0P`M^_e`PIdQsz+kgg7+{ddKFBXI5HGYt)m3H00#5c4SIb+5qI%q(7a
zbx^%(zS6V4cVi|RTP~hUPfynpVuVi^_!~F}hc_n-mnLE_yz1l1PPFT=cy34im_vU2
z0!?Q$FR|Sf6>I=KxwFDwt|jEmuiH-wdPLn{dG;xVKQU{U%-~(p3-7*}tmS6u;Ss?V
z&rDfY5l6l?fl7!%(&s$8z>&SwgOTwso?JGOXHAHE4eOz*=y9r8RDnX2NYr~YdsPo>
zywqQ4Mf2%r#*1Q8Xzih|Y}p&GE~A}AQ?qF^Y7Uj{`LcjK&>J{ZCXb||@-wpU7dfdq
zG#3S$KvTOHJl8Itw?Cl5T>FW!2dx>pRlm_ikm1}!c5lS;W2Q4bRqv_fFuTK!+il^_
z^dhpaAnP)^leU>4G5RmoOS^$Ewo8@(&>M8(N2$IfEfBEUD(D%hTrL;tI}7*x(cG`A
zghXCa>xH;u1-9wSlk5H&>wNVJ!RwZM9T~*7(`g=YjO0v!_<=jFT?VO(&dMkK<YVra
zJv=?TK*$IAaBumbpV66iQDi;WYB)@zI8isn&4ExZ=!4S$@zyPJ#{S_(4&ITft7^%~
zPG@&NrkL74N>b8&ObU#tz`ZkFwhZ@|QWhH_Pk$2}g;IvVL8XCQUa{1yLXpQ$XC)mC
znEC{V)57_0&^lAK>V+Sn=%q4(1R{5M#`nlV#v2?nk~4@mZBO;~b?uOrC;z5#-+vsR
z{zxH6>zMjO(ta{&F+%U~RdO2VSrbe@2InC*k+~YvB5EXWQZZQ3ABf5DX{$^fo@#Kr
z>@}BO6X#_$cO*(Kdq!js!c@zls)JT`H_xCovmF%^ElF-rezZ%LL0y%^TD^v%x<drW
zOYHE~&z~#M)*3W>u?`@g1OdlzNDqUI)VNkT;QSe*l<~gs+DGbcgm}7T%_HuBUd7FI
zsr^|`_?wyW6$IV{mtj>V(32z`#Sp&Zc6d=dm)}`93?{OzeFGcKzBe=wa-Dre#Jyh4
zh(v&d?)FFNP2WoFCrXdE&b11=!YPhKEEZVAU9yn5($+_iljTz()qA~z3ks)F7!IMj
z?!zmdKP$>$XCFmG#*QtGZ3<R4^ud0vTw&JmsfFQ_tWYg3UD<V|?otK+qo<0t0+L|W
zgJb}MhidqBB4S=k>u|&ZWjJFVYG`Wam><Y`_Uu^;UDMpwHWe*M;g4aM)#~xert{5R
zqR4^FbxSwxq%w<uw)TWML;55rxDR)6rdc97J~^?~bL%TP#HMn*|0z-hIhoVQZ#S_w
zmU6J{QJl7z5#VTB!z{v)OyxWe1PkmoO7222+aty4+1U#fcI{A;c@$WdSvc!r{rc*=
zvjhr`1820LQse|wyP$nKkbp9ptXL@0?&`1LkQWgIUtvkKjjv7D(ufI^aQE?TN|JjD
zC1@<!{S%Tz`9n{`4G%@E7Y*WPK^DaEti0O?VvoHb%iPW@!<+WHxySMHbT@tKGH2Ol
zDnr#XxFHoNjrj;_CYq3<FbDnT!srKepoeGSKw$-+O*K1|gq|n#EUZw1vXb+0hX`Uw
z-xD+Wb(<8btsni8rpsquEE(M^wv)8Siq_oTBD=LU7sEJH;rj35Z+*&Hcn42w>$W!{
zYwCFdnGuPe?(P@au477A;Kuxi0x1(O;!T>-qEe_WcL9-23L}ZUenvP{t-Pul`TgDb
z^~K#sS27>;|HNDmd1s%zw9u~=k)*%On5ebKd)vB&ODS&}NuRkSq64hDUD$LMSoTm=
zFiMU9v+1?O3}ERXb<uiN)FYvcl6r=`<Wv_nzqaT)e{6dY_PFwFq?eSEXl8yMcgJ-k
z%41(ri*?te@~_ArK<Caf)>kL#u0!$=M6D;7VfLmpBmfHl@Z0u|4~=d_5vHk@tziF)
z^U1X{_2Nv-zwwl}cLlY|QDa-%$ik3RChNzp_v^D+JipnG&&xOth?jLN_b&&LsCmEo
zAB1t_LCpTx8g|IToqBpPfh;N8D)otVpZa*iI0m6=ua%tNP44P2=^8_HVIF`|=A&WC
z;S-X$$@#%IBlGLm7S)Q&{JraYeY}0!WYq)*rDuh6>q~uAM_=5`xmO@jR|Q}s>U|z@
zF5WO4rx(xWM$}VCH`f!S9JfamF3Mbm@b&jo#Q#K8{SVG6|9|?HXN3cz7jXv6!-qd`
zS900k>2}UbuzKX4HFDRYaGlBR9zaXQ^!vnm#C~$}#$NXd>i736ZJc_%SzPbsfWpcZ
z4gHzRtBa%_x#tFX{iRl=22*b~vyHd)=MR{zT)54%2>)5`73qgk9&a@yer8vm?x{0R
znkW1!2J}}>qcs)iNe2*fCr~!Iv=Ksf=KbnTJJD<lu;;&rF#p$A??xQNKd6Gk=NH<8
zdI%N4<%_4G8|F5nq9AUlqJ?>*cjYsH6IXShc62;OHFzi?f<mL2>@Cpym%qFhyFqlQ
z@$38P?gW(4wL@u1QN1UzhCQIe8IMPc7Q1TkLZ1|nETEh`0#R;43&)@?b)U3D$WJ@e
z9IBf<QPpI*9is+9LdDc?_bWqxgf3K`N0d=Q916wDZ9f1@Ad-guzW_g@`hzXhh$s=Y
zgzL*_)n=N1yA~?mYnp;wa?&c{&*|O>!io$g$SVV?MgjR?pd6(Ms;||-%c8~XP$+N#
zq|-QhMmVZJ_sGF3qJlrlj$rfh^IPDMP>P}!<5h9dsBkM<bO()9tbM*;uLcQv_!BtZ
z3e3Del-!`_ev&~HiW>)b(*}}LYAkNnNo)E=e4fq^Fbrm5`AOlg(vA^F(0a))=Y@*8
zOfW9>_oPB+-!vA{BEe=*$!mCTaFp<Olqk<N%PAD5)1Y`m6%Lc<t$~&n66-G(vK<O;
z;=fTi>)exPfu2d<0j2pr=trA}A4KP?PIkn2ibDLOW@WEoyB^PL5pw?pX}GR>1qv_F
zssvP;iv@b(r1X}n@XXimY%@{R_<wqgViYfOI)1lL?CEeIkr2f2jLGT3AFuxW-vE`l
BW?TRO

literal 16826
zcmeHvcT`hbx9>(&^eAvV9z~?7fGDV_s7Q^<QA7k(n$)Nu2na~8A*e?=7C@=eL3$Ty
z5=ept>4Y93kbpn}gb-Q?0g}A6JzshEz5Bg;#~b&4f86mo1_btAd#$zCnrqJA?>8qt
zyJn~>{L|i_AP5rHyZqY?2okJ^pbec{HiIijKk*9ivH6a^?r-1|{Nw1<qQS@32bV3p
zAm|rw_)CE7Ama}~`yjpF&YSor&kP5>>Gw(D&l4s0+`cRj`yeLtS~SCa&y^~?4c8*+
zPIisu!JoId<%rc-#m{!hRLoW(T7An;dkmks_#pgT@6p!Jfs-ScH?HytyKi6E;_y&m
z!^H99VH6qak@Dvq?L%W)D*>7%bX+2n?M3h-S93E6^q>!Pbqx&-K`^sY4_2i#JX2fY
z4SOIc^uU%|;Fst7LIoh`&?5-ERr;2H?*{pam0OLxU1LQ1I%m;E3tkV?eW8B07pd(Q
z*g&5%u7f7@>)>ME27&DmbZ*DS<KUNzKi7lrzn+J{$k#UhH{G~b!~%mr+>g}ur1H5|
zWJggX>WJRS#U;@`|F^u`%>6o^>v~MW8imj6?d$6cOt#$tW_4~}8nx6_#}wu5Ym_>!
zi4JOGI^43e`$}Zwv2JodJynn<#Fi^Jnex2@-0C=SGnGW!N@=I+t+&7e-AhSHQ5)0A
zoiN@>PL388x1xKjW{XPtUgoW*GAC5hR!y{7354Fj<YB7xkM@z83zw_h{V%QKUSj0X
z+=#85h10~|z?eU0($!yN5v|J5DC)Oc3esT@DrxxMH`3MBE!G`yg}35R>hbA>RSd@4
z*SA0L<uH}+2wwO3Tq1glnU$5*?qk+(G-f}@>V%23S2*=}E_LY=1CM}N9+yf<nY!Bj
zE({&pnL2Y?rrwLz(cYP%Q~qG|Q$<yk!}834_}K@y4jLxkq;~;Jc(gtmPHD0!Bsz?x
zr6ghZ?se8Od~Mqkf6jfd;;<hD<(FfWmiqMR(^IOdv^Et&+uMsToVE6Xsg%AadPAia
z<p+-*J?iA_JUyuDIo?!cU2~GbU<{K<U0tsyTI2n&{l!*91{(}?($&?amBOR3Jty(j
z(I{o^9m5Z^E*wImG}Dq&XQ#W28-IQkGh1l0hT|9d46H>ddG-k+^rGq(l~k!EeJ;^*
zYrHYB#q!*==4h)bhXtZrD=t3dgdHy+1jm^ioIy43;>@5Bm!m4GsvVj7B~5C4Ui;8`
zld|WWnArHUb}L(toYc0(PeC2nF%pMu8cFq{)FSX}rNoKBu#h+{_73yXnbWAPORFrp
zigRzC+uCMP>0~)4S62fsFE1i5A|j&g&52A>5R$)2WOH8uW5V}!4J7avCnOd+qmAy~
zveQ|XI(vv;wj9riti-n!=cYyZMF!|Z*@Ip6t>CkXf<8Q7<eEGF<m9YV?nD+HpX#oS
zIb)Q9P%;n}R$F>O95*Lu#bc8a-(#^_#_p?Z8rm^8_Xu-xvgxRjIZX*+@tUu-P`~}&
zJ^zM;5O%tttw*X@1ecY@KM*g)@2F#soKSNe+RsT*DsA`DsG<$YiZDs&cc&K;MMmVD
z?$7Xc__3O1S65fBrfSx>ti9HiZWdM@>dY+%UVuyD(o>ssjI?8;MFQ;Nk^}GTot|F!
zj4UN|A#vJhUUASuW_b>4<)*UcS^=))7{ZeMTL^2&i&K9)Y`j$guORH>!wF+@kBgDL
zC@(_X$Ey0*r+3oqn`{ZljMc};QuIJVT*p~#ql%O&RapEQfBnfvcS&_Q@aCK)EaK<-
zptqx+!yLJ*tl?$Xf%o!0K0e~7?p>?qvREw63;fE95{A5f+;~A@NuVipBLuDH#uC;K
z>Me}&PnXcA+IL!AlG6)WUQ(I*Hnow|kvll(tKPGEBZSxXRIb@+OoX963%;Y8A?@HV
zb0sUq_P`d(7kLpK9Th&_<5R^}JqhELD@D`9C9V2Q-N=pt(?ZNJv(}d}_)-R;?IW!v
z0zq|;R7Hjr^gOwQ5b~}#QSEt?DEqcos2fQ0_*F_(LxC_}4{YxAWR+<la`}T$-7LpS
z#eJ1iP%=iYn(m4ZX7YpYEGf13u8>qPS_w32Y+?CrE0?Hot4QrQFcHZLHr+OQKB?|2
z0TPu`KehAV)g~4ZnReJX$G#`Wpk|>($&1IK5d%*_&@IuqWkJ!V{hyywbJ9v%{%9?^
z+DhN-mxAT4*Yb^`?629`JyjKL0SARD%s-_pD6zM^(zWqLMBg=b@EbRGcLGz&`_{6X
za|;Wx762cg?auexp1P6>P8u80^r8e<Eb#!2sha<{&Jw#e4mMM9Ev!N)z#vV)FlwOO
z(Q)i+c!Fkt=dlkXkAjynW!i6tNtw~hz#-Q<r(<I?o1~_e_GYE(l<2)Suqe$MBAWXs
zE1k`1^RHBWvCNKqnI_Vu?S%4Tl2@CANL!S#VQ^Pjhn8pXQ{f>c*<-aQ3sv3PxH6mv
z<e<4Xn$;*>avdLMu@uHE9=xiRpr98ap`y5Glo%sSs+tg9n@K?@E|sizjq+Vod3p@-
zX>~uAN#3WU7VqVZbFMaL$J-_@c4(CLj{%>B8ZW7xZTX}oHs0dL<rOZleSOf3ZKCfn
z!QHi<%OrAluF8FKne3X)s<PUIfe;O2OeT&|f_fe`a2iU9*`sLdIBc$edzt_UnEq<?
z{^HgYQN%(47mJZ2+$Y5uq1kITl-49UBxmZm+|;VGd;#+`N)*JzdZmF69(Mh%fy<`c
zq%{%AH-ej0Y11<NQ(|bgkoc+e{X(KriP?_HsbxB0OB{;3^S!2B3Ux|$PfeDq2M0B7
zAn~+R-Cgy-=X7bWkBZV3#khFj=kSxyT2WI)pD9u0;H2xP_}x8+LN%b*6pXY}UJFTm
z&_!o<q$bURH#dF9#H~>)4jV7bGmjY14UriwXe;&*HJQuE<|Ms9SkbQsHi+K>0mNSn
zxxlfWYot-<0=9YeJ`}1eaD22@uFM~I6j!lN^OobZP4HvbDJ$JZ(PU#7N2L3Qhlg{9
zF7lPpXO45GqKzDbf<PFX79*2S9tfM!IHemY5sxLD&FmjrN`$e-N-;`Er0s<J**p;Y
zB$PE@V@l>LJuVGtffxMm#tr`rZ1S%${yUN1|KQbKHGtef>~l=^`1)eGy0UU2(WRuh
zH{W=%41inx_+!P>>fnUjg|ROHqXEkUjw)#0dPQ|L{vK*}s774f`yLDo`}_ObQnfS$
z!6EzdY-izn5WCKaJBMuc?|OSNK^ql_<1*aw_gvwivNl_*f?8cVb?Q`BO^upkYJkJ;
zqn0MIK46%Z>x(M&piL{lmkUj@k>l<>syn8*_}_<oSX=jd_a+A^0Qb40=vFo)C9|T-
z05~Co?LlA8=~RLBN2{x1X)kT@A*f5?d;44|BOf0}%M}+FV{IvF4>Z)(7tnkrpRr6v
zv4{Os5yIdxt^t1Qhz&q4+IJis9ZiYIYe;0;A_0MSS$ilX-4u^o%6?AU1UwLWnkb+d
zyfRmW#N$Y~qgGW}B_+omZx)_rt#(Y)=_;;+6(x+XU%y&=Hp`pxVAj@qrCsI`%kxsC
zbc$)AnNN`V(qN^FO33;OU<<53DG<v%m*|i*SkZ%5-@Kxi;<dqw!%n=UZlUFBnl@IE
z4tj~nL?Sn&$dx-1IsidZ30N4PsiuT*CUp^spG7o^oPV^*DDg5kdZIWKC9i+-ZcB`H
z;8eC6vu@ED#uS~6>S@+#jQC&-aE1ClLCokJBU=;tiNIpeF}wtQ2mT|D_7BP1zi#~J
ziSGZ{wYug35t39(6`26zhngElZuo%3V(F({`im+mD}6}Si!Ps^3ahQnl&=N40dvc{
zY-dD1b+0ph=37~&OHh5NfbtqoFZv+MO%cBPUV78qqfOho!6}zZIAM3Gq=uT<OT;fV
zkL{4XH2%#lvY7lkXm@Rs)=wi&JvlzyI{W6Rwq&((!sZQNeB8Bpop(=#CF1gpbC(+c
zU4=Rw6-7qclT>LKdP!?JL7ODD2(F$>%R3t?cI4)(Y8<OLLB-LKi24XOrm!ZZM^aF$
zJ!u`kZ{6w1nTV$$`7ET@kVj*7e*V-&rd6H4hv{ZO4?>+m5An*#)T*kg<OiR&>ZRi4
z0)c8_Dg$&1s38A4gqw)7TD=2~AT(*biy{bL4tlG-bwsJ;!&_yMZQ$41;_v$D(c`$a
z*;b>Fg=j#;aeq3^kw@DW8qw>GSZVn2f^5*)=u-*L)v4UvjEwz&|0-JoXV`?`jUmkL
zo&`<v;+`u%OhuqWN-)y}U}a?^5j=lez$0_Z%T=VLq%10(O}|D+C5eX2Jw}~S0#D3T
zJr`WT^XPcbui}odGZT{bzoo9O9wi48BOn#vR{Kh+Z(kfMSWO4bJ0*4rJpi)6=Iwhr
zE4frvyqZ<D8>5O*7zJ)lx`Q<yi$rJxzltO-lPJ99BUV*KF13}Fr;Rk{fAh%!;W|_?
z|7<8?{+ZGEWM<?nhQMCxF(Mq!O^%|M)E?u{%kW=E8)>UWD;SmhbwI?XRj}dhez?Vy
zwloBydC@Txh%Ic#>E3+3C{^5ixJAXi>pr>gX4n?2Y<~cld<wG-C@M+goG@fNc*IDH
zWdf@pzyaS~nHx2g`UT83u<$#Ss_IVGpQX6txhDY=2OgQRpapJSQvznSyu4gOEV3x?
z22swH(orfbab|p`oG?qT0T!NqKno+}-<f`Xpv+E}xZHt!ymaHBL1J%hK6pI6CGgk!
zZM%<L27Elq{_9I|V7x9jko<0iNZprvUfV@#RR!(@W_oVd(xK2?gQR}eBJ8rgsBPL3
zTfp<H7aBxAVu0D$R+NH$D>N_d0IO;nn*sMd=u3Bb1+=L6xVShuLL)OvFZ!hK%(s`4
zY9|$vA6&4svMN-U0}YqV8U%uS3{)o`Z9CSKtmYLZhgj{3TxfR>Q8Bj-xNBBooq$>%
zP)Osi(+af@!4F)DoE0DsI3m^tj@V#M0qUh6HT2-Ku=r_b6Qa3{4T{y5aQyZK2STdN
zupYbwVv5>=>zz8H69(>q#>8dt3a|}%yE4J?vM6^j`1^J@z7&rJu~JSCh&1tih2}IZ
zW`GfK6+VVOXjsWx65Ay-2iE!|2yDSzI+CzeA^BBKPEOsJ6!UV_{crE~(^C9w4K)}Q
z1){Jvh`m0qRZmL|a*I1^6=!UyCJ)Ga4TqYfa;rX6uKAB0vL}H?G=%1JO=W%s%>y&P
ztJDh%3tPOqb-v1NII-BOnu=luA&G;c=#|f~mIQoL;WUt1piZk@X#yc1>mLdtE8N#t
z=6A{D78f6VvHNI?jo)x=M@PpG<9aEs(%swJTZsc=7w`#EFd)h?NQ%ph)W(pQYzH{3
zpPcc9%v;59b5DTPXv0Gfl1N?ycjG6z^_AO^+&-kd9<?JSDMKgBVf51zp#QYb43w)p
z7^%OSe75Y*sGk8GLt#ERKI|bcBZsy(uK~1xwL-PUHWe=>mhC{Qavi8}GM)v#17HO$
z@QzMyZXF4ZsR<Wf?3ULnC4!AQq-!9!GL@TV2vAaWYd(9>rBqJO#>NId5x|H_S~H^U
z$D3YT)Og+nGYrB4Uvo~)IdJO|fIe@)sVpJkV5F6B%IU^3cosk(?aWKz(^cC3b0czu
zG_6@i;9rHSBi0|@xvVuJcpl5-Af|1|1+b#4FKjo>fsHQ@SftM2`8=$&{n|3a0yxmr
zpxK&a@Wt1<WyjtD6y#c0wl8-R@KXsIes0w$ou}KU2IfE)OF^}}>RaQL31HnZI41i^
zo}{XqH2{z?vJq$Bz8yocYbot@0gDUIjDioi2a!maWl<SzO**lZxlI1QK-yvr0afC%
z1}50<sJ)o*Tw4eX!8LDSLkQl=m^dONSVU{7GnnCm)JJjAGT(cRjoN{Pm4ICTE~ceT
z{@(DyYq4FG_FSoCayMS@WZ=?dk!588iHmU|mb9<Tj~Sbr6X~p!J|dB3!wdvJgEP_&
z79eO{4Gc6%0Z#J8G0dSw8i4@!)p9sDD<dPrh4KA`a)&u?A+{|+C0DmGkecAQ5bho<
zi)Q<w*>6?e>z(wT?7aHs_<kZ1jRJamxyRU7>wt+^7-*aJ|JY|6T8Z29;U_@fqM74Y
zO@WhsB?lzamSqO1wx2XO_?{B7b{8D%QvaeLCRP>AXQL@c-y+}L>w3F=@2NwiYIg!a
zmx~N{w8k2v8rh*vFLI-dl1MBYi}&_=Iad>+^p&xvz)>VC;bb7_Te**RjW$V{<6Yc1
z89%*DRxn7bgQ>P<gTPRCcN7Tz-l)n>{`L`|Y0hBT<bk>XFCPU|EG%;$dI6CuiB}-=
z;H^tBhmAe^OwAddsQK>=YExO!fv4ePYhu-4^bj}qX`2dIW7sD-rudiv6D_|#@V@9~
z*#Y=e7FUxKfk$CC6oCick-*yC{`_<TRvf`Hdachl0vShhp-HZJf0$&HkWBD#yRR?v
z|8ml6Sz7h7mtHn8swt`A5~|d)M(I|tR6Sto;9NKpV7o9avjT)-zs^(prdRv>5&;K*
zR@>Y*&H3TihKKye!UI-D!+OYMUKF#2yG#l?4Fq+_#2B9UR8G>u8lJz#Sta7h40<HE
zk&Yl0=boeZ5N*_EKa8#3fioqOJAuQo1#y`-BFalwb^se0s%`86;Tf*(tvyDFVX<=1
z>RcL#nzXirXy910D=RCNsXy9vWd?;3)M&{ixmteS-s6Qxu5)5ilGjXGBURYlUuUlb
zGWbnlDGq~)hQp0A2sdbeX>tG;a#_pEjJOhYpIoN~2-IX!Gd#WLAa1eALGDNb=p8~8
zd=CnxwI`X!0$ClZFU;2!7$TF&ti;nUW`l!+7zT+`2^KpUgb|dMDrjPor5k@v6d00*
z@66<&^zmEutvGadmL4_<_SpqRx&n2#K0fmH_LkFweVI#1%|RAR8MWa5<&Pa|%irt@
zRrkZaoR8e#KUm>pgHQ(2nu=q`O!-4_IbGRg{pTW(CCXN&f?q?mO+DbePkrj;f0X{o
zFc+xvq4#UDdX4|Lja!&#Rq^=#1*91q0Ti%VYYdl&zuOvT>iQS_B)Br)C_@0p*lRK)
zjOv11vz(^$*M81T4RnSL3HShf-e2q|JeK$AqL@t#XOZRzqP$)dNMDpxzEH-Uh5aR%
zdGZn4x(|WIKi1wfc=O6v3HLU@vi)j80EDAek=*Eg*TB16a|NKv&dv@TYClx{>M8)t
zEN3_*=K*l$-x71UX1buzw+81w|2^5eTMzpqU06^Ic1O8x?$b+YEp<TI+btYZVgY6W
zi0cU0(I${T@uU}5JFuqv{*LtgzrDL1F#cK>#u}37^?AK0*+|WkdY(=D!^S@f$$-3r
zUBTbbL$cg|csSrNtoPoPR5j#Fauq*mf0?`xS3UlRY(x{6O-aj53&9L70#iC?Y9d*0
zq|LDx(VVMK_W%BDco{t+DAUASW2PavJqXl5?r&f(&m9HL(zr8@Fuw(sSwrFh=RFR>
zM02|;CE2AETjeYWFXm~JV9~j7Rq`8<`N^%RsabotSHi39MB_Muzld0z^}}m`#V;EC
zTRi6h)D!(sP-xbZ8!9N$^jQ=!RG0zV)FAK|>^hoO4XfYEuZB1BSd<Em#q?iy1-St=
z-ylW<mpc@+=bv3oQ9lhL-xXtHV;IP2FzZynIySpj5BH0egL&1v6`l=k1n?Tz(?pE1
zVe?RpS8cJd<k{n}udgMuN(zKweAfrKl0q{e77G9ZJ5031!mu31C9h?5sdJx2jlckv
zjz91$sTeYh!eB6xfs+|(OPxA}YNDWnrDwxGjex+Zz+ZOf$Hgj{=U}m1d5RZs+HB|J
zu;JtYR1Y&9U~MWeI+RfDkSBl}GdsW=8=<C+YsMlpPC9jG;oiVrZuPeYieMc`LP+`z
zd7w_fq+}@p*xvUq&tUOq?app+Wl{xzG=L-*bkR$?12|L=vJGq+gnA1Mky*>c1&>Kk
zn%>?1><JLjT?&eAfa1Q;Se-5`Eolh_?XC%f$3f^w_+fJI-G(tmR@EibI<S3Ino9oZ
zc-jOYQ;_T-+=T-HE!oSn^Ssdb7dE!G5l`wozRx{?{^e@Y87&wYd_F7ycxD4g`*Me$
z?^5#ozDAC%4kyLvHOk3~-i#OAjLL4rQ^=|7jd<l|#5$KkEkxGJ>PmY2CS$k_s<Q;X
zPKMK@3*dbmY*}=gxhJENxEy#E$}<TH3esMA$fnNp^iR0DleuG(S{^PT?w|UDCo6WA
zuRBIMlwy|)$giMt2EeDvgshB~Ob&A#0RzIZvCO*^tSU<}1o`0GxAD#{D=!c*saz%+
z%@o}N^_l?hMq3LR7TkUgT@}cheyMrOsxQKCWRCXUp*^ucN)*aVd5#xa$p$nKKD+@S
zh8238g}%80T`0elUNktrcUl7w<Li}U0#IeSm4K!@xvrw3LN0)ZIOW$J3Tg=ctOWpk
zB&N)+(E(sPwb9>qb`K6(_U#9&on8%WAju7LSj_stc`Ar0a-6QnkQn!n#UwcQQ|>v@
z0wY$=It_tjorfP#Rpjl>B7`iLTVTAMz5<6SR|9w`OqpQkuNMVEj5?e}0{FF1<1pCk
zr3(#?p)$Z`>!u9crw45TdgwSi_OgTlnm2yF09s7?0rsE?!T^Zn@})t;^dAOci&bX^
z5P~rMX`o@r+GBxq`~c=?0$RtDz>Z#B9sfWv(q0d1PPYkecLzdU2=2Zi4mEtSbT%ET
zo8DYG8!fwlvy411Q*buau{|-iIa&eOP)G9<G21u)hE1S(G<0f@p+@oa;|=d<q%xF_
zZ-sVGF4}IbN})R;KIA8FUEKu*boBc4GyBeZn&8S;v{A@<x&16<X7vq2BeiwemB8xy
z0(eU}iR7%c0af%J(gqkEMkdLgEr;_<O9`;c5)j-hr*{c(KKNt<Z&wO9?F^^?NLo^?
z=(=kfFMJo`vbyg)MOlfcDT8PFTX~K)Mq@DJ1Vr=^WP&Dk|A70=w1~}6?@fTHIy+x&
z+O~WA1#<lkkQ#en{;~X7hIiaOpw;RI5*b@6wmG!Gl0Cl24_f*{Bk&G3Q&QvalcTl|
z)cShKH;?QOyPL+1C}R8!J#iTvdR0`w`W48Uf-9SVejD6{n;&arJMV!e^_7!HxgQ@l
z*P@x6G@5mk=E6PCF@!>RE!Vw)OyZ}r$NeH!FEBPk`%;Wb2+Oqy8~TY!ozv^)T2LKE
zB*<<7^O1mn=np6KbadX|w)0%@XdXFVnJNNJx|Eo1PTvzA-i)`7kz8E1UyI1XY9(+U
z+V_XKXo*8}&JQ;434J@GdMW|sR-6mO+>anpWNVxne2~*%SDCs2nv4W>AEk#QOM`g(
z&HLV?bX{d4HbEvvr5pS~jx1n0KeyoYHfU*ZO5tGGOG9;yLa#a@b2*{Nf{rpMba1d3
zvc&}lfQp1SR4R45jCP$1|AWWad$4B(!a`7NppL$0{oJ?`l4-Ni&-b9}a6)gNd{}rf
zIT$*}FbTd+1SC8atP)K}X7!pjr`d@5SZ4heaWFqOH4*k{g=WPq-^zF{0-=K|<&g*D
zk6!NJ((Q-#H;4Yjb7c?3jcc&wg(3}~GLzcW*c?#dvm6XO;(rQ2oZGZ{Hg>(QX#81j
zbRYhz`d|>R)|Jux2R90MyQ8JdT5;eDk$~`zg{wLMfxgxn!CmTs)%+rJ9f5PF>q77}
zWUM-s(U-Of%ELZq3b@_Bf0focWMT@0ZXjLR3d6{@&&&fdN;oNRZsXH;j3}mTn;dPZ
zo)3EX;i3UhhYwO$rmo{uceWb6sJxp@P6e^<Zo~6kum<CXW%U=?Vlm1NptI0n(3z;j
znImv@ibF?o0$4q5Jx>5W4(f@hjla@d8F?(O;&9b83=VW;aAKqF0S<&wG|X%?0Um^7
zjHRVzcX#&<1_w+pUC~!56nPvV@$KR#vn>C_#RJ6<4v<t&_{|xEm9@bKmH?h77k-2v
zNK_2@X$cTRIMTs#08Fw0!bdK^Z_CvKP71M@Kp4*mlWwsB(B*Hz&@nU(K(sneyZh`5
z3%DZ!Kqkx(c5QhuX94%-FD$@j0}2Km3$4n89lDaQvCD&Wd5S9FDz>6<Q0=?n0Ria^
z-(hT|6$^w&+{~IU%ljU<llKH1J7xZY^5Uo{`YalFnkc}IG=SoUIw1z`o0v8TtVgG<
z%K=<BJdq6t?oJs~0DrgcJeUQe|G%*Wr2ps>&PpR9RZ7EmsX-HNsGx}3xC*dk>!&B%
z$^m4Q44NquSM#)=rXz>C%#kRczw4GvK*xdDa6<3BQ_n%lRv=B*OJ@L)5paOD|5LG)
zP7&ZW>1sy-W3dimRN5f$QE=EKe*R$`|0W6oW@rp9_~DquADPyfXNou>e{x-bX#k8~
z>JmZd1>&U*DC%*@23&-MOF>GUx!*M4m^MuGf`Q=0K<}+f>E|E62Nhu@4B$oUX4Vw`
z%jLDdsj@;aW}12S&nqhXhp9z4;|6340-SNu1b9zeJJ_GRMrKp9q7APA%_b@fs8QR*
zj<reo4ZVhyBS4hYRwpxSL-5sf@Dec@IRbTH(Sp~<WC$=;Bci#)f%kXywORQ_K)OB*
zYlKnu|6vXfs6T+_d5%03`T<D6$Z(MotjvmP&Yg?;Q77{MOzHKg)Q3rWUIyxyvbU={
zsp_aICD_YQeJV?iAmpj9ub-Hhh#{-|y^`xkTF-x%wtOeCs=~chtp8X(YtPcT)~7<<
zH|YDJ&^vwW(j#KQlR7e2TwGjW9#IF9i$)9U+>fvrwT%1tSV%z+sGmVt^3c#wB|bA#
zmOm%T@2tcEJhNTmOwpAzkx6|k^{TraEoiv!`U(Sl+mfZI0Q74iW#b%N0RwbFhmVgo
zamE!83Hrxs<n+3~eM5Z#qMsWHYlA)y<XpG9`LAwkwtw6mY2ZzjNq9B(PzqUl^r9Fk
z8UO|WWj{1zXFWK_unGs#nJ0nTa38$<=P*$MDEF$xy0^w1%`Yr$1-2cC#gzlX7(Ahb
zoEpS|(vd|+^s=KVFJ^9Ve#}v(`KS@uihct4o<`@r^{T_(#;Jff0W_m#vj+6}8|LQm
z;9S@*O?JWTp?b}SSYnd|qEl{3NePjzQ3wVB)iV8++<HM#`$iGR*w7Z#^hhLiNHzE!
zoC3rL&M_AY!?fWsLND~43WaqhRT=(Dqu2Lm3e@z+A}v*}q>}(!CI>GxJApI<DDG3a
z{pojWG?$fd^6nK-lNqB_T6dC`h7O5Z3Y1D$-tM+1hAXtan6-!7?Y?(&f8a(a{nlyW
z5kZl&`_k6O6=3o6E}%QGQNyMHGeH|>f>w|(m5>Ii?GzmG+~rnXpi{d!ItSIlIwZ&@
zi_VRP$)p0#793AEFfd2}>H`4EeF}e?JFXw*AoYd+y+xq^@Vx%(23Y=Uzkjt9g*51s
zSF3!ydjq!zrWco&_shb5eB*I$OQfoM#|}hrdcR%gPKHr$AUGy@e?SLCHLKN(nBT6`
zPejQnRl60k05z_6iGdD^t$*(tns0M=)&j}Wb7{ZS&z;nM$H#F}KDJj%XCg7IXCDJ@
z8N&cEG<7Gqc@Q7sZ;0md;G33smJcXu4b^&HdLE4P?-v{YkIKP@!}Zpear_tKg>)CK
zO`YaBT&8^+55Q$$o1LQfhZ2hht?7qDaqoSCiWQ(=;4*2{tQX&wzLiIk2vb0#Jd1Nf
z2Y)o)E*>MP`Jvyq-lZ{)(Oil)2G8*>L4xNH4|>1ZQ?G+gZ8btx{TDbScqoehPoPBb
zpHqhY?;NyMQwYspqY%<~Q)#{=BFy7~qG+6J*xsnUTn6;wih@6VL|MOJ-7VmFIYTt<
z_{SpZc_dzq7tLX^GsEJ;gWDixs01y~YzH#UP6flFmLHWbld6uP%J}sKNkHvEo#;O-
zQ7-sV^P;xeAU*=SPS~}BOiL=vFJ}kwz)t~Uo|T`U??P#MEekViP(#BG#AalS_Hjai
zxqppFI+N{X98}ymva7;p`{f0UWu(B8%4yLNAgWEgZ#~w!%iQ`klD7~)+}~e{OuL?W
z$%mv0CtmbG7%U|wB_-7^S65fl))xuPK_06VodT3Me0-V+<X`Nx_A$;|yUUDz!haJ(
zXY)b4Z0N&TXX$$6YIpoekeCOW|D-%oLZUYE+rWTDA1j5828k3#mHSBj!&+%c$ltVr
zf6v7?7-bCf{cg_7ay%4Nkhpmcs4edSM3?mGy{=rN{ST^P_U_%gkirbkby<T|Y3+sA
zkxNt{SeSO!)W}LjGV?%?Fh)fQ_*AC8Men^v7YOC$<p`{UU^NbSG6CjPa#NF&mjTN+
z)ME9vrD&-Btz5SBN4YGY&lnD}?)AYn&a>O<w?E#ukG${VDI4L;wzZsbdDfep$ZnJO
z9Vf1+T2?Q7h>K0prDfPuNR%)d!t!0>8&4F+s^?}6>$*Di23japni=f;MQ_u|%`xYX
zsnA5Py6TuusEV!QQX_>#_fS>Sm_p)V0el>V!u0DieLryD0LQ<2Cdr_ztSr|}_u|E7
z+Ty0MGysH0-kuGtnZPKV=ym;F-n=Z!PiJ^j=!2cz<>wq79Hv38;)a`BiKmxWY;<&H
zZ?DM&m0AG6+VjXr;%TCFW~B?Ug~@Imo4*GT8wL;|`TS-!9NwU|9Se4$1p8{z-D>{X
zXauN|vaIpU1__zBU%vcikfv=jT<cdo69P9475}6N<>ur#0|gDSn%XmE;_F-WzPwxl
zb=8;&(-P>-{w)1;Fj)xNvU>xJe4+X+8*jm%p*yzJ!=I4s&u}pi1YOz(i|Y{d@*D)5
zCIszy2*LFW(D4ldaH0dcweepa{1evTk1ale$1nGJI5GB4#hOMD6nw5==k_w9vm_ez
zAFCA2ij3aZeX>NoeVBjmGvik_2ho2!8dLSkO)pAaS11PowVU0(``XiY{d}t*_JH2B
zG^OwTpl_1R@OR814TgxYp0vG%&F{88cc1xEwThK_3(0bawWoffKPTOPu6MPMaB}uW
zu8UjT{WElD!F`bIWksbd3-%Rpoh)<WW#6P<C4z3B4i1@0y5Q7s(us}8n(`CMGW|WF
zI`T4IKWpl6_e;YYMVX@<FPt-~JK~BH&%C?MB;P%_-f~lCU9fGHaWxiQO||vHQ?0d(
zUoT&-{G-B6*8*E=OTw>RSgg{_AYlx;wY18Q6oz-InVIDw5QrOEPuqLPbMde3qkE(r
zEl_w$?t)J3lGALa;?Zw!aXZB<%_Ja!Js;nHzv#3i+?~~y(>E26T`{TVL>(5sjR_KU
zS=I4ZEPPEqE5}{Fa+rDtSt5Dkl2X2Y#VPmdigoA1kFZmrQ|k*Z4m|7PJ(%AFNi5-S
zXLX{44!h4-3Ws>T{OY?mKjw{yfW)?sWlk*Aw92Sn_HxRv?@kQb^1Q+mV{@z1FlO1<
z>O2D)4>US1+z82@<t;xu#XIQYeDTtm-qm}Vr&&Mas$%qYb|(ECLU>?dqO*mJDw?W<
zP6Y`mzL<D7Ah<UK{j-F4=HYlsQJo-DDEP=mGn8v(#BNkcuXgpu;3Gl58*4o!XS!?K
zh|QHj(Laaokkn2afBwG2_Kr)lv73$7?+Ql6y)8~Q*xzzZY+Y=$p!R-0g$TvN7@N+o
z`p?~RnFW!XT{SmeGBB-qdxx}<@nL9k#XjY%+3uJ!$oS{bF3qw>eUfIpk^nMv_VLj_
zlz%#7Xz#ssG(hl_<KO+omST+hVW!knxogJkMksWVT3py)ir^^A<A;p3h}MM-8gT_3
z=3>)TomWjHIafG^vr*OGFXveaywrg%-RtYUY=n@X@O&f$om+g(A$uR?HYx@<E1KN2
z$O^DZ?R+WB7oV>Bd3HU-)I9Fq*mHg-c|Ep(W%fHpw*IX3=B>cxZl_2iib|D!>TQWl
zPc7biteQPPdUSz9@5?n%HtjW;#y*xfb?|PMEjA=y(M2hCd9c!Qjg_?O`t?BNAMa)@
zKi$(Vz$u$uY__ejlvB*m!N^TDkt4_1eT6<W-t#}a>5(dLf;L<T=ResZT3OjmHRw+4
z>wbB}X_y{Gk(n;}c~-Ns+juQ2sp?7sDNBBnFhwx*;JvAreMiwX##~~!Nq+TJD}C*r
zylRE%8U{XI<G_`MM$vrTVAb}z1;ZLi!^(&&nm%gsc#7Ef*&T|nyF3D!Pv!#`@ufGw
zdA!1Q?_a%0bkZdsxcD1JKFPqMmD^%DYaV$Uzul)cwCs(*(@WJ?I`1AcKiBBr;LxLH
z*F0-;1JCOYec0`zla(JWpR14YShtYr9+1kut@jZ7`$e(pC-H8WFWV#VmfD+^!^yOx
zh~8XWsRX`9*@-H4qj_t-Y1TMP&t2yAvQ&JW^}wXg0oNTZmuCulTe~+9E3%f}v<h0{
z)m#$3SxI4r#i!4Fx)Gf8FvD&C{Q;8hp*s(@>AR@^^fpU*FXFS?Rr_tl(q&i27IK^&
zc#j8-aYolg;%)cxGzRRA)6ST?kdC6YDp3|Jzts^^uuZ&o@tp&99qKv7c3BgMFsyM_
zHtU%D)}vW3(=4odyi8U*yH-%IOZEu!C(@Q^G~KDOc!#|8&SPo^GEXgu6{0-{<dV2Y
zF7AE3$Bs+rI@muL7$0+}T9WIp<z^D2(JOWH5wyL1`jzC|iT*(bO{AW?s=pk)?Uz9X
z;JpI+)C3>kzwIfWB=;5U_IQ{hAK6tgenLRXE<^&<)p*81)`4$F7hU7^depD@ggCW1
z<iDFPV!fR)Y@a!PeP{Q<jp<g&kyTQS`L2x~8E%5@FR_LC9VPv<dEx6p`cvWM$U6=z
zcf}LEIsRB4i+?4mKf*fytYd51efC7x@Lq0*eYU(eaqzlQhgy!T)|C#Ytp+ZVw4-Xa
zEiLb=sH2~6bf}TF6mr8&XuJf)Zwmk$ts=4Vj^4r1U$45latRO1%Nwdpz8p2y8Q#3%
z5Bbh5qK%Phyw$L5Lbm)b`ppl{xY~5f+DJHr$dKi=yyDWgLt;+B)55v9SyC0zXcqb2
zD`B-%BELx5m^E)e-xU-wfP5oc$(~9na#d%G-d2G2g462%ZRyz(x9e$XwX5sO(20xh
z=YMUmxL^9p0d=_H;$<V-{aRm~A*eq;FTZzK!a+&>?d@g?9-np;Z<v;J=VO&4YmT}p
z(0xd~oymT6VCG1{W0Nt~WM-`2p-%%34pfN78B&uEo_jOew|Dg8RLnuLoX_OGT~hZ}
z$c&x~q@ACvnw#Z28HUGR46UO+JCYrtcjx+r*aZb0g?^2=64M|vdL@d!gb1Q^x(*-A
zUs>6IbbUrODocW#u%6k7@}Zkm`c#WtY~3O#fvhv|Ic<s=eJIN6(`{w3a?C54?#hMq
z%DP;EPn&vNwupk*CPclmOZu?(d5mA&DNmEU+6!kT5)Xynm({`}R~vkr59aH>^!}s}
zK-U?T3U!!if5Zx`o<s>7NEf}#dKfrMTby{O7K8POi1631oNA&NxmwW+*Sdt-Og}J;
z6bI)cqDns4GHOM;>htyO%Qa&7jT#dyNfDnQ$nehQJsG>Cj-p<nSC@tDbcov$<1g1P
z)lUe^AokQZ--S*M8>BQ&t}MMZt{$`TDELq~*UNZj^VR8&m?8D4j6Or<x<#vzvX18^
zg~^D!;mlvf{pr65*ZJtZvlM)=*&%etuZ<Bm>uAw;OD^5434XFfJW2m)p-$jIHx9M7
zgho;~xoRRBAKLz+sHo^YQs{bKB9bP2OKZdNM~#l7;wh0b#utu=mDTo!OR1Pe81~z!
z^jx1Xzq;MgY12MXYhe(v?Za*tW7A$!#WN9jt9BcJ)E}wa785f++~m0lm>qy*4?T^P
z0W}|GryJHoDu+ItSB=<n+d%90;+dKI4ZBz5(>4XwM-p4{J()n=I)&XG>QHK4YJ0_r
zc>QAIb3yGm0@HXlgSGce@Xh+h3%42?8uDp<+@)dFj|rQ!>rmmBBEJU+-h2kVOkWv(
zd-2N6<sc_IHOEFI9b-8w7OyzW=>AqFsT`cIZ)8OLcAlT}hK7$-OnTs%5Vg~}W+Yvj
z!}|RE#f$Qi^H6)-=h4o$4Mr|4x;>S-!N~frhY_P)T1hh3R!0hZYyFs^;g3f+pSKa5
zS_|?Ohw*B+?0Wl5oPuVpD_uwXK6psFi`;w`eQW<AkPpZRrxWQNYqLGGw6xc<2QQr&
zIF|R7y|%1)?66)nMctH8Vtp)$@ku(8I+dNR#QM@>G_A_UrU&NQ`z7<4!^YPRBC$q#
zHX4)9bknZ&R}?|MKc&|soC=5<a;tt%9a1K+r7vm>$MuyI%on^{lJIeke>*jxvecO=
zYGEdk5&QaehN@X<Z>d9jf6n6u&zfSJx7OBk*_PGx^!#cE=h5hz_oRf<4Mp~gUf!ro
zP!^ck)^>GbeEc5xF~f}(b;H<Lwv-t3F%pQ1#vseJ1^0AejjQfE^Cno#<A_>04&v=9
zB^D+o?$vkcClZzR?|+E5ihl>#bRGsX*qvBM5!moT{MN>lo~f9gsqH=z_Z;q@NxGvp
zFCGec)DC;r7AFZj6}LVXZ+*->j|4psxM}=@ef<Bhj5rfe0E-P`rxL3+ZQ2xvoRfrk
z%Fq5^g+@UBo5diRgQAverlx&r)y1WyM3AH{s;D>>pqaYQ2QJNk-aVQZID7Ks4Nwr;
z11d(AKqfrZi?mD0>>emu+<W|1==bm6ANB5so7VsIRp^#pASkwqdGjXk$u_Zqyu7?F
z*Ou1S)9z%=nqjx2mKBGZq<{<sL5a_yBZv3zKM$&cUlHkjpfE&Y*nzjsQB+duc`T}(
z7ZVe+&+o?s?kOfwnCvjjvj^AA&2!1vAQxIkN(+HN1Oj(#LPCCWvShflmgUsclrsf~
z1%G9oIXO9bt$G_g%7=4pP};9Mc4RcX5YvyUX&~YtNZ<ne{{qb7`5`Ga$XlR&L1AH!
z3PDq61<qoHIy{GhpMruxmruWM9jtOqI&=TqK0WXkD)9>o3!prIp8-e*e);nqj_uY@
z=Qs1XYq4>0IeB@90yL#7T}fuy+1Y~_pTgCpsh!iy@GKy)XW}F1?reRr_EfF4UY`*C
zD_3G-V=w+?zzFe?OD1M!it_RoCnhF5YpfNI95Dflt24YCemH&W8v(c9?%ywxMz4G5
zb%Pwn7Ce0Y!PZ-wS0usVj7?5HlGMI?+6=x58T|HFW-8zL<HCPEb)Dm1*+BJ{BU0Mn
P0rV~!{#J6~)}#LgIZKXP

diff --git a/docs/articles/vignette-07-backends_files/figure-html/unnamed-chunk-5-2.png b/docs/articles/vignette-07-backends_files/figure-html/unnamed-chunk-5-2.png
index 0fcfe84e1b7e80341d818ee8e6cc6ca28b32e728..7443dd2a43c4b7130d26d80dc5f77016c35f7442 100644
GIT binary patch
literal 20790
zcmeIacUV(vmp8ltm7`MhSP-SCpeP8aH0gLy5fD&%M@4#--a$RWIVhq*KspvcYNYp~
z2uO|6dq6q?A|-(YNWQi8d1mJO-uIoE@49B*zlJLq+`0F?@3P8o{nonk_f1U|);&k}
zAPB;$di{zvf-uz~2;<k^cEJ@3Z_Xlk+jUo6<qEvR4@J2$7~XcfUN>?_kl#JfUxra@
zULOQGj;LO_tm~7wFyf!s>zPES&q#=|oNu|c`_>L+)n18%R|%!r4o=}Q)An(<&y@LY
zU~=;Uv)do{mYC@XnY5NHCY&t0cHG{|*N~mN?Yxu5MW<T_uAJX1xBm<all#W@Wz7P$
zjs{bXzPg#9+L7r-)0B13z-ftI4c}pf3~d$ofnnWdHZUMCpnsU`tAn?L4-uF~>b@Q4
z;XjX$hrmZ?b}+CYh|-DQZo{QLmk}lex%nHT&tB*9shFonB?Jqt+p|b#;4|G2Wb|RN
zL!b3rXPQx=wMMQA+?~q!A9i&FK|=oc|M&EN|CT>PjoD#5Uk~J%RZ6<oEJxhf{1Fz#
zucJj%iV&FRq_Q4){`Q%_%jo9ldb+M)k6)tm{yLc9xIC#_#}{Wo_k1?&be%@|QJSHp
ztlStQ+3T0XbEl^eH`y}BF_{%iS97kS#H=liRQB>YNuY~!&=^)zbF)sB(pI+Zwx#ga
zUS2bDttN22xrtb=$Pqj(Mrmu{=<`d=oj!US*5qrl&-cF_4M}e{whySO(&n5dRw5ko
zGD&AfPQr?Z{9#h!Xf%4!=Vorx?D8)1Qc}b*sdv@Y)xB>We?2bw{Cs?c_E~Oj?qVLx
zYzm$eUBUiwmxn!rgYnYt)059+m%h|JU=s6PndI{7zf;V!7xwWL`PnlM{!Dxa@2MjB
zp>d{-aG|`Pt!z3Hi;9ZwQP<|?x-zsHBY7<d-=IfEO2V<)*vc3Y+<atG)F_Em%#-ol
zYgQ+<FUKgVl2{{~k?dzbPat%4cfah?Qj~X}>BuG7!ooXR%6cH8CMD<ULs``=EiGqC
z#_PjumcKXHIXDb0R4*<r8Wq{zn&}HaBqX`X1>$t0JvWy!lQNw0jZTbbso8RnrdALq
z1+)~q%4@tP8rLcpvg2GwN@Yv8*If+W;jJv%_1L!sRac(5jJ|^LX<=AsrP!%&9OiBA
zgm|%K*;1<?Qx>h}^b11xV0&Mx8M(*Q56`5Ei76>T`;y+Qtw~8p@ww<9?BtbfR=LD#
zGv6YIt(r-c4SGd4UeT>_r9blB3^tE-;Pn+BA`0h?_Ok`<S!;<F61)5PulAC~qApG7
zEha`>Lc-J8#l;2p@~n=eKQUo+Eq$JSH-fAgE^>2IvR#Ist0em?1Rk9iE7QVn`A}R|
zCZ=kea|KeKdoLE-VD4G^`%9K=FE1-`2lw}q!&G*S$Wz{Nd2f5KZfC^IP$D(eEdv9u
z1uw>Lk*8w2u=y8`4Gy}sXwZx2#IO~uD~)9GaDk6q4JEs2h0cnXubwC1UDHd6`9sG#
zR$FRnHwgl(Q*TFKcoVA)G>;Z)(H2@9R#zHXrDsx^vNqXEu6fNOnwZeg_TUELk%G#+
zJdpzenhOEjpATqFh8}Z%Kko=?fg_KF+pK<Z$IX-ljwVtBcoL}EXBFuqEnMUX!s;^b
zaFj*jh4R)-zo}6J&ou3z{O<$kXjf73mwKjnnv|xk5KPCpkIS!DPrGOLl}?t<BqcFa
zOC`Og))Dn-&2fisNSC*YahLZc&*_!UeIAoy=X%~qr3R2knRomk+kSa|!l>MB%6h0b
z*CbXu^JPuorgKR(3^rxsvw(~RE=5t%z|6BpGVTC^%*GwA-Pq)BW~K++m>EfEVm}to
z@4NPxql=cs<cL-1?xkk&<zi?0MDx@ziQ?8CGeKHM#i*W48zHZiWhP1!<0>xVvP(<7
zJ>C^p%ZrLjW$-mW^TEz@Q;&|l;8U?~#+*tpBUT1bR-SUpN}ej^VMdVAx&&sj7RHo5
zi<PZk8MjH%!YoLn2G6Udb#K2Gj>(}04$^+Se>)U&ZaZpo#A!w&)_Qb}a4KoD+NoIw
zyE#e^Y?_sd&m3o0@u<9rYkv}J-9wGxZ)Q*SlK~+bc#H3ly|;B*c4Pn8114tc;j;VX
zG%7*NuJc`vQQ>I=b6+b3>bJ;VKXI_(vaVm2Dy>p3Gr4zRWADwzV%t6>dC$xFZ8mss
z>tWuxC@Y7GliEQ?3TJ9@H&Upf;-$_3?KSi237<TaO-yKaBS%mn>r<o>0_d4jb3<RB
z>k;}4?xp8-MDbV8&tKdUGO!|EW|CdMFd@n&OyuqKx)-Lvn8*)i12^B3Eq4F=XH(aY
zJEd3K#0rTuo4tPL&{5sA5S*&XJm5aM!PM2AeL`J-Tft>0urz?k)HKr{I66$Jk=63G
zVB?bE7cF-DGVZ^WOcCJPj1Nq5|IyDf-zBSD8(<ZaXP(5Yxc0HipIE)J5*jh!()TfL
zQ;L|^f5R+JK=s+M!xd|8sT0|~SX}b<<`iC_N-WsFF2$umF>_Hl(^B30v0xcxb;hRi
zooe~kN~=4KGSlY=wkK6>se+K#7I*)0Uciinb5u%c8o|M${6&w!=_)ZB?y_J#8^g-`
zmR{vs^r=5f7m^5?K3cJg>oEhXE0aY1ceLQ7nCj~O@krDyZk-j<&-s!RshHm$WPY~!
z388PMP;r~sF|%^IS&BZB&p#JX^m}xvvIP_Pj33wZvF1sK9LxvP72RD^6b76!(oMHp
z>CLAU1CIy#_t+30Fcpp!MTyyug=EpUP<<fHdRt=|rnYl>OUT_lPPI+9c7mI}Y0JAc
z#=P3_geD>E<b*AsaFa;7Uq_Rk2$@fwuQ~<W)<PSo^;>$(g11~>r2Z{4<Wd2H)vX)*
z_{bfWLwa3hZ|6#u%KH@+A#?MD!l1I_DK(Rw#*v-E&(K*t_HMSaXxqsvEx_EC7^Uou
zcS_*dz0_Q@oUKyW7ecesa$7c_$rGi?B;x7nYSfP_jZ27&PwwT|c2MXQprp*0e+exi
z8%K8Yab0w@C3DK14~yW83d%^*C)}^l__=T&#}oqHHj@3j{hsnnM9XT&9`$<4_dSo{
zKZiyBeKhdDg)jee%YQB8`kzJqUm@~;xs3hEbD)mUo7-xakwlJP%+1X$pN@BMYdM=S
zS<g#zJ$%ma<~L7(U*WG`_cD2Scuc0nnzw+H=<Mw!$CHHe%1$ydv$Z8kyLWbWVgTu7
zXr(A{Dh35y?L|?3Xgn`NvaGkWCZOQ0A7P3C?St_W4sPos-ZR}gS{i|Yf${P2Zq&I9
zo2B9MLq_2EQZLyD?eJNdY>t!jbwd$t0JSS+)X{06z=~7S(eUW8W3r{Jyb5wpI7Br~
zOiW7q2n2#`Dg4YZD{fq#AvE#$pVQFL=-G&d=f8gt&-_DZe%{>wyx&wyH0Oo8pRI?y
z2JJHVRn^tiOZzOi%K$f0nu52Q+DB!5evIXAIl)|K?fh;aE~?7)(|leY9=aaehdez!
zDYGhygB0rPFrJpaS{kWfW+x2RX8s$4+`TW)b4<(K3Y}mX)iZU~@HK&z)iis!2c5$V
zAv849)N8=%5G?}wejgo=-$Bsgly<upbIzz>VB~eIs6}$VWfKGqp8ni`kiXADpRQwd
zPmf87^jQQ4%E-tR_>tyKeW#+OA(jv-U<0YRzXp&w3Psm(5)S&OH4|T%c@3PhN%plS
zk9gz5bJ2K;CGDB;$f2yUX<SN*YSihel~z$40AdTmeRzr>!N=K3(v&=49Qlneno81|
z$PmdQ9U19;tMvd3WQZCXz5M4RxBqgn{2$%_zo7g7cNY220w$m7wzxySX#DJ`ZnkHv
zq@<(+YuA;YR$S~^=!B&i7u)M~$PjYuajek9%5JtLfl6LXfC$yAW~nE;FC~Ef=k5a)
zQitGEUBg&LAM(P0or{YW=0@DbQ+=lH8ld9zvE{F?mEAw@SJ8F;1KPb05!v};vN=k&
zHbpTwC|xtjxcC8FPCeREv%5F-@*~v5=~^B}TgJlx7YB>W0ko}js*CDb@}O74;VL}G
zmtcaA=M=N{Fy{b(g~*%5<}m~n<ATVAoKnuFH7n8BNk#gqqWk0vmGu1lWeH+}l#BV*
zT{}=6IwKpxA#D7zx3`xoaP6C_WYX`|UW<d_>?o-dt?Py0-CBivUV*M6Sy>AXC@*r`
zUpE?|%^TDjM~pBcNS(gW=D}J1a)7-n*%sS^5S(L;On?RvDI0qjFv-U|M|Op~Z=SVc
z{=q6W{5(oaL7GG&K}hW1?g!SDwuq<Yo~(zCmC{~9unC6Stw+zmYN9d1cRJ2yC0(i-
z4ftU#ahOx;7wx;>SJAhL*v1^_^g~Yz+gQr>f{ipprrLth1)EEi8P(PDs_-}3GGKEE
z-*J`0On3P*^q(pVtzn{TfY?__L7R)Q)6>)K=Ki|BLs#+|0}KJ(w^XKTHkSwifnQF9
z-5dY(I0<WFY`n5ih|{;0fkyA27cgy9t#-8D3FGedO(Hf6q$D2DymX?^%-bwzYl3%X
zV<Hmw`R@Zl3Men+p)%n8K;iped@FBxx6c4wq*gHZ+Yf*I=4c}9tSUyTmeL*vp2sQ!
z-3e#G&EH%xRTJFAFU5W43HQ3%vDWSJ(K9K*3av51t?1mriifj7l#Pt&c_Fw&=~d5C
zVnM42_>3!RX`P4h=q?p^rXkJ_4uq|ENhf1t6BDT+@;p{ATYoC2up^5+3@vWPh?p@A
zm%88`hs*L;N!h%Q*8EQ2c=fe96FyA+RB^Db&<3aJ@9&Sf*Yff#>T0zO)~U>qVQk#-
ze9tbuaXSFJuC&wRY>0qX@-St4i?S-_Ip3Yd3bR0ZZjv=VpbS}`BVl&-OJ|xoghV*q
zd+2(+FUs{}>1>M=6LakUz@PBl`_rGhIVJ3ML)m$g-$p9P2n!1z;zswPB6|U;nM+S_
z`?!<iUgJ|vRgA^fcMb||y*5P%TXhljt907-YJ(`m+>g+~czetp9W$t0>r^ki{qf;b
zPO;OfZuP8E_dfr%KPgh7f;yiq<~&?>BSoRQAPQ|S^)hhB#tRjmW|PsCn2x^H85|C<
zMK7fym`*bS4$3MWmNI0|S=o^FK?iJ(NvVix!QF=Nb7mFdO<e9bW~l0-)!n{<C|9_4
z5ahbZuInO%fC1~3`1Wrqyh)9Xja7?IF6x*Z<6=SB!5~eUxp(30D0Mn=i!4z$ljyE_
zSj6-dDm&nV%PG?dUD*a${$?-_VgVaJ<Tl109pXKRF2&mgzB*8sIC+1s_=JS6;o-#f
zsu}qt2SqvsSLn0+o#m+b`!~uz7a<xZbtsU!=G$HZ&N`+#1!TP~Nmj~x#M1yg^vW~u
zDn6a`8>%QXMFlwI*2L|7Z>;VC3IW(;7e2Hr;FEshs5{wAW!g1sX`*Sc9_W(?2M1TC
zM1!ScErNU%D9u`@RrAWehqG{9I`2N+hT@cTt(0ViK<d4%jpbhu=}3#h&I*jNv9TLY
zejSw~J6qf9#L;SZ08%rf{={}Fi2xcAKODF|f<guz3yYNW^z_6a>Kv*P7>{2#<GmeH
z`{UDIJA3=V1&{gBYJUM$wKVM*VUw6U^`V~0Y)Sm}VeGi?4NtmYu8Thh|292C32f04
zYAPj%$d7opjvqLE1ATU(+f>WbBNy%#l9Idz&o-yOZ5lCivA3Te9vvM`2KLaSh6a<2
z3)-f7u1bM0r)JFRBT`aQHeh%MJp0Tl<-dfo9=qr=T2%#+v@~oY?b7XE0L+nGQAxns
zf{i+4V46Qr!oal4!^sEs&#x+X&)C%`hpk`<l5m@Wrv#gxaz>TA&rGL2ViR--R@|P`
z5`#rd^d9w<{!zyyTguLPkzdFlZ+M~lf5B>$PaHL0DO7a%8aOOWdx@yg#OZ-R;pA3t
zG(O`_`1V$1Mix)eoS&zaJvpEycY&EVz%@BwrCEds!}di5ZM6p*<XO0(f%z9+2m4Q6
zCrvwZh=A2MFo^Z8Sxri^c&ier>F(|x8;p8IzY?&0Z=Q&t$QoUPMAuKdGqG5rY2_qF
zeWl$cNKDMSZML(<JYcyYfas5Jt=(#^t(yMJ5rfJJSo2m=pGC^EXU}rXt7TC71_j*V
zIVcL`0u*R=uyUtzC`+?T@a7O0Xd%_k{{C02$0R-SN=r(tW6ivp!D{1OE!Ic<T@=>K
zaZZQ7VPAoypiUT?-RsRUqKvZ#|H7Dw_ZzKoKNgp$F>OGSMV;1YnToYYUd1T_SqO7+
z#LFa`#IeQ(mh@V#T~7<z?%K$FVlRKKG@82(sN2>8t5z9a;k8<(GVmqaR8s3;v4fE8
zoaAY6f2oE$e#9)EVK611{fc(1XJA&`K=NIko+_JoM(Di>D|OS(3cr*3X|HJ8mlu4f
zL4x6Mg1>A1^my-Tzh$Hq1QO+>UXwzPxla2w)F_zbE+9POwl~pW;zp8;gbKH(E@pD<
z5xXV^e2;?f#&}p^pSi!D$8~5to0sdy_5%eHn~SA3r1>6S(5XVu2bN=!DD#}X7Jv?7
zzsaFG?2+@gm3FZlxtVDMms0IZ1Ve%$R+(-*=F&D1shA8Pr!Nq<Sc)Ib{fIWewAAcf
z(~uf@iSAvSckSj#J5Mlb*5wEEM|M*4DwkToY8plOqKySx6d0@SHDk6&kU5aB23ZWr
z0*lEm_oJ=~bhhUeL?1|2`T~-C!SY(VMgq{rG4^e!kJ9^=<`9BY*j(tSlt;s&Aw10s
z4YB|j-2-!RL%H6;+nv;7VBiHwCS*1gP!fF$M}JpbT)awx=q8@usM5=c$A2Mz{sxLu
z4t1v?H6R$-@(`UNx_SeBsCO(piRwp)Q>Hq@FjxnQgwtSAQAvq&$zqR1wtjAlkxi1`
zVmUNPJ!dAfv%&>mGf;Nla)w}D?YG^$uJ!8tYkwX1ROzwj*$^9CauS(L=5GeR+pS%q
z?_J&!iJga?TSi=k>;9Fg)2IXAGpmhgCCinsZlZkKe0QeavEN`gS+5roZjbz16oqD*
zLKa{A9}`X<LLmtW39DPHI0O4%;O`$c1gtAq!j|VJlx0dI!O89A5XtHP7jPwplZyho
zZ@o7Ajnd&=6ID^6xbz(u?Y581yomxS0R{hX6A~{sA3PDMKmiFDB^?YuxR)e{SW`FE
zQEG3gHTc^&fQM7*=$uec@*_fO^p%dLwAB*=^y21Rxme=NitcT+z84}TF8=QATgz72
zoNRRm2Zz#7@I)bMLb4$O8u2f)^D6A05LTz#Tkkh<_vOBZdr~`FYIbe>H|j<K(<RkV
zKq^BnwesLnJ0L-mgxZJx4|mYuJZ%Q`?BQoNkB;#h)dmHkvtc=M{`e-bHkeIjHWl@}
zzT=NM1m$^AQMxj~ur=(CvjR2%m*2!wIY3^oCI+I;U^SRdq;E%g|BY^@F`#JC>b1@r
ziI*-o^yQ5>ZdP~_P|r6xo#4^|aLy#30Ts00(aEFA_vY%8&hG9a%Cn<K-9Wi#1QaQ%
zqoh1<D-fGQ$NqtV`08o_=>o-^BAiDo8FKdBT5hV8Kv99NF@V*!3r&-~CN3l2srBrN
ziOTHqMJCFqyrxA1iar%<im6^7m7PiQHrJG%dI^5-WJE-SQU2YlO8{WoUR~fT9q}@e
zu!fBc$%T!Kgp4N0Q`>HV#OFYn+$}6d{W(}wUvMZ!{~SKs7(W_PB>T;0o-?nyq#kol
z6&wT#;F#Dh9ft=*C+~;xRXAlcUP3+fwMd>M@JR-4zrq!eWB;Xc0X)cA0brM+;^OKh
zf9Kk5vNP*H;|&WyD==Oi6utwHYdnjC{v0^~x#I$+#mXn1$zHtk>Cb=Iky~W<5R{wX
zi<B!kS_lo5dR}tZO*BQ~0HEnc{4#oQE5t^VE_AXBY9;^WR*5>;Rf`(fP^+)0m!(&}
z@QDGv(%n;C@)lb|oeyV0wKw`Egh0MCNzxz}QT!;)4aqw}g@4%uV)eSdB^&boJwPLA
zRMH1*Qj+A!{lZN(GiGKq$G|}B0=;)RPL9$IVwM`N!-x!N4?ylze#A0TKE@Ki+QD_n
zTH`cVG7H}6suy9iT_m_BSD3OK+>V=3T*pz!-b-IDGjr=U*fE4SrKvM)y=D(QgHmAR
zK?m8|U_NjQPNiW?A#PwCr?!dy-W@V_SJ{SL2<z_s8`E12Wzzv;?<@;vzh!h*dJ|h%
zcO%_bVi`A<n|O&&M-9NU%eQZ@<5NCv_j(8gS>zMn-&^}KJq$DU1s^`N&CQ5N^rRs{
zB+pu4rBek=Dm?5$i6u36H-`ehw4zwKfy|6_YYyNTg5)7gGlI{PA#PJ!U_(mK?C`uA
z_n{$+Kr{crp&_9XMI_Y=oDd{l<L)Tr2WlY5uUdfm&{G3z91*(@oW9rn{uZY~fX_1p
z>MSH@b^qb=0DGh-IJfNNFntc2$*&r;{KR5%6HgDyt6FJ6MR9v`9mi3-(eN)Q+w!>x
zXyiG=563MVpQ>VpUF%uVKnacE`I~zTERya|G+F^U6`k}GXTJuV9VxeR9Rh+<kD6qN
z$2Daai#qzCU+=<wAPIhn1(665blIrX+01LycP7U$|D<}XNc)8*uC8(#1m&LQys)nt
ztDrWl&GPi5?w(UBG^>@5wW?I-@e8%EIuz0;PRh6^httqyhW=@4$r|pmuU2kTYB<O!
za1B=9KmRSgxlrg11_%Q63oiacLl}P4S2ZcpmJ>ZXRUs${S;RN8W*6j7HDuAAvJz4>
zRLtojUl$jf@qsdw#X?oQLIUGPBV8?sV(ZqBQCfv7i4e7^r^aTBUqIxq)`#{QM=C`F
z(D_fsAn>??s|NR$mFcEEDa-v#s+)&tv;v=j^o_uw2hERXPN!+FEI*05`Xq`&|5PPQ
z*1RpE6an4;BW9@~7h}^PWeKL?A2AD^wwaNacQTS=S4g<=@^HE$`?1=&Sr^RKDy4bQ
zvpFQFI6+9&>*xm;D$&i=DGWJN)>uqx#u#f|mVj=tvt1cDhmr_x*%Fg_L`kCB?a%cG
zVNu@O4e3N1(-^sBUAKu~epZuhi-x5k&q&J^o3NlvFDB&qvl)Hs`jM1e(h=QTtaKfA
zWT>~sMbm54u8Z{pkMYayK7nj~fhOL#h3?YcE(WA$+SCf!^Y--#N-)Q%FkXyf|Er?n
zQ#$zA8=JRt1(=YMohl&=XI`4^I^BiKBppPQgfB6&)Ui#Kb3_TG<e?X|_(K@_U&|HV
z?b65-w5o3$9Oc@JyiauA=?WGzIhaOKYiC48yjJp?75Ji_J!@GYjyGYdRdlT_5G5xk
zJw(*F=nPmgEgc=5P!`V4@<6cJMrE#d8bU$Y7p2b)I9*?9CwhrKl$F6GEZtB`yNACO
zoxqN~Pb$!w_XFnSqSr!yz!($nmkb3`Cd!!*O=XuqA8`Nik=X|FxttO$_`}Hih{&Y4
z0ADY+4#iWJWL3ZDWsAz#zEg@7q67Z-F5Yv!IFYEap9OKGgM|L4KtssIwYj7KH{*g8
z%4DVuhFp>tTv9B|S)qP${Kviige)mO<ijFtq|~U7B^pEu=cPb+LIL7^7yv@*@1aW-
zbJ?b0yxJ5aBxFzs{4{XW9$SAu$(o;^?}R|~pD==L3?+=SvhxPNQ>%Pe-G=Wsc{==}
z849!>)Bpk%bDR8v>c36m{cCVzf$R7uIdZ=x8eckBlX@YVFFINXcm)_N*$`&dqb~)L
zeNO6RXtv){;g072W${59wRnOy#uq^G|COB3#JpAxn<fvrpvm9WwJt3iz}+akfcQ{<
zjsn&q^(2@%iwn2wLQu1>{mVBw>LGZ54&}>@#|j~pE<C!Q3VjfJZ^I2YRUv5b0}e9o
z{GC5|0)Bi%D^s9X>?R7e<W&b$LPTpezRQ7|8VIIuXJ0^LGTp0i*}`Z1@j@v+=wAVM
zP+#R}Gz6V`|KJXDWp+}3)U)FZfC}XD_<DJ6q;j25<1+c>#Rv~D^@w2F0zw-h2F0Re
zQwESzDFOJHB{r1kEC&BauUFhsR@nH+K24-85j(LKwJ+7u?Ay?B9RebX4pd;v!<~Ce
zeCEaA(%Ka({0;|5*W20IWi~<{9Jwh06v<-_^H=UTd8uPk&N+>8=+#43DU5+bPNh~I
ziP9+Ie0J&kUo4O|@$|$l&Sqn4Do!$?w8km$=<Es`x1l^Ig&)eRB5W7|!hJ9U$OX=K
z)V>Wc52A+#!SMb;e!g@67<=HXvYnGtJM7sIW&0TFDl6q7=Z)y9{l8ovf($)@*diOd
z>B)+6IYVy?unfQiSYI7Q!;~TISjIANwW|(;MRw5Nh~rRpt$wa?bYX`~_Mkg7Q37|j
z#{-fnh2Ts&`l|g%)F@&KY~Xr2ii7tMv=LmgyeuSeR553D(kJ5`@>I=62)uGUG9I?*
zd|Nb7ka6g18j_6WVb)P5sykCBTN3p+y2=AmA1kI}OtLQgBd$Sfu8<)<^zHv!m-b(h
z-8gD5k0=>{<V~W06m%t0T-;;Z!i<%bbpST6l?o4*Jsr%q!><T}xx{t$nVAa0j7@vE
z-r1h$iyLF(X84bx<RbY>1zttEC8a4c_@bw$W&i_nMv#GJ%~l}j$Gc&(^=xq*SCV*d
z8ej9?sin!OUQ_B6)|#ZAy-EqLF4_YK>4UuMnH{tFJU(L&hD#?uil-EtboNICn=KAM
zk5#NusR?S=^MyJFA*tiA8i$}~!}VczEpwS7Z|$tQ4!^!M=Y=|mLYWTrbN*VcHA9C!
zthz~xS~u~uQ(zR(zTu1P5Q9DW6@%Fa9mt6mGMFY<(EWIPCtt7t1w#y4uXG2SlBPT8
z(VbCTECD1<Gq?Bf3$z#oa(~hbAL?Zk@gqMur`PEWRUiS00B++T*rm;-(9WU>OB;pN
zIB`kI>DjlDni$jj4Tm@{x!wjHkAl?$Hc}qQy5^1LiC_7$3*Bo6XRTUeeg+#1%v{f*
z<Oe0Hns|c)c>HEx^Edy;c}c$=J}H6Z=2#%_I{FBGSX{dlJ`pgpNOQUp&p&dLc$8s8
zA(In`v^f@w#qX*;byFf+>NGw79AzNuotD`O$D^ps`Y48K^Yz7JgHqEbxY4Ahn3|#p
z=BY#V`tmQKu#OQqlM`>jv=z<{J_c01d8v|^tjs3O-<(c(r1JDg{=nwH$zmqN$IC#Z
zUpljzLJycBG(6$30nB}fx)>5pxW4kj;uNlg|B?Tx_Qhslp{NFjC?Rt2NMw<RW7{x(
z<i$DinRmu<Rnt<^#l6Z<m(gR%xIragwZM;BLtL&EIImN_P|%9uXJt)@?Em&He}J%*
zSNjII8*nlc)<fohTa6Xdw&#wyH|<?FAlb;}f2Rp6GHsN{XasAH^2JrybPm3#nz3ot
zA!NxM0ExjuzBx_%EaYfWYBc>u9EL>bv-xK^7SRL%AjCGnJPj?946?CU9LnT(^Z|uu
zqd?A&)Y~Zile<aO!tZ3MoVSZ-@RvawOj79&Y^D@RGo2VUyn1&fGV=7c1--}yL42-i
z{DQW?^CZA;rMn4G@6bKuQ=|Wn6dj~GfaCP437UN!RsfP`Pv=8QdPV#AQw(2jEhaZN
z1ee$iMJ8hxZv_8{tH!lS``6T>Tvu)2hC5}EFSgf@1xZb{KU#;<ZSG6uGt;2!?ZBwK
zKxYaPDoxv%Ksnier}XNyYH8nt#s0$JOFR8+i5kH)$DVXW3%X#KlYIuYW}vvQ4$a$c
z*VJg`sz3p44(P~`bLEsy3RSmeYf2QYi{!j`?>dxhmT>`<lSGVD2=qf!9)BMY_(}N1
z*(8?_yck@{V;aizUckr|ou-ruMET<QZJu3kTu787yLAr7ta-+IBZ#EIszV6-rLhn$
z*E$9hQx`|aHk6YG#wRNW?l$*q79fwVk-|&M-DQDp)Bo4nUnQJpdsj25Hg}jd=wG+<
z6M}T#XoviA{~LFF;+A#b6uHz*`C|VB2I>o1_TNsOkqK~J>Eeju5c+Yw;FLoYK2M80
zQogy#h}^E6{O$Z#ch=eH#0w`W8*e7N<uTu6Pm$83Xv0E-Q8|Op&pBz<kcK87ewzy9
zK|;90kzIZzd|s;8)-;^W<A!Mq>F*b_@Zv5qep?SlqC^*Bt1`WN8h0S6GY9IFaO>0Q
zlT}iEe!v%&xc^Uu5J0T|uvnY5m?z}BEeZti8`o&0PW9DpY;cltq)b*0t}l9iCFYh(
z2R_A9A0>VSH6XFJ@P5bnp(|il*b-blf<~~x<_ZqR1$VE~{A(3OT?7^v%Yz3B0=zrl
z1xyp^P~7<<QrE^1GNNJpAp@tafiu1L?Mrn5t&QGQ&AlillZ~g51E|C7Gq3m}+aM)j
z5(g|$2zRmrw63{8tjI|bFTW@|WH(dgF^r!a^j#G!RQxrsq0dxOdvk}Zfn)GuC2MO`
zEky&3vyp`j7~u8Hg34^6wCcBS3pRnI^Ri|9hx6~&cW_v-C_P>4^Px8ikb>+bh8s>!
z{m%nvm`PcD9*G~YE?bA?^M2=IPkzGSpo9FiJH-{dj)5n9ilp20mtCL?Y&YwuM-lOH
zRh#n7e+!&S6UNICOmOlkF58nePM7!-?EFi5!A2?>?Ktl{J5}s^XQC0}(2}Z;-##_*
zXs}6(`wk67)ykZtj5VuT5+}<hvSl~`>~S%^GW{{BQ$EpvH&_p=`j*)e#ueYaJ&su$
z33Z~b4mzOK@;5IqzDmBcIH(vzZz)d0$tQRe38W2<WG@C*vUQ{^cIy_V;G;&U3!1*R
z8o3z&u<{Pnt+A}r#f?aJFWS1Uj4>sYEK0vgUryjZfEVW%sI}{h@A#^d>o{rz)pj4u
z_CxE4h~`bL>!Ivg$>ue7Lzy}bvM`$u%FA50{X**_uPZD!-)o=0H<mg%yc+fPbhn=%
zY%lZo-_95JbidrP{m5%kyV*YoksGq`m5af8$cI*LtOZMbFcGmHn)7*N!w_q^P1V5d
z20m-{Btu)t1={n;aSMIt;PsJ&ElP@kGaQPjW79$kowgR^nypgQW8{0Pw>>=gr2=J*
zXdWvD#P^@?!TMe<+kJSbF(^1WU~Qr~L$P?S(O{dIKTX?rvy7h6yyVaC;;X(=F3QhJ
zBSMY4N7BRFL(;Qj_sCM1LqeEh(O5d=faVSIkm6h)cBK#?vaB2oMo;imeW=rj%{r&q
z0FYxLf876xd;h)j4*w;%^4C#_uQ$;X7q|Ua0S-+&SZo?QmyL%8Y#m5>-ycL!OMA*C
zeG$#BLMA}4#Nze=6y=(ALt(!d9fzOVRIE{~W{&L0ij|zC(na)^m_3SYwuMWlh9hap
zH4@;+Y2;D^XM6}ix{$nz+3&cinI{SE-2%KfGV^Ne6=xbUg5wTo$y|t(Tf?ISZZ_FH
z)z%cPNHzNk`9J!`V4wRKbKZOVt<%k8^xA6K$jGdkbw!#&;k{mdu{iz(*}vW>83}f$
z84I>FR457EozgF=K8*FtAM1HI^!D~<{k(AeVM1QWV2b`_zu@=PONR#caB%D!V7$}x
z=fd;Ai!<vghYu9E{t?<%SCt&u<lEM^-bM*h6l4F*cU68SVDqUiHeamBV_+pphV8lS
zZ*?G4Av-A)irZ%A%SqKm%<{v)+FYEJ(;_B6<jQ3fxFYBNx^nX3f|>Wu*X>KXfaYQ&
znE4qQ(MSe4eqtY5s}Uli8((SWV7`$lsOD~Iw&%I)&yJN0&THuLG<V^XxBvFa@zs<#
zl$M|O+x1K59)UeUJhXkkF+@)<K5ymF{;=+<s=f3q((0Ohw8EIFbRv&D-K9;QA^#eT
z;5b{((5c)p$Lq)QxP>u-4#xNyI0rC-x5$id;j3`IpZAQb<68oIph0y<+&guf+OwRY
z&H;@t`L`$1Oo~a5nkqlIA*|?tK0m<hv9mGC$ws~=hyR`_*-2H%XvOZP(<*4xO7&u|
zx!TN)uZIVcoXrAOQ~2v`$)i!#q31(N_WzvYnOSMaH@gyL&7=u5EdShQX|6U#>T9Ng
z)IyB2pX8nfi3`V-D~K(#<lTvazn-QWDgK3nAQ;l6;#=(TzKqbcsOV>JiF6!sb?)Op
zQq6YUMv(7y<xFYEHq1Exif?CrNB66AXv)bu^Q{Tj{*tdb+tssN!RK!LX*}uH6P;9s
z&upb{(2()Xn2VML7Ct=P5v22|5qA@oE_yrCDqm!;(gCzrBsI#*#m8(@GEYCUum-!N
zEZVujYt#H;pm>^+pnGe@GpCj5W+e0od#5?CI2{(z%B2d^c^{|s*d&m+K`0QXRpjIe
zvML>D1Dl2*hhJUP3vhMMX+^&7{`r8qhptn%dU|!e27F!&9_8HSu#4&2m0#j=y@QM2
z!kAaj;H5{i4@H^~Jvh-O!nybQ!CyBr)_<^@^>*nM2#yyP)ArYH5*)0(B}=Eg*%dfr
zp*4R=>CO0Q@ODVZ)p`NCc1l%1C9S~qDfL-}v1)Sndvfjegw0FoSy`DPlCqh{YOC93
z*V{qG=d-^=`Kn}sGCpou<q+@S8*FLdJ6BI5-AyS+FA|S}(L`P=hOhZpPYuH_p=H~<
zc1isKkImI;yt_!!F_Hp9z@wj~M{u0%0RkF}AlHBUIc7xZkFle-5#;$-H8r(0sMVPJ
zF<yUE-p`|(T0G!*^y0n0pw57!Sb+gaQbyNc_u1e7w*44b+qtTxh^-D=SH%Ra<j&n9
z)b>8LwftyJQgjGOrg!b#^Z1Wjgky;7U3BcnPi$L%bSb*ukTK{=@n4aV6R={@NlzXv
zF*UI8N>eai&v%r}Q(y`){GI`-cm2|#Lx=p_nh*Hh9p)1d-crMyP?lof{`=w7SwG+_
zGzTvsh{Mlr#~txCX7!z6EyOj?W*?57xGs903I3EN12Tvz7qUAeJ^h+UWp-=hYX3-9
zR?bqvGJ@pt{oHlr_diT`j(HM#?SMRKAN(}_!Q5wjry1l<P`;oDQ1tsHX7E7o!(G#+
zY1;}UQ?HVapeZ@|isjvp`Nj=ncOsXbm6h-H@bxVRdg~5kk+Y5b_4Q*IAA~>x|Eivz
zo*iwA;#N6wRMyK;mkg5T(Y@v!obJ6I%&#h#DbgcP2+9@RBT1kTw-p+pkah5Yl-SRF
zl?1SE)=p_@X(;6+VcEzsu;}GmZQh<Jts`2y<@U2LF2e1BKI+C-#iDXo2g{*BY~S33
zGX17+*&i0MGjk)`ymoH*zXhxAXOP=AOclZmuuB2TR^uy2^x2S0$z`t?Xy2cR-a8Ts
zN#%5ij;rzmfn}Nh_BITI!5kg(TVKjuSa9s=?e+4j9dRbcvjfMX$$3__nfi@B{O+xU
zO7SKYSG4&>!Hh}|djEH|Xm5*ZYf+kQWd{q(x7?Xw{#bcc5&!EGl@UGu%(T;8Pm2az
z7Sq-AM1|%*X<_<5?a$y327TLSbC)MXNJxkTp_3gDs)Mz`!G(e4<>m83r7ne%DOl(7
zhgn%!n#6U#>WD>1&rqRY>DvYsGWTj)TExi=Sk#Q8wOtvCW-BlAd<abyp6zCX`ia7&
zBWY;|V!k|oeSwp&cwB`jRijv2pv22%`j$w(q!nZ{^q_Uv;O&nlE3@VO$1^2+gp;l!
zyS|#XG5vtrx=f(RnkVc6_xpvt_5C~D?wV(C7a7eXwhs<Fdk;Nk0-tbe)9A6ie)P3s
z-29`*+P;L^1J^tMH0PL)zicp%lck?j_4EAhuBM@8w_MZy*hIpPB+j~*ttUHH;aJa6
z2ILz@7gFU((2sukau#x<PM*Qnu7$d}yC)P!$HumvXYrLw)+D;LIF}M0u`AdHRNrmb
zGvd*ukwozUUE0%D@YhLZ`!|C&{EFwU4zkr@Jgn%!b}V(4VzoVYgL~66#EDUl<7HR$
zM--<ur5!J96t|piw)5C7%V4z0*^BIpyn8f6k-p{b)c@f!6f?cxzj|_iTzV$3YiERn
z)S+0Tt0PG^!3+z@F7t>Z7ybyn<KY*q6ME;2v~>A$@zA0^e<#@!_{Xs|J>$SDigep&
zRcw(h^@pf^>(+BvMIVjO`JiCKvhn@G)P)Px{tbHqGlohn^)t3jc*>o3H#lzjOv@af
zTQoFaI6nG$P2?-2XZy1BvWwiN)N)PBwV^gmA8ONz;lHH0E|mvLN^h@4bOM_x*d?x~
zt^Ed$4yr+&%=34eM9@#$_6XkLycO5S`lcl}x|p?kP)(TP#;Ri2UUGN7!SO`#Z@13d
z?U9p8z2BPsCDLjD37PndB2lL?;3zrZ*e*ta;yV&-!W@`By)y?pq2^PJ6{-^KH^G6#
z`|-_aR96$_&4yPQ&H62?LNk)m6p5y3M!}Ujv1>Jk^tNq$<9hFvPZx3yB42OaKNPaL
zPQ<G41{N8aXlrA@ilFCNi~Vlctb?qIdy+z$1vdw38v7UwveberDwj#@%;v9o*{V+N
z0t+A`1eI6N?(;*tcp2?_%oXGpj5$6YpjV{9%VXMQHPI7aEbLsj8=3V(&G&QH)1<gm
zF~&z$S=(vP-&m<Ei;$K{^O$@DaYF4}$e&T;s(q<1X@2)NQ?#gt1E-{^M0poNI4w45
z4H_*z$xgU^;z8VgyG3KWR_o@Jb4lJc5aQ)zGa`wNF{T?PW~5%0T-@f3SW?_^Y&~~z
z4c|r5+_)&eD$d2<DAHo%gLahmn_fj^?DM{m{r=I>ox!v}1}oB*irYoxSKoOMg0JfY
z9~>G4Ma^|iW(oA^u~-}IsGWt=YX*An1nmP2LWM|!dGy{-F1W5ugB)#d7wc~7E+p|1
z<N3S9Ie~XhCTXtq#^Iz$^@4Z%`O-cr-_<_~qBmVqDL(Pk0sVb5y56}vQl(l8QJO)=
zK-|*H%CxfzJ5l{3$D^S~+dljy#<-Lc@Vj>>ip{=W-f{c0KL5$+;(RvCclySa>CO%<
z)=}C=8Sigf+S{*|bcdFN3LCjy(xT0P2q61Pf2;dTI3ah*+CW;mGhhAIC9clL_-N`{
zl$!dx&d!P^40R@K^<A}pPUIAc=RkZyo1oO6(=SguWuCp*o@c}Z`-#*wt031FMuNgT
zX1F~mr6*iv61A^|uC@B7y?a*@nVU&DeT4Ve^J9Z*YVkx|Yi@#X$(Ub;PGMK)!99=9
zxd1olCKKm`;??i03`g>%?2C$K@7OK#b-q@{jEF=W6KF5b-u@w^s+n<cPmhy+-2Tj*
zDusiw*H1a^n8n;BW#(x+ss8On9L<HSe9Owd%*;&gJ?*b=UROxb+SMnC4p0F4T^=8r
zco$1Q7tI}Qx>V7XqIJ0bG-cb<1ZTcjLwbEjSwdLM45Mpgl=<-CLpCR{4*H+JJGyIY
zYlkVj8ym+jc=>14qIDO?_roq&1kel(zMd`}%7tQ}Oam-g*<I(w^(zex4KYmaK|wVI
zg@q)jC^P2a+xOd<gL}@(%a6FZlKrX$Rt`ZINQOA$dDq`TMZpW66&BV|4mGsEl~q+$
z)fcu`dMy?NbcvJ>jf@XW-#&jPq223^1)56P=f30iaW?Rj@B-7(v~mGlhM=I}{7u`@
z5~l?4B<DyjM?xQdylr`0VN4K%*^_oJ?B)08+T7aVg-gE5CgI9zC&ghojw|d7dC>Le
zh_iv8UqxqEmoTqoy!C`5I_|WuiLtR6+_GMVqHR{$?DZWkf7~y89&p1yTzN&Be~B43
zDdY@j-tM;ZER+BRc##%|`d03byBdnoDIw;ACB}lReNHO&HNK;c*)N{`nI=D_-T%F!
zC3IfdO=!IAgm|gs>cVjefrs<<cJ<bCrCMrTBC!F<{k>6#pEt6^%H{Q=XKtQFA}p!(
zTx?--c0rjg%$TsV#T&gQumm;e_X1=VlvrJI7CtW}C;C2yY+%S+Z|UN@hn!#7*3=>A
zzt|?3@<&Ze3{T(*DC}Mf)YH$^Eq&mAXui&f{hM|CekXj%Ep?8nT$3QvtZz0ZTUt<g
zvJZvcX7xAClt}$)&fKpA_a>0U+MISy>szO=$Expta{04!5lfrCb#k23l3PPs=XVZ9
z<LzmcqZjV}8B=?Z5MJ{lDJge1M`YJ0**!hNUcQh^;$mhvq7>kML_5CFIjVuvMn~J%
zX_a$|uPM&ABO62ZaxIZACUH&|KccJ1haC}KdbhZ~X?0}B%mLDpafUOE5^M{_$@wB_
zk3T+QCgfQtq?Y0-Y0j@YA1JCokf08cO=1v>s)9e9c3~=Pmw2IX9UCdOnrNcOQI=qs
znP!}sCS2hylf-QvBnT45#;lliYOd~1Z60^<<(yXAu8rlHxt`c6%+Gj7C@7mTWS3^b
zLA9)*oQtQTCD<GLd}}k@Bg6))qP#RUpGmrmI{N<A-Q9io<0X0d>XESplAl)PfOVJn
zq=$^9*|mg3&5V;Tx(Kq&$e0HJ>5z`|{1k3bu8Q;vHm0ev)a|m7dXQCJLql6%N@nPo
zV|;I(lUKPZcd`@Pu15t!i%DC5fQ*0VU^s%@t(#bWZ-6xi$Qe6jY46~mXKLyweV*gs
zLEENh+>mOEN~5vX-#otWlJiXkFY>^|-^(&rn$Da(`;F5k@%8ZVu=Qe3PfrEZ`;j2?
znhk8c@rT<F01Nh((W4E_zYUkWOSq0H30u^d0vvD{mU0=D6NWT!o=qwPQnw5%J9h9+
z>CiD{8<m%_u~K^*ni4!aa5w`muaYU6JWu45(ezv2oif^|Db?CVxUZ5LIy-mnjF<J!
zv$wZ5UZBru1!p;rRGbwR)dHknvGM)M)cU4`@5((Vfb*@fqJ4lr00?D3EsO0YdGR8P
zG8miQR~IaQt3{B+U9cgp@vDB-D}L4UT^W&6mVn*E!^367hpMbv59=8in9SqPzq^<6
zav@fGOIs~!Cc=qUu`Vtbd1a8eO4}s%0jN5ox@Xs;cJkB=WKuXRbgy204EqBw$P%4s
zPmt1z642^}0yGaoZfVPuTP7TLlYu<8o}uCUn`hnuCs|5`b91*Tm0&EEpFlii`*Ga$
z3?H8|Cnsl2<OK~KD=VwT(8-Yj-ub%G2tj#4?z2(hYVVjVObCj?x!c=NCe{t5ti?gw
zesH48s+}B2nTkCs=koyorsWW@_L)#{cZQ42VWB@iUNN`^j()x0+6r(QF1z*d_RV9G
zj$fe?STv#Cn&J;K@pA-!?WsRMxxNW2nP8?kq)VQu)h5pzODKNqK=}0I#oMgCjigA{
z6XoNSkM+Itwu{%yrKP3OIVBYDWMoj)HZqFqfC^|h46oPn<;&L)2Bx(RS>l)TEb`FC
z56b3pSs`lt&I8i7Plv7Y;El|o!ByMVv}j>5jlu?rDpv1qXG4*BIQst;j;MSsqeI%P
z8<KsmT`%Rp7e1IkdG%8$anH-iIR`RrXmJ!!4l58XV?}*f3&p9)NpS}uVd3a@YaKT?
zw<qgi^>?bF%q3pVx8Sg-dCm#t(Dxrc+{9Ls;~}NVAZ%RpzUO`k5fdc0xzQ&kHIlNB
zy@_&&t}7^?kx}j5A*}*gyf~1zqc#sgd~-mS#5)kFVvjlP5h$3)yz9v}aA*iW1d0_e
z<Cz7iw;D+%S8dF_SK8(eiq=y})GD5{!W&f;>!uENbD1~K@TcJV>=(hXIDsXsv%`YY
z#9zAqCTN>rK>(pLxP18ml;L>^o0Q}<__wyUf&lmVFnxdui#gb<J~*>c>_@UUgY$h(
z&$?<$J&3}dxNL`-@{;n&>#Kw~_-+7Gd{9x2aY_(1^JG)yh-dY?cl@obS`v=1<7~|I
zH74YD4*DKO<o8d<T&i9}{k!0)Q&+3~*L$IMUm|F0<1i6YKuzNQ<b}pQIDXO#O6&=>
z0d`;xs1v+&Te~WLfCt@9`fJC??yTIKvWGZ0ylp9&NwVJat)jI>l7G!R)2vRNlS=Xg
zKOD2kj2ut?xth^Xewh8?!-uPtLyNs|h^7cCmc1fVpx&UcPcuWu$jGRrr6t@xDK~$>
zQB7T42e1H?+uWs;fl^8`xdjCUWiNuywy)aXzg#x5#D<e(W_+Me{KASTo%{KKKb3uZ
zvT|~sdUD=x2E*n}9GxGjEd6EJbalI54@^4^m1JgRwP@2J3OT~!b?K77T=Ic!kfBQ=
zs0_XQu$=vtIvJ%U4&QBH7~7fUdg)w3``}Z1xer5&5GkT=e9H1hmoCKWQ+0_cDcGti
zxyAnc`jsO*kr@qV`#;Vf+)4I{p57xk$tKvYS6lvQPx(VNRkfJ*4@qD8fRZ?mZu;5r
zMhDExmj`Pzy7Me*HRU^n?wU2KS{(@uP)LCnWHTKoWgzdLK7C?q2ZvB)Bq1RYLoMZ$
z+-de;nqh$3cK@r|0fY6~0)eDi0|w;M)nC(B@5t^QiQLlE$uNHlRWXefiPq$$k?gCr
z<reFBeIA1LcTD~GO>67)i(+CZ6iI)^BcJu~A>&kasYN-KsCU`w#h&bNT|>Ih0s-6V
zUi|i6#?*xXHW$kyc?lY9-HPcF9U@YAUR^hgMW|uW9z;{`=WI?0A0P1w%om9}uL-4>
z(V<CV*Cb6GwfxO$0?Heoapvh|oA<SMsr7XE)zsfV68d^_^4=){f$a6KEPCtnE^}q^
z*A1`gH3+^xr+Ycd)zr|ad&|=@M5K56P3|6*Pcc;QelAjB`vZNgCI!j!QdM4q8_|1_
z?+l6&prB7**u=F|V6UVE<lnH!3A}aDQuaV;u+|k07K;p(8TBBG8|s(kGumI@(5kp7
z&s|nxl>SVYx0vhIUZE!*sSD3b4%wLS)Dzc)bc2n(95O7!j`DWR^D?bS2ewHvDZTu;
zss|%YDtr?8uiiBa&x_+C?AgZ}%96%^O6b&CBPL9)#gM6<N$#7h@75WX9J+^j-2J>(
zw>O{h=zcCD+w|GFb9sA7k9^Y@>05KDx?>0u&BYc}%@k7d;{YJsu^nX#s`E?jy53ll
zJD$(_N?MP~;43p{TOnt%AMQoqpmtF;iH&}G?(b<06l<FYwalzGpMT}F%a!_*eDc_z
za7U*uIx9p-?~yjuL?N%BB&(;+^re^XsYgQb!ty;o<OBp%((Rn=L}FWe#=mm0?6N9t
zA|?$mq)NcnAt9G$N-Dd%`po15M-qyMRQx9^PIb@!M7ts)OJ7%XWW5%?f;&Ig0~8PA
z1EsJ>g?)s@we|gx&kir#(zyEf!)le!6$g>U<0uuD7z)tca(3abm6a8?$J5|vtZYnb
zPrMphblxA{)zvk`+WtW4__TMRORU}e#Ei9|sDJJ?Gm=n}*qox(L<e5U^t;PxFuq6s
zE%FAud-dv7?mZn~C-po%i_>UMk_mvk!>r@8v;M#g3v38`JilYx58ukO{IPxT!`&Kk
zP>mYLXpwH_$b>_?8-mUg@jwM1xRHw&FP=jM9}JDm!Fb3n76<wtliJ~so5zner9C&T
zc@F;$Ya~|&oG6d#GFebxd8a1es{A47ZARA6^Swo01GwV#La8GAk7pvb$4@3nj7xG}
z<HDIU?G8oB&=3)uBcWSCweSM6(}q$836#cx>2#@W8LuJl_e!$shu+WIK7c}v%<KG)
zxgrT>A{NGk*8{wmO4ShOL2GyZ=A^_-B7Jku;@0G;?bWoQwdNFS2V91b=XeCJHD#2(
zqxLNFi$ktETR-O^C?=*87q=*?SWF37vvBFFJ+IND%X7=h$n1EYy^oyn4-Jz*f2(av
z-(`mAfa)GUGm&7+<6e|M<|Wro9$MPIVd1iQQ)o$6=WWK;wHZ^Hi~}iLO5#76Uspdd
z7HRpjZhqP?6BX)K?!0g<&T83HySq{2ll#=St?Ae0G~a=m?TdWH?bOWN%ucGojQtmi
z*wjke)y*!3S>$K#<LjOTJ|Lu*R^E9j(h`onN?-A!DIddA6$fjVHR7zVEKS~ss8|X5
zB)X3oMaPKJL9Py?jEL;u-0BqiRm{QJ`!Ux{gCch&`U|(7WVrvZ_HVS2z$DQe%PO?=
z?IU*d1^;8}v(7KP!`U@&^QPIpDkYxy$;u|()GyOwuJu!id(_c|empvJL*{kbOC&Mi
zr$D40K*{o0x?99W<|LILJHK-QAOV(kNBxJh?msB`B1|DWO<%&5f6H(Glydm;G{&E8
zk>~q<4jQ--B?OJB5&j+j%=mm@J1>4#_`SJ^K~C#m>P+%WhOzrfMSgp#d*y{;?Ecu$
zP3DG*Kg*(jKbkwL>>LjM>g@9i3caU33!8X{adB+iHM?e7Kyq({Hiy<!Ig<Ai1i$NC
zbITF>%hEY=0Y5!+KYKXY&|u%8P`h~RoF{*&V5?TbQB(xgJOrrX_{&oo9Z!hn_k*_R
zA4-4x=ZF4t_y2e7{(nyw7=TtqmQ8Hx#P;UqTUJ(6s(EM~9<tAAH*bi8y}h{Cf}xpP
z^C>}5QJ3{<h{NMAT7^SxXF3qfB3Wo{9<q-WnkH#xpyvnd;BcWI9IhS=BtivGHq>eU
z0q5Ox;2|7(Kl}YF^iTFW%vv22JFo?I-`-puEOoKSCY}`$>6Kp}=8J548^+`J{)rIE
z@E{(C84R`0A3-6u-mP1&0ozsftV12E5+noRD|;%t;pnbN-p@gNHTuV1FoVT{(|WyN
zHi7&<x`?l}K#55dT-kN<!<SK*j#PU0l{gJelKd>qM9{~iK6=HFH{*I~4YILn7$~I5
zO{a^Vx9Q+|Tz^M?>qq^c7HBKf%O|`#f9D~bPR{~Z4~-Qt{P0y5S6!+}F&&EZ05ku^
z;tRiq<ND^(PU;arOPwF?)K7X<qJ(E#+YN4o07Ixn>w_5Rt{Hj|1bom1rhunVTVxOE
ziwMHj^wiWG9E;)Kg1kB)m1mq|`drd(R{}T5diU?=n~KDu^bdURTvNq4^mZkI^Gw6x
z;L+vbR0igqEU_x;2PDCPcww*(8M;4K$3R*b0L~Zn5<i=xJ{5j;Klqtg`)`u74SH<f
WJ;SJ73iIf*RIh4Y$@}B>!~X~6_}q8^

literal 20604
zcmeIacUV*Fn>HF`3l@rQ3n~IC2ntf9h;$UCDWFsVAt)ULq}KrI7Ho)8rGtP-3(|!E
zK~d?wL=p%nozR4Y9yrf(&-~_mXXbb2yUulH{+R3Ams=$3U2DDTeV=mQ_w%egy`inf
zy64y)6bi+vp?*adg<`Blp>}laVuDZ5-bYK|2h-haYFFSV{I{lp3xXfJ@2eZTqfmc(
zApbLvZFqc8s1qoSE0^?r5@!be-}HGVZT})1KXON%A^Lt)@Quh>qa&JScr&_j!rjaY
z6aIa>oZoR&n8wa^@|4V#DK&c)OSukSxOzWKxktFU&Tn$K@{QB!xWjiY@3MIyuwzV2
zEOdxx{9N&~w$_0Wg;ig9ETtxXe#O1rn_Ny$Yp3{sq{zz2$uYvr3S4QW+{jGLS$7;k
zp@L8CvVdQnod{+?q0T%+!BeH~`i~BfADcQ?oVKig*}A#5<fwqGhrx+p?>kH5t;W@U
zo)?^m`jngS;rks7EGU%n{y)Uvm#h0~;qQZ&P;loPfBY96u*;*tO;Get<E;tB3(e@Z
zyl~7p4e=!^d#}&itvmB4Ropf;=G!-hwzqnEdwcy7&G*Bslz(w!sGU{w>{}=5xy0my
z{9ER2EG#X*5oX`hZqYx72ynMY7mGF+Y<u`RSFOd&loHHKx$Vn$TfhR@B_$=vjHqOd
z=^Y>^MzS9@rMRwVvLE+S-`Y%`ACpK~*H@&)wfFcX4vusG9v?2hs9xscqq<r1qLMF&
z{&M%)q7<RWFY51^boS*LM@rAi<@H%o{Z&?oLULaBbkx+;^40pCkgYhAEH)jtURmkk
z<<;l+VsL!h79RK6LVVCJLsL`J!y>oe$jyD^RS9KlEwS%*qjsth{LaBF#ki7^rmlB=
z2n~wvNS=}6sdb-dYwbu=DZc;Xb4gj5&GJnD(Tn#j__Y&nQ98jA9&Y{!8)`7iCD@Fl
zBqdZIJLaIE{o1@cR@sGEa@KnY<NZ!2B{?J{L{eIMqD7*;<?Yqy4hqL$Dg{LZ4^)A1
zF~6{|u)TxBG*Q}Zv?1?yh4}33>>!!c+4*{`Io7+nFW;0fy8<^7cXD!U9@?5%aht^7
zj>L%4?`nUXb6jh$=bkqis<KpDt{HuHy>c$sY@=p7*Ry{kLddO`QAs1BYDq|X9NX&{
z$+xjJ!m(7GmC_h#T4u9IaBi-N4LonT<r`SDHc6bpG#*-;!6>Oml$4d*&R@ee$ZT)5
z4s14vx-D>Uj6Q8OHFtfN+_Lo9zpZ+Nw6<apKJGq*Q^Ic)5XOk1fiVgz`;7`_q%gaw
z>okglL(hBX4JLN(L_T{bCoOk(cf!`omoHo1oKGkDqqo-yEA&@jO!(f;{<!U>G0w$~
zNF8enOO<7=i)XeAmt)t$OYtrFSt$|T;l3&nR<NsHCEF_mM$auT^o9#wd~(h{Yb=9;
zPj*qPyr7e$B&5a4Dnorv7&U5Fh^<bDFRHFq&~sT|nFzAY$~recIoTjAWHce9Wc+$t
zVX>CQ&d%p+T;NJ7qq%D`$4fdbW&2bt*LGXg%J6v^r-74eNkRp!?s8=l1H5eWq@WK{
zi}7s3eD?Qdw)T6|8s^s5*RLnbS2%9GR^x7Dl^*EGDuyUPC($X%4JtZ{(UEMvmNAKb
zcaKd^FV>+8+B?xTia}fX{)_3w?`W&HMCCVfYOo?oCM#D0X#?(SwRb{Cn+5O!te&20
zq4RVx4zl~u^FWu8vcApf0~1ya=I!XT^+)J}p#Hd+wu{yE5?s>btVeHbZ$AFya$J@V
zo_vj3t+cP!|Lu>uP+R&sZE)GCzv#55r{_^gyBp<n8jZI09KZS;JA3+C%xF$7*3V%4
z4-{%WE4qD?Ut{sdwiK2!)q22Gl}{sZnJO{$ed-TVTNaV%CELBO8@SaH!q;d&!Zs^A
z2Y+K4QHlXy7O<=NR$vRoOIt77+Dbe<MyK*kyW>VnSM#O`RE64fweYqagWSr&d7RfQ
z@db~Pq9?Zk8wyqRP?~K}K*NpR9!NeE!>;&CKf<cz6-F0r6uUlD_BDqUuK}AoJy~Xu
zi(dZ7Ts605D&exemV+H3mrZxZ2Fz~<+@%V&_N<a5Diz`;#-nqK@0dD9jGBfk#=u05
zm#k3CBY!1SeQQUtbID2`;J@BLBcM~x>b<k-ey3Hj*d*k>wKhTUlSH8`*cX-=*&9yQ
zg^a&TDQJ4qjJ@7WVe(F@rf=f5bt0^8SXzcivp0dGjI(Y_iZXH@D=u}afBv%f#!A2&
zXBU_Dc`gr&WoL&bn$m_Z;yx@6_gX@nNIBwqIujZZ3^dwN-!%gnpYI)5%a*n3bis|#
z5@uhm6b9{x{$g9(k#Avf^1f~&ABP_XP}6x@?}Tz0QVPMz6_i!X%;pkgWK!O&mPxYP
zwZNh@%CX6x7+URUl=<~a+N*lLEt{u-q{`A>oKJFZ;bd!x(9;g+ENoM73kYEy5aJcV
ziRVf?x6pak?vwo&-pH3@)W}uaHOAD?`F!Gag*X9?mz)xUOFHpUtfaCr){U8@pm-{F
zv-8Kcqx6=>?9pjiZ<_x5&%(w<e0uM$H)h6~$1k<X74(fj%)*RfadS<dWjIEgyy;uH
ziz{B9L9=_=iz)-UaBj;aa#xnbJ#r!MhS|EP%$OGQM5P{?Ff!!d{qXDB*8)Q3kD%Kz
z)mc8RDGZP>ePk$o`OQh}N{c!4>Po)$d!!hhpp^}?q2>fW^xD)jzNvMG(>e1}7<R~s
zwL<+LT`l{Z{FV*q2^(z3-!L*sPfYV{OL7FQFdvmnJ;}_@6`yIFm|Un5N?jXraj<JR
zBv2(}8KS>jP9)0dqVW^SYKLke=5#7<{$QV2;-KRp&fzDYHe;sp>V_hU!PBoLd0Q)E
zFkgc<1auUWUNdujR0~RPOHP=BCpY*oU$Ze@a#nBg*ZesxiZ(i}DadqXU_6mNo4J<o
zT*;Jj)9>q13rHY79O%Wh+Y9v*;|sof-FqG|SLN6~+$a{^9<@CxVU^;+x7ISY9kr6Q
z+NBdo)<bYas!v#0*xJC=ZPB0$Vrx^8I=24)kjADt$Yk+Tp)+!lYT=x*)$JG4`$nko
z2-aB5$1t<CoOikS9&#V2sQl|n?5|Q+)d2-~!2fF8@ZW(={?9%BE0N#-;MHB%Lb*Wh
zvrTmU_I$a#v~(=q5!=|4t+!MN;PzUqNdB}ec!)KEeF2R6F8A9?1Z{1Wl$Yb}FmnSH
zM`b<i5MbEX*VmG)Ajb$!_T}k;+#<+b%10dnnS46mUX4@4_|?#7UGR2JNK)CD!%1M)
zsgjbC85I>Ww#mLWhlNe_qdnm^O*fY$YGF(>;LF7Z-tbYEt#Oyi{QUnI@>QF=*x5}I
zhXD7vU?`?$Bq75*bpTH204unbL+Uug=EL=Mj+7VX_&|(m*s)%xq2W)DBKh+3^Q+B=
zWbVt!$}R?N&u`BzlQAoU-qK2}a2j0;zkApWkc;A7TU%QL0{R9TowC%fgm>I{z|7qc
zTSLu!Hn9^TP|!4iK|Wx0ArFnOA=L<*mSteEB9E9@r)leL(-ew?6R`w4`|aDe+is1g
z4Ym;0)}z_V^N{AI8qS?$kZb7aFH0qsI!Xj?t^!-2`EfxmbE8sF+_0kj*WbLN6yOzM
z#gQP6ty*llo}ySSNP(+FrK8b1lK6^k+uML3N%$^~&Xf-Yu1%^b#n-XP<vIKwlZJIS
z(s?X65OMmNxOG$1ZNI5Znfa<E2Lw}e)XS#aPMM7jsDw~>&C`EA@|}*kKIOb$K3o`C
zg5FdA8Ato4<n8|)_}?eG|6|wchAXN)LApS29x#5ekzV+YkJZ)H6e-8PypqyVPg41k
zV_gWV%*IUdhMzN-+k16O9kQfdN9xS?!gNRf+F%CJjV+Bxewwo&a<qth=fcCCd%M8P
z`Qpx7s$wg~<9i5rYU9X$Ue(d>mf`v2zhK;roeFz~?YrN3(yOc*BU%z=irblX!2N4(
z{8IT4!pa$wt(Ud@73eCe!&Z=OxHUm~qLPAbE^b#OaV){7$|>(J26LRd^{TvvmLDf!
zt4+XsLIw}fCb@@`FzekZZGY`P;I=jvJAvM&q1^{u>oc>nr>@SMR{6LZFabRXc7Q9y
zilURt%E}V&f8MQ;jOX(M)xv-WbP7~X_D@vJSd8iVU2s9vq~4(+M&x7vw~D)mg|Hvr
zin8s6Uvc?Aucbzg)@;l*>jW-F0uitIJ7^9++`CvGv{_%xjhvp72^aej!s)g?m6es2
zb`tons0jqa2BobLB}40rFywvAGtmdhN<o3x%IO?fS>AA^Egy5>ky*vX5?owd#-$Dh
z-(GShum>(Y!kia^GiJ)3F|OjbD0sJTM{SW{#?0+wAuB5z!G{ndkP2$ndke<jJ{QSZ
zPle$foQIh2gDk+ra-^e_J}!-yF)eqVEt}1afP<4Ru*M_dN{SFy(S&8v(AM%f)3Q8A
zTxqG4j{L7Hp6?)C2McCj3|9K}RA+QDJ$$aReTCYs(|$H9F@l1{iERJk*?t|VqbL(8
zpo9I_8!GKwhYfG_u31WINl{X2T(S)Yv1P?}x+hyBLb~Qxm~n~SP0uW3Gt6^VS?&Xq
zPnvH5MI~W$Efn1fr|2lq^br*V0<iV!!Vd$kKVi0hxj#|K(k|p{b3-n8x;QX#h{&^x
z3UKI#5SZ2S@-mi?XI$I`S<Y#wt$>yD!stwK`y8bLEIjp;LM5|LN9v{iLQ6HmavS;)
zRhM5Yz6X~Lr>8dg9jx7Z_?$ZM@d&GLFOGuoI_jdgy9C13UL1LC8Lm*~cL2;(`4II?
zaF$j=A8iQ<+3w>lik!RP{PoLU**{f+*_f9Uz`o@g6|}*snn$N0*Cu*V9AAMJ6&n*1
z!`EIvpP>;c?ltrM#c>&NfyDclO-xO5W%*zr)$l7rK)06k*u%Xd4T&=D5qwJPo#BhE
zE`bt8=Dya3*xPZK<$i&b?ahf?#WTnWs^N1C<bGSFjsA0Hm6AZcu0;&ouVXzb<)BY6
z;xWU}dgH|IT()UXHXGDHmViT0k#Vy_g?&uR#ZQm043B`=kcZPexR-IUjn=<ycm0c_
zk&r9-G(e<@?aeisP?-1CA*>^A=oy44c~p)=%nPvA;*i(^=oEDOZh^#C@7}$u8sVB(
zkGS{!!^w#xZ*y(A*^(S~L>sKW`Aea8BH7<LM%Xk)Pg~|RkoStU@dOEr+F-uMC;NHD
zK_eOn+NK-u9E9P1=|3fESy@?)KUiEUa~_P(H!UB>%=@DWME0Q7Iz&r?*pxdBQVXiX
z^iT>z;HTOLjOY@V&DCFrc(U^Ig`Xc5ZZh*8jBaad`)yn=j!Ja(^z@Y0fY=2wfeQgd
z@3azP(!&)iN%Jc<NLm-4@j@9jbJQ$62CGp-ZcQYS-21I#$GWbGwxa31=+heGZAl4f
zDxo$%K0gNir*)>kSmOS0?e)Zqg@0%LGzbj2+2B4a1MWIDEpJ`}v_P~%nWYv9_xb7-
z8&a85e~G=`9K;U53JUOy_Rh|2akj~ESDzm~tx-ULjXI;I#ke|^m7)zOsk}LRh3Hto
zr(tGhhIj-R5w<xk(rUEfwQ+@;HO$bz8e)yIjDz29DnOq%;FVZX4T7}dB<*#V;aq?|
zis`Ch(`AZ23&VWvDGGD5e*Y?5oxA<<gX2cM(w6Ho*?-<V@G`9E`g8N03$XFUzDwgX
z`0cG~ZmW&uSz`#G$^LT{iSYlnf`#SEeo&BcPAgvYoe-zu<h-5BF)AS}Q~e8Y393Kt
zP;GOpXgjQXWzGD`<M+p<oo@qxjN*NH@$K6Y%nEL(waRyC@u^P0DHnedsdAZ}M^T<a
zD41U+|2vSjRMC#XI&Q!OTWuAW;+`oo!!77WUsq5{TdN~Sl>!6U6sR2mvySL%7<v%T
z&mKLUR*<mb(3?Lw6counzdm<gYL%XNCWM_ljMotNqfX|Tl;)7=m5u~#>*}u&JtHFm
zg_hJyAWWFe`@z4#M_OS4{5NIbMw3GjB%g~^o>`h`Z%3}xbT%s^EiKJ)_UCiaHlv!w
z=$1H%EVX*S@i^PXFqZ({pcU_+mA4W_8sZ+49oOH8og|=xFrc>=yN-Oj?K>8Y0BwW5
z-}`Jffpva1_!#u92_yWv0R-t+d?1-NEzgp0eIz2WXee;Q8XUX8C(nPrS~_TZC1^<a
zE&799=UbLzl4nTeF71FWmu6jp<Thr@(QUGnJm;YwBofWcxX4Q5-SxPLYogVX;EH6S
z8Xgqtd$Fftg(69G%_F~lGIn~IETENAg-~siuM7iST@jG{Jusym+bqMNY0gxyko&8A
z-8}^`(TL1_<~fR7O3*|Z1?;9qoz-*eH87fW!~FXBRc0z9(oYI;TYb~7IuB|VKJS%)
zHAZ5RZIY)Um}v36{-V5Fg{Kg&%r7U$Lqu8mng=I*Ai>(+sS6oH6h~M__sw7RAmhj{
zHt;o`3_Tvf%o8AH`Rzsazj!*1rcleP`l4aBd?-1LQ+m9xp1T>AsvD*be&NuL*o8r%
zDG-X2Dj`-4uTJ*H0|!8@w>a-y@WyWp4)~0O`L6y5?Ix4ABIYaT%Orm(5Y$onddR$|
z-X$z<;I}u{)(Lp>EF~Ncq$-i}vy_KC31%{LA4k^jB9)TK0}yb`Aun$Yvu~w}+Q3Ez
zE9$wT@Uu=X&D}amgB-;$N;w6x=0r<eBm~yX($Z4V@!#V*)BS_nWhN4_Sqk1B9;3Nv
zx<h<Ig8NKi{Wz<OkIFGlbikY3f||<8NF?2eLb?e8O!E%7kmJVt^p~0u_sCT;K%gcQ
z8j<NegS@pu_IE+I!_`rkoS*)|6RinG(IBg%YIC#I7zW5>GA&-p(U3?aR?d>vN@1}R
zA&p=Zq+y7DhFYvLI~bCj*UTi5Tg;-ixhAM9L!&wYiP<@MY7A8tpB{R6c<^Z;vCI)$
z!B3-!VitYAJlQX^{M|BF`Xq9dOW`|wh$Z%BN}?dGN!Yf{6hDBEQ$_VR_vJyAC{uJC
zehpSMa7F4qwaM!Lto)N^#8Bl)>654R=>6X|ZgH$x!u97DC^Ogs6foPU595rrZjLc<
z`WO7fxcaM}rybnbeKIX{+!4KDGELdu*q4>;=YSXz!~p#Jf3csi=&jFJIn1KgmL_Z=
z%WFhH`2t(|T(ssQ5-(xqiRa9#KKkkXUVDQ8rqyqpE-k>ay(<C%!cCNsoGHCG;8|`s
z0jRRHv;?R2#?-E_17M~(ASwAhfHR+_sIwK*Ik{dHH9o(7CVF&fRR1m)=H$cfh*m9p
ze!;CE3zWUp*fuE|Fbg2AbFiZgP(N{_<d@sfrhESl>G^#BaL0G_wHksojt6c2(um*<
zmlxM?Yd9G?`iYqb>JFAU|3D9keE+F)AZb{m$ca=o;6-veDsH7tUW_RpeZu>)fxa@7
zl9duzNn8R`QZ~>(UaO<HX2mAIP@CxU^XcGn&@dxU!`8-piW0qB3Dci-1@=-|7>1?L
zXKWFE3zk_y+JfLc3TdLTReC7Vv7oxlfe~3uDg6NUg)nLI8>swbRa8`LJUGVb-g3Tv
zRB3xjX=%<IF9(aCNBl=V-vZPV`GApm&W#?-$ktHDt~8LFhS(Gl;>AkUghB<a&(ynO
zYxoiSIJm|1zwQcp2j<sTjtTg<O+fM2)9XpHQjqyH_4M=*AR{+lB?0T$=u|$~$59OP
zs&&r27+eqV8tiE-N>96Spu!!O&wBi#7!vDoWEwVy6~TACz?B3Vfmkd81hyG#ibh~L
zf=gcWs*NwyvFpGMN^O7RSyDbqJEF3(^0?n*nhdo=C0B+WE<k-cxMvs=r@;2I%XUn(
zkkPy9YWnFRcL>@m4q}MmyaQB^FdeWq2?QN-O1GVEhXXVF;fZxH(|Q#nFXhDTyE1Ct
zAW?4p%GV*rDkwr6_a1P?oJU9rH4bd==a;95__J~MFnls80U!;K<f2*-Rjt1U<Bx8E
zO+#Vcf+5c1=4%2*IENZO+^KT|L}WdLS<%AKpUbUJ=N4d_f??bZRyZ9}N8E3dD{>yJ
z%rh;=POBjHQEG7f@4(Z-fHHt$PrEge2qqHU-8wEYkA7LH#?|9V9bzY3d?{b9CtOfK
zkU`y92H=@rq3lZ^e0E64?dJwLx;%^&rBTnv%YG}CkqMJoj~^l@Z`R{Q8<jTcL*uz<
z9Ix7O*DE~Qdr?&;5bJo>8q@&Z#~_wPnJ{vjEhQ}bT|~Xt_xJZ#Tz#-IKGWSd=Hx=A
zj~rKUbqsX*+!ruea-ewAHr%G5dO3&u3Z<+CF_kB9^#^uxaLpDNP|Zs9y!BAEG)*uN
zegFIS(T+}2cSxAy^!cEmdG=kX9({;z6B~hpj4aAQWq!2j7xEUSy)V6o7bc2qTH|xL
z*irA3p5d8SGl2#o&g%;~EZ6lS>bn!@LfHkBJmRlo({ezJZ<dZQph}BP8RT8aRV5`Q
ze7;*slHOgx&_nQdF97h7s6xwn8^CrlKmIz<MI@T^o`ls-Ee9J&aIQSdar^$IamXor
zYn|bNQ7(Z?2}tdy*lnx{L9A<=6G~{>CFBHY!L4H%?Sae1#+4rS-ymS}RR9k~C=(L=
zH6kFz$RbrFz^}P-XJN0Ym%rKu^MK7(O=-DI6U_lVv|Su|ft`in^>vqF#H2T{2YE;X
zkjqaO_zzNk8-#ha?7VNge`Q|^48(3cVo1gJR|>}kskpVTM6R!oejL(K+zf3@H4A8U
z0iiCi=AL#9X7FmkTq>$+nyGXyl6SGjB>WOj&c$Hc*7)egNCB{+w#LUC=5PKNHW9Qn
z5tQ7mEtfz2Xvc>MQXxjgt3<IoE68%AOrR?+HgJ!Jn%bg3P~_(HQ>%{J3C7iLK{|n(
z#a44eGwW|=<&vA1o!V)gUx2rSkw^{-J1}`aQHp@k5oD6+)^s+zpr9QISqzL!d>X32
z`4G>BXjcH7c4n>bTuMST`=(RMR@fmWby=@pL!v?ic^>?{kEz>_`pC-4(RQWCbLcqv
z>XZF0w^Cj*p?YosqUz{)wR7*`(dX#RyC5}oBm85zj%Q2E#kbj17mbcC;9v^QF<BX1
z+V-b@nP}hQZyZXFy+@8<>Bo7wpKcsJ8ETzEf0;MC5B1oQIPfYjhjtKTO~%!opx*{`
z*8Cc&UvW5sn!F~O_=EoGQ6nyBel2C<c7*(*o!f|#Kp2kh@|8^5PF)%Gez|^mmI-wt
zNeA1$j8ihBoS#&Y+B8x?RaLV2TP{|9YF9#09*k+Is1)6?bX#<79KKYV%!ZnD#2PTA
z9tjI;#NUoOzO-z$@iL=YA#Uw~RbQy1!co+M!+oYB!Ea}zCF3y8`B@<5eu6@gxn6Pr
z|Jql}(&Qbe$#CfVC^#El;J<a!sJ95C=EU=ICrV$ZV22OX$b6@>vvQ>NqNv0v0shbz
z+Ol%F?p4f2e9YlFZH1^{A~7FjUK8L89SLv7$H!TC6ssJ!Ke~<-!Jg%?GNYpXRIa(z
zE{qDH=gt1`_TrZwjO*z=9U7KT4nQf->Id8;00~cqRhm%YS-+vU)~GZ7h-cG+K+Mic
zjz?lzu3>)D_rfhlhG61qaX5dhuzDMvVl{BGF?i3G)5<{1sNBkF=5Xzh`Ggjkl{M(_
zSq^}R_`d}ply@@CMQ`@zjXuqa?8RS~CHimSoMs!J&?6w)2^Y-cjzTO%1K}SDlePhY
z9(P-bPVGk2{5&HS24$(LK>P$cT6TQ4H)SX4ef6_>2IqVCu1_=%=o^604bqi4D}ro$
z4P8+>apI?q%shKV2ppp3iILi}*>Hv5AsT=>;z*t8YPM5dnKG2T;w~~d8FHQV*Jp<i
z4aT{e)|Y2qT`6h<7YiPQi?MTFIEQqn*t8|a!Rje$xB>X!*I<(weI>s-{OG8J&2@uN
zB+-$<V<W8q2O=mMVK(xB2az0OVq((O)ulVT2GdIw^b!h2ivbd6IVzrE@;5H-&qvh&
zNp(Y>e3n>>3phmuJWtO3gq#pB7`TTD#1P4Jh#Y{BY(V(vMdWGe8W5zEmg3q+(?YqM
zOaZ$5BN^HTrvOB&T5ENgdv1(es2z~WeBcVMIACqD=FPvb0J9w!1avGmi?;7qJN~+Q
znMgT3Bn@1}oE=H3y}GU_klyfZdO8ZxAVk*8Y<ST;?BK}z$KZCN+l!)0KiGpV20^5W
z0Cpq?iW}xU2OQTo_{y*unX<_TxNdMP6G_}1JO%)NcOT%-K+yj`SOVJT$K$otdZlD8
zIWOv0d1Rn~V9g)vfGwLpKi*pmkny<xOyN-(H>+t1dZ5z?jq&_9-Eto2IEW4BHHz%J
z`G<CcG+E1?1|lPHfQ|pBV#%Gvu+^ZZ7y-s|+kdvyObH)>q)kHIZ{zr1vLG-+J@~-8
zW}fi7(mL~0u!h-(T;*%vi=daPcu2h<UYbEuk4+|U5mHSOn)AZFhJM?WL4rF11h4ve
z?53t(dQ=1*Vc1!SqE#~+0{_3uYyVJXnGwu1^Yp*3sqCMY7Lke@$QJEL#YrCU-ciK>
zAMyryWjrHNTN5;!hzwAp_Hu}{aCr~BMwBBUN;2z{>9|0AIRzdfD(xLZ6)al7<_J$a
zg4GB?bV7fT^)*FWwhl<wXAzAs!s<V*;eq-CG|z4L0rPJ_3U@}Dln`Z>U4B71;&-3O
z|7%IFTcS4fxSKktU!opPE~GMH>7f93ZP}>|zIJA}YuB#D$H!NaCH}3G>vvht=aAy*
z4nkRpOEX8GNH*;VRk=BYxl5OF5*2*6cawXVBVba6N7K>K5#bS4P+Zhm+@ya(#Hi((
zPmh=dG(i3IuO<%+43y&2(|NZS*ta`Ms{zlja9+sMOktb6W-41|-PRO@T)VbJ3*SOD
zVP`-c%pLk;0qKANUC`#!!<}oR0!ne$L@M|+y1swMd;!tVnN)2SbP42K=c-@doHxv$
zSPy^QnkwXUukRkpr5(|Tm`#!c7y!zCaNvPj@MA=UgL0-gs15hv;p;-#+o9bHSMAaq
zBb=R^+YGktS6x#KgfU=D2t7rtf!2{F+n{AzgRQ6q;;#`~o<?CEvMJ>}#GZPGW3|$Q
z9(u_@oB++pm~DVQuWMu!3w~imo$N%|L-~d$0c(~3*(nQ)#S$oTxo{Kcmgy^{*D|tO
z)w9_~2RC7+hr`DQqys)6B|yC2!u(Qhs3LNY;LE*|!HCWz&9l8)@BZ_tz<BB-wx%*C
zk~m<?#DK*{dnhwNbDzYW&wt$xqSNBUtu>)1GfJqSN_-+EC@_NR$CWC|;(R2E&dice
zJ3{YrImXoI_XjG~LW*^mk?rD%l+95AMEtY{x`P-sVhRWoG@Fl8@b?;TPl0Yb0b4wM
zxmgW#YG+#qe;lGiLN%Fv;YTP>GH|wlSc;aGRvf4g0F-+L{x<iTK7@nR=Kj|<f&No}
z{XYj_`EUI7&4e8#N141{=IPPn$KscoUtHYBi^TCi#OO`o(k^ZLl>$=xEISU&>h$=5
zo4kL5;%ApHm(iL3ev@)Of={U2IhO|1xYjWW#m}+%vvc5Ai;IH-6iby;{;XXPSNyPD
zvzGL!wUj#@T}gZT$=5;;0f@oL2jCz-KF~)yh`xm!G{Mt6p{X@k;aR~YxX*vwZ2Ui}
z1{;R-TdUV>KOfDdI4bPyFnULyw`$n}xD2-0!G3QbK96{tayGc8$kRVx0QD!*CXJbM
z-!`Z0-Xd{^3It)?^0R`8zX!7%jXEy>u}`nou|8(Du^>nf&f8K&F)AzF@A+;ueiJTg
zE*H4&v$#e=ZZ+hy2U-OGd)l!714mmoKm~1Y47I0hO{I8|2nde{MNuzHyR}|%xe)Z?
zlAOO|L}4Fel?4RewSfu6*hg%#?@3z~?h`hX9cjiN@85x(A>qF~!vZqR0Rip2rr(_}
zlhU@qqWI0l<3R19PxPO*DChj{d68NFdUO~Gok(bhN=q`rFXxDOh*Ln!GqSU@9fulT
z^CHX|dT3Vss?*Yb^sWVixqo|^a3RxO&p*F@_)v)_i~3@vNjL*lLW+GDM76P^W|8JY
zMz>$1w-#du`}zvdDL2zqJxS6?;Y9<|U;!Z^Apy5sUS2-2xzs*S+@ci(C4usWk4<4i
z`Q+Pge2Vc<>@=jD_t`nC!o-i~4Sq1^z+H=8?}`<N!aQjHlcx!soVf7s{r$$hw4{|F
zD5T7mxeV7nz;Pc(`52UJ+c|m#VDv!W?^=6Nj7RzB#4{~`+EN4%{kUh(P0<R)f6@(e
z?AWozr1^j>#|=c4R$P1?P8|oq!l0v~f|o0N{yiiKJxm0HXKC`=pkp^E48dFP-!W`L
zvKj@R+>UT6zNyJcbzu403bdY<BsuASw96L!ZkOHOo*ncJ(DSU>{<YGWENK?uacAg{
zOZ)Vdf8@M)_EStu{MoaQn}JW7pKu()zJEfQf7lYt$tKa}bLhyS30c`avUr9>jQ*Sb
zA3uI*Fg)}8Ow?SafBb+_SIl7(b$en)D`m|mQ7yvVt!AQoAF*mlVPb<z{7T535+8p|
z$elCi&lk=W*8R9~VxQgfSU|wTFW$6!Hw6}m60N`RZJgBLxr*C^D&)^eSJp|EHwJ+y
z4GON=pwu})Yt(daZY=Wa+#pWq^y#Z0!#sdqB(t55E*(02c)HgJ-3yW;zlJawsrsXy
z@Z7{ReN>EW<crL{!yg3Ne#yOl@j;Jnt;Mwf(D-mY_@M14P6*>beF=TDr?%RU1{KEY
z9*d+_n{EIhl3>27EA~@f$Oz?tC(~y;kbj{{4(&j2FY1KIKBV;rg>rp>Lb?-BPxhXd
z+I9HnEwK~%rUFhw{1pc>pNqcsTYlzcfE<?Hw?pi0LDg-)Riz9rl1jLijeTQ~MmNtZ
ze)7ISA~M3`NaBvLN&neH_7iN4^X$UWhjdSzsIIz2tQI2nBDb16G0C7xrLTqS8W`|$
zvP^|pcS#X68#Q!w-Rv~fU$LaOHUld}b=2$+-f#9GSYNm6OPPcHAhAO?rIGJ+FVoL>
z(&%UiCBl)=EAlkrD{);3Ar`@JuA)>u;;yHC!WI?r?c{VWzLjxvD6(=T>e-&mw{JU~
z?zGa=`@0^{fABPq$)1!XPz<oFts1nx(*kFYuq4-*!8K5`I~b01dT(E!N{dU|H08V~
zY<N~!%a6Ld(Z<nlC?{v+)9RPt+c@lp4?mV?c+-+6#0DHn9YP2~DeWA~KWTo}rClUs
z5x*RRMVfz@*7a~A<Ep6o5Jo7jwd=|Lu1DdCvYVO_y@G22X{T?0f=4qs@N949XZ{@?
zcOFz48NRRla0AMt-B9G?T0R*O8JTb>9;2A@=bwLC-FK3J3wJo|s(r%B@$%KyhFNyW
zm7;!?ET!_pb$k_rX_?mudWxPuYPaH}k_9ycMMaIAot=wgFdNHf_u04<x7s***)983
zizl=S2Wr5jf_;=xs`p$^U%@@(mpRVnIZHG!!zIf~5;%7Kp?~AXJ}-;5#>SgAHt#^2
zKDK;bS66p*X)P)~{+QQ&uZr8Yi%1s*YW5hz5xcAWks9U&M(_vv>i7ENfwJ4IT$mrk
zNzYuu9cNQ3_D@jY4!_R-jt$IIdE+X))AXu&ETVZ#cV%v{j+|9lX|lGq*1D%v&U;S(
z>eZ{aW;sJnhS%$HJYnx&I&GR9W%oMboNm!o&7zPw0`^*w`i0BLTHIx1I5K&n#im$S
zPcItgwXFEb;sdr^MnFIyuOl%jDSr^>9g2Q4sDB_Nzh1@l(W`*Z>wM3nEM6aBNA_sH
zDvIgnn>$A|)zxoYzWfk~`~Ky?Ks>rUXt^Y?*gaZ*lwJ6%LH)e+t(=Uz_!7MSVyH|(
zQCj?~0riVuIAJt-Q}xzgsYSZP5-q=Fe1^tI<{!2mlwa@E>zJ--`+FeSCrb^*)N$|d
z^Q+ngrqi7-M)ybQn`7KAI-e6GklXX4YR>N6KBB$Um}Qajy*c1+dvWu17k9Zu&v(==
zibmy3k&uCTq3%(PO(Wq&G50CqBP@1=wpKsOIkn1_m{Y7=2ghT9^PdphRf{5qq}@>V
zjL&>w;}N~~NJ^naYv;JCUrKn4<^!=uPx>#)h=q19!Nn7V&O{upPJU~UtJL@Eq)N%A
zbSHN5V@u2Ru$q9s##>wTRH=(l3cJX<LoE2a&3V28sgg4-+!Lu8UDF{7O<%L7eA}?E
z*d&X~Wh#Dgk9%UaGeqbwB2~M*aD(D350~QBn`LtY1KJsnK{cr4nvBL!4Tjc`u0u&z
zQp!0?kQ%{&V-pWD98s>fRM!*>jcIG0bi(eD?7{`aCr|JmN~<+%*Q`#9_8Z9<&o^|0
zw{RBys*7%|8VGBIzO3}@Zn&0irW&dGvg+3S*0<A7Az&%z{nR^z>M4rnOA#*lOXy5?
z#vQSme$U4u;%&}ue6x9Z=OC2|&#3ST2&5c*0i_k4qlsoBMxhiZHYQf%{64$bPQKjU
z9OYEYYA7svNd94gvfHf^LHg6A@-IVHxeeXc8Q<o!-a<k`dgkWl-@A2snI<g6svHw$
zz5M^S>lPdLz36@5Pv60?gU{TTO`0Z_=2m*mf<eM7VS;DQ?2;3k7$0{4V<!R^N31(7
zDk`d*Y3zW1N>baIb`_6@pFdwt0?}L6ZB+dcEjl{d4_Fu#+AYrV@~XhARlU&XGhK!U
znh3l+W}P~f4z;gb+xBEn5=i*zfM>pWumc*qr~pB|<bX|ag^0%mHG6yeM>hg0%_kBQ
zk6YW=gbdbnPqE-)*KT=XpIzfCzxIrJl=#Ww29(f{tLEh7XcGgWS|JGc?C<Xv6cXwl
zi$#l+lky?m+XKNgsR^in7rWADX6}V?E0{nYj<|D#SniSny>_arJ5i~OCjc!nPq<Z#
zx>fX4`>%Xh`L#iBYHxQjbXfx7&A{0BsNqJ)$<zGmPo2*;3ZL%9q08I2{IqB3%ELWQ
zicqRW`X2Z+xWltPmBFMzQ8a~VMD{vF_d4`~q9tY9b{uN;+MpM8P^Y@B%SpY}{@$#|
zB-MR=eIX(GLAi<s68Wha0%XrGBCs06eBget5}Mbp?e$Nw5PKgH5uq0{iiz9qNMlSe
z@VG80@_KP^rb=j?;}AuGwhVnL_8?XfKf0#ZY}4nZzn6JZ3~93@Beu|V=fPkGM#!oB
zYxF=?9>v??3d?Vx^qUS-bs|?8bBC)XTyW}wrUb#r#)Hd90JzJ+a0In)f5@GNS+C|<
zFX@kgWi>6L+j?<mXXg)XXs$hVU+WL<`6g9fda`19t6Z%vH@)ydRZ&q9(j3V|fuPH$
zar*S>V`3J;f$$2<Sz%!<Sl2%2+p;|x0wF(kIu^fa?ZXVD=6pD~!xO>h#*2K9WAuEd
zMzf>q4zn=nRECs$U700Ly@d{rxA2xvj}yAuuM{-dv-qI`c+JMKG2(Q*yl>HAQS)=S
z>Z4277$KEjLNTE{d=&hb^$iUTA9+Gh7LU8GdHuTl)Eb1ypbgTHr1QW<VUrSd2>IX{
z$<R-upa80~33NK8*Y}qiRa$?WDzL@k70#ceo`#jM!+>ikhkDOdOiV+O{Dy&nftk(%
zhzybtc`Lbx*U3I3tgXokrV%9$y@L6-3NPCf8v`Wo0okeF3RTw6$9u+lLX)|M6|Wep
zx9As`uWi=%Sc(?aKB0H{^5yhWnD)E<0A;w$RaHZfU4g<p1(2k!vom%Mza3Z&9by6s
z<s~KhVBDUJ!C_o-h9H5~IOFMF`6i{BP{OiXyjWeSp)<f)i8l7!B#wXn8A6bd{z$H{
zp%+GIV&Ju?J&DdADaS4(2+vGqIdFv6^FDv9L`Ma$ennqGDc=TBn_bee?;Pw{Hgras
zl-RvieR<rsYYB>HDfTcY&v4t8H)$YzPC?67H#jvF;IUxdNPk|0=;@Nb(kE*DdxBP<
zL(H)2qv`=K>4Z{qo!pblkAfZ_h@q^=;%;L<keE>?Vtzkq6Noa=v9aD=exY?~vFJc!
zgtVMpNr6_@NA-re2G2P}`=H5)*TmNHI)7x9c>(+ZA&O%;(AwHsi)l15F;SHOdAi*B
z#D{Lzi^p5%oU}DG>^-hHqfl>==xWDN(~Q*z-zsT-qM@CoH(2SZSMEIc`5!rZ+KF7@
zo+T(K_}O7djKx<z*@oFRtK0XTC<p^t5D!r$dlT@A?`RZ)BQ&ITie%md^WoR?QOGH*
zy>~%KO3DD}(Vh7NEKCL_CddAU99~{!Uh`zG<@1T!6B=0-{(_NFMMI|x26<EJcnhj+
zTzy81{iCd|MKy{gH8ulYl|T#&Cq`3ExL5XFA=7K*8|!qU$>CXXljwh+-QihUVbWzz
z2@vmK(pGOq3$r9y8eT?4?pXTelxB+S_j)G^u`KBnbe0ew=Zau^W+u=*U!F?m1Z{71
ziey120h`??o#ZaPm1tF5)Ele9TT#i$SSpqJVP!YdBU!JRH*W6k(R6}~$0(h#Nx0p}
z{A$gjgR@RRz(?v6FOIq08%K5@HZnWHM&r-w8Sta{#d12Yoar2oIr%PSNAU8q-|Jp=
zwhL2!crvGxS2A!jTykOHj(FT#e<B4MSfm2JThO;w<|%WW&wEa9284Etv8X(K`V`z#
zzkJI+XND-0N6*<2IA(P4Ou4*H1?{t|xo==yy(dGZGHKDffoCDKx8|!=_LP&hJ=J68
z0Hgb?W$x4P+o*keSDj$brQ75L-hb6O>_9hina=fe(fBmE`~_32lzNTXrQ@wkgxtxT
z%+!!xtDBV)22*V|w@G=a&xFDsbG=z=TJn59DmQzsfo0=}bFOJke0LG292d#Ro04Cl
zb@QRr1F{Y3i5LXsR5MrEjAk`;-4n#j$#1TmSk!Bah8q#WH#N$x(6m{$iQ~_vHDUge
z7rErh<>&wEF6MMr@|iK+`S3@zcRqi!y??Bp@X_g#GNZxZJAgY6@8y|CrJmgk0eEqd
z(W_W1>Gst0=$`F?6=kv1w@p9!{bC&nha`Qf@T;k16Y~yLpLF=DyY3sfxw!$x00j8z
zP6vr4b%Yn`CMww<c}Mq|g7Mj9tEA{2p&5EpuFx^omNl}a;EZXX#dJ1t!`Rq(4^v%2
ze6w@8w49{BJa=4cp`5rJN-_jC54CUPxDBq7&qFu%FKyPogiu*)Jb^IzFJJ(1Y@GLe
zAM~z%&S?~5$;`-*l)DIieHhPOIC*O>&>;FyQ@0bKCx-ibz_L-nYz_A=5FWRUpAi;z
zGPL^}l&>;AlsWz~^0F(u5`9+1T~kxDM-8KtX>lV(1ZUPCNAdWSdzZ@xrt$~!y#V1`
zdJ``(xA;}JnC||$Edd<tM*Wyu<Ct5v|H|B*`H_{4=#-S9ykV#vuo)N{hV<4nT1C{4
z*c<A&Uwy0-F~Wj=Y$MLXTD2H>C;+FGU?~5N5gqvcrgv4^sCJ>0tqkfnl{9I%TL^R1
zf@J%*N9*c#<#hy%Hte4z_I?Sr2<x?WZ=73hOdgILoQ+H;WBecLy!_G=S=OJ}*maa;
zSD*Xj6>BO^pF*xuW3E|M=4q=|t}8&Nx~mQe)IIq@c0AGco2qt<v(&vJ&Bl;s4cn6y
z>c|D)$=LFWtyHvt2s6I9H~CRT&c&GAhbpFrcCFUsHe1Cgd?NYSd2Nio=SZa#NEqL9
z<9kPG3PPtTP=eCi1Km^l@Cqs$O6csHdO^b84h0ecnW!s&LgEH~6VmHo;OAG_)vcpI
z?<rI1+RM``zl+Jhz&vj;;>HQ=`)Y+4{{XYr^~HEn*FzOS`6}Mc%9*)J<$F2{XVbT>
zo8%TAv(0aY8p_Kks}v*^xZ{;_cPhyQs{>L3fzncf#W+c$NH#L+4N-1paL1OWLDVQ0
z`#0KW_qD>rkhF^Nz{XE52U+x`?XoI!E^`L0yjnMtHCX<`B_l`{)v62u7KJ)?r|A3W
z^*KIUx2$fZ3PZ8?GYZ>X<5Eh!(g{XFM_3wOzrpV1m%qBY^DgT5Gd&4TH|8wy?G4~s
zU0jLCh-#a#Ygn!C-FTbu=J4((`!{v2kj@IfCS%l0wki+qQdjVo!iQ-mW%6jh65%>D
zHNCI-0*40A@tD5x$J{ITL+}Otuh#k1r9EQ&a3)!bsM)@npN~vR){IPwFco~+L7GEP
zN}a^7Psa|Zbu*E%&6T(REcDus8T;B#EE+d<qId6&$oSG@E-0iZ7<Bb1U*kQ`h9BC|
zd`{+4=meuycDyU>HQ}0Vc%`0#=*PJisu@fjk~_qRs;;jkmjdm3th-F-)%jI4uirE0
ze2`ialu|-9u&*a{nXQy)W<NO}bn#SvTg=?_$-zNAP03KQ(vO@y>Iu&?D{vW8)fqp|
z<{JtIFOM}R3Rlt!R_vB{N>Dp|yRD^+qb<S)IIMdgIAEf22H`Gm3QRW!-}^Ibh20`O
zNw$>Ppw@HL+tnWGmYV9PN3d{_X-^f6-4eR|!aIfw*>js^jFMdMvobLXt~%os(QALb
zv~lWKpm1DSW=8!y&Tz!Za>}vmJCC}wwzIL^qqgY7H7B(20ter1us^wud7>o~-Q_=|
zfxl?vmNHc!R4myfM03>$pMMpTC}6yLf$NF)rJYik*?Yp(0fnkIEH0%ZxBJHa%s;w~
z6`@-sep@-7&OPUx?{%&uteqUFT*tq4r8;EjRynr)Nka4+;<YJTv*WM1Nz~#kEu;3^
z=}Mc_yIfqAm+Y%W9b6Mom)}HlI*p*Wqxf%g#+Bvu-7rlF{4;KOAMV0I_1F!QeSwsz
z>0OV9x#~^Ghl9Q5m4BU#W7W$I88UflSne!Kk<FJ`AQ&ngHZk%vs90nR%r>ucw);R9
zyRJF!uW?1k=z{URWB1ppiQ8+>o|NtmM!hKebC<f8+9+`;Q}gl8*N*Ik*&j3)iKht3
z?fi%IuVNKYvxd%BlVx$@#2KF?v!jqs9g9cOJoRG^gI)L85XLHXE8qXTrJZnm?-Pzc
zEJ}0|KOIv#wdT^g!KAF9%DJU0cBJ7w;|~mDi7gULQGn7;pm-}c1MOP{r=DwdV*fw|
ze~4admZGHS<TNKdlAKI>PU{*s8f9NiH#<yx`j!`0vP_dRF!bOfW};j}Q7<NQ@}iRI
z*(DJSAidAlkKzY1jcNkT9!Fm#8U@@yYz3#Z6>b;%_!b9CPSr)zlyh`EvDzyxr|H>S
zv;F(L-!Sb4&rf!_#^$JVGp5QboXJ<Edp2LL>e2&?(vZxgg11d~>0=4uQrP+<)t+NQ
zf>^=6+9C5i@0Ea5^b3>se_R_K|5J72W_*zP)qb3{&O&3i(+9q(hB=)>Ec!yFD_4td
zCBfGlP!$+DzinUs5|SfmPq*!obGhcN6K%Muy@x5??9g4aJJHnc>r=$Au2HS7pLrd!
zD7y!Irf;2BySXauz8>KcZtnaxBkDop!4meO>AKp_RYEmux`U$v;{)foiV~bOJy)mS
z>Fr@+you%Ryz6<IBC3><%Dy$UFPg)aGD)VHR4@1wt8CKF8D?1>iLc~)pMCA&vHM@~
zYPj1coI4p=!f1Xyz0NMZmgl5dlEh<XzKfXoyq3J1L9K4iFy+^{fTPy!)|%@gO?4bf
zGWFi`T(NDJ_F&Ja+wpG6I~)et{m}e$l@#V>$(r+L?AHy?;w~?mWEKTeLwR*=o|dc*
zsCpM3I~=yiYE*z_G%c{xcr<8R>aa0u9O3f5l7{02nKmP;<aMXWNt-xXW2*|ulW(2k
zi;s8peN+4HX?lRVY)~*AFI8^0PE;!?KB31}Z07PJ&^UyJ^_rYcg7d2B;F5cjJ8i{<
zFzuft8bg|PUVNkK?f4~#O;5-13W=gFzq{*wF)zvHPBN9<eT0<fRj#_dv9Uh#+{kf|
z(>r$<H!E(+Rk>O3k|OC8amV59Yt_|(*V{5(fAW&dLx0|H3OI=m=N!AQihKCv1WH8$
z=V?!~ta21jl0~6?UR&kJoJ`Z3=$K6?ND)V&-%R#(ZDDu~PQ}}6#=6FLS>Mj(wsba8
zPT{dXh2qKIHAp5J#CMr3GH1F{GBKH2WfNlklGP+y{Kflsh4i1_&+x{pP7M}3^*U;2
zdmS@dL<^hYDDB(Uh#))qJ&Wr!Q=9r$%s=VvA1JEf=iMTHniQe!M=q=4I)-zIVDek1
z4CclrFTNT7#Mb}L$HegB=f2Zd<7EYloVXRw3Cn)Xnh;^Ef4_1yGP?9fc|?ZL{Kkz7
zMw-uG_6tcZDlx~~@bNdg)d_csFy=qeOWn_@srhoKK!LVqSN0{_%1*P=%rDR7{e*X$
z*Hi`nLvd7M+V1MW8{R=a>uUl|=&CUTvy!_n-m!hWO2L}tf6=3EoeHEL^V_`ES3>5v
zTKFd_>?Df!Lcp@lStU`Apzd!SPAl9IpG*&Gt>&o5S7>RLsVm8-`9;C^l3dZgX<azC
zwtA`JT$<k2Ry#WD?E4W*y%rL+YCJC`3xB)#vb$)2{^0`xy_a&%3H4+(JI8)Kd~Db3
ze8Kg**AJZdwm>>o_{k;`i#MF^(~|W$0bi_iwljZsp)hXz;N*+=?o)(`Q0KKoh7&v9
zZP5dx*PE%Gt3y+?Bo|6~oDg*gH`JX^ckW4xm@gTAo}v*E_`yN<5#4LB>gH5ik6qM@
zBMXg=<xObg93J^i_Stohr^^A$G2+{_ckAXYw#QBwc2tLa7n>HJ`q8g@GcD@p>$#M(
zVHL7OgOf34bCKUNG(_j312?0s)Q=6v2=34|_AJR|?0AWJRcgY+wV5k2@0P9V$(_xy
z<%DesG459MRaa|zIbF=Vt+z6e6O8}#aJAH>dqei4Z?tTRd!{T6f1TIG_*|V;9}zv2
zR`j{9Q-oca)ep3Kkyz?VX+6)jMF7>|(#BmW4fd2nCJtd1_m854y`AP=1uG~PgCht1
zi=NF57_F?XD-tHx6<NkZLi|{9=6bQml~^0c7blmEIJb4JMq8gtEslS6GB_6O2jzl<
zi(Y)6hXUN<9dF+%v&yEHCpJhmK#UIq?UKIj<Ou2E5|(jx-?L}WJS+(BF354ubDmh-
zRcj~rVAspZ+iusBukaUACclMqin_Nqw8q(sur|%wL>Z!V@qu(J=Wn<3gbhYT4eyEv
zl@oC<Uc9)VU~XWX7I9z=w-bSaPIsw)?e0<{Tzyw)aYG{|Adkn~)4+<T(h`5cDgMvm
zhq@<r3<byP3w_MI0d4bkzNRQ*HM~KNRuHeG3>T{82j&40i5|-DJCt5vN3%@xJS`yL
z*x}0|=p96&?%>&)<_fEhUt!&+0V(WhEQd(cL@!#ClTdz7Hh9rZ2JL_KQIb~}Pdn{1
z*C4#lE(I@~G}C{U>$>h`ifzQ+=&Ev|e$SK9)Yg7^`vHn5T-=}cSX3c5L9zOp++K0h
z*^4_5J+Abo*BuL(+Q-TzThjaTQQ1;%3O<#!VCZeSKcSp>elt3A-C^urra;tGyyHjm
z@h5ytpDrgH0o_T_xZ1>9ej(VaYBjONxt!%_s{^l+dwGPMcb}8=(A$Yv;n#O0U4lw<
zT&jw$FQw3*1}@jO8#qN?A79uW+*|B+@IBjm_aW!_?JogZN#do4gm!k22QE&j$|PxP
zVeCToEbSdDlHu+1vbT;~<{B(WKC0kBD8bPt4}`1cIb9*VznM-YlbSt-T9|45(xFz~
zs>=t*RE$Yc@ts=b1M<du1B-O|56}&i`Ho(~VqJQmFFzfcNgScl7BAy&Q_?(j10+OX
zHGMPm%|LvWx#%ZpZ78-u%iQ$`5+7p{&C#eVJwwBHO-)Tt%8bp-gv0CWhVt3+-oCv&
zHa523DY9KonevwTh*O(QaidM~RUAu+1k?auKWIml{F&8#HcP@yq4o*87tYKJr$Dqu
z*)g?cBR~W7W#?~N^dD8&{tb!%3qYZMLe&ZRxsPL4Z3M8H5};)tk`tn%5B=Kx`(+$d
zzUUm|eFKB&5OyAHF1AuLGcz+Qz}eHY6uxOj7Tz(q@%Zs$C;gMiMNa(rQxRpFZ)oQF
z^^olS;b167ujwgzL(?YGjW2ok%Pz3J@|%y5vBBz3Hc;m@bd=~N0g5Sy&)rIW#LRwS
zJ@Mhg9SZO^oSSTt3Dn3rjga>iP<Cf`eBcYOPX~oGJ1_4#d>er?RR+Xq128*o-_iZF
z$<FjJA*6DLDv8?9!t|(9<kaC+P)fOE^-as}b;36v(4Z~aZgI2`QO4j+<XmXm%7v0W
z5j3RRB1R?6NNXiZ@_<6E)VagY&HbiMYjnX&fBM`058&mk8z3FEE36#<fiy{?h-Y`)
zSAe2$AAI*jDM+#y=#-Dhm?f1vw#G|Ujz(OA5~=jnip$8@+PI$wClAaF;*3DMiGsrI
zojYk=-QAwl$(v3CA8*3zpU=H~k%7SySJ)SJBl<er?}Kyl5J=rxNwU!kdy%8F#}#VB
z&WMN@hI1=Gh3o`w5DV2c=<U?*6!|@%V+R!_ed_%Ao9*&TFZiaQvGmj?^r4`XnU&cW
z&K{TbP&cdeczx$cr^xRI?YR7}b})PpX$<|1P5+%iAP0g!{!R7%KiSk42i-$_H6y9O
Q0=bmNRqZR-%N7s+FIcy8kN^Mx

diff --git a/docs/articles/vignette-08-nonslendr-tskit.html b/docs/articles/vignette-08-nonslendr-tskit.html
index c2da2a340..299f24998 100644
--- a/docs/articles/vignette-08-nonslendr-tskit.html
+++ b/docs/articles/vignette-08-nonslendr-tskit.html
@@ -33,7 +33,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -189,7 +189,7 @@ <h4 id="non-spatial-slim-tree-sequences">Non-spatial SLiM tree sequences<a class
 }
 
 6000 late() {
-    sim.treeSeqOutput("/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmpx99kMg/filee61c7b568e93");
+    sim.treeSeqOutput("/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmpxdy6r9/file1112e7d444348");
 }</code></pre>
 <p>After we run this script in SLiM, we can use <em>slendr</em> to load
 the output tree sequence (saved to
@@ -394,7 +394,7 @@ <h4 id="spatial-slim-non-slendr-tree-sequences">Spatial SLiM (non-slendr) tree s
 }
 
 10000 late() {
-    sim.treeSeqOutput("/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmpx99kMg/filee61c27e4c5b7");
+    sim.treeSeqOutput("/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmpxdy6r9/file1112e332bd700");
 }</code></pre>
 <p>We can then load and simplify the output tree sequence in just as we
 did above in this vignette (or anywhere in the <em>slendr</em>
diff --git a/docs/articles/vignette-09-paper.html b/docs/articles/vignette-09-paper.html
index e64ab3643..2c23ebcb3 100644
--- a/docs/articles/vignette-09-paper.html
+++ b/docs/articles/vignette-09-paper.html
@@ -33,7 +33,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -359,7 +359,7 @@ <h3 id="script-from-panel-a-2">Script from panel A<a class="anchor" aria-label="
 <div class="sourceCode" id="cb7"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="va">map</span> <span class="op">&lt;-</span> <span class="fu"><a href="../reference/world.html">world</a></span><span class="op">(</span>xrange <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="op">-</span><span class="fl">13</span>, <span class="fl">70</span><span class="op">)</span>, yrange <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">18</span>, <span class="fl">65</span><span class="op">)</span>, crs <span class="op">=</span> <span class="fl">3035</span><span class="op">)</span></span>
 <span><span class="co">#&gt; Reading layer `ne_110m_land' from data source </span></span>
-<span><span class="co">#&gt;   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpdblWjc/naturalearth/ne_110m_land.shp' </span></span>
+<span><span class="co">#&gt;   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpHixcFC/naturalearth/ne_110m_land.shp' </span></span>
 <span><span class="co">#&gt;   using driver `ESRI Shapefile'</span></span>
 <span><span class="co">#&gt; Simple feature collection with 127 features and 3 fields</span></span>
 <span><span class="co">#&gt; Geometry type: POLYGON</span></span>
@@ -395,7 +395,7 @@ <h3 id="script-from-panel-a-2">Script from panel A<a class="anchor" aria-label="
 <div class="sourceCode" id="cb8"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span><span class="va">map</span> <span class="op">&lt;-</span> <span class="fu"><a href="../reference/world.html">world</a></span><span class="op">(</span>xrange <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="op">-</span><span class="fl">13</span>, <span class="fl">70</span><span class="op">)</span>, yrange <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">18</span>, <span class="fl">65</span><span class="op">)</span>, crs <span class="op">=</span> <span class="fl">3035</span><span class="op">)</span></span>
 <span><span class="co">#&gt; Reading layer `ne_110m_land' from data source </span></span>
-<span><span class="co">#&gt;   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpdblWjc/naturalearth/ne_110m_land.shp' </span></span>
+<span><span class="co">#&gt;   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpHixcFC/naturalearth/ne_110m_land.shp' </span></span>
 <span><span class="co">#&gt;   using driver `ESRI Shapefile'</span></span>
 <span><span class="co">#&gt; Simple feature collection with 127 features and 3 fields</span></span>
 <span><span class="co">#&gt; Geometry type: POLYGON</span></span>
@@ -638,22 +638,22 @@ <h2 id="run-time-of-each-code-example-from-the-paper">Run time of each code exam
 <tbody>
 <tr class="odd">
 <td align="left">ex1</td>
-<td align="right">9.1276798</td>
+<td align="right">5.7988890</td>
 <td align="left">mins</td>
 </tr>
 <tr class="even">
 <td align="left">ex2</td>
-<td align="right">13.1026733</td>
+<td align="right">11.4464792</td>
 <td align="left">mins</td>
 </tr>
 <tr class="odd">
 <td align="left">ex3</td>
-<td align="right">2.2592016</td>
+<td align="right">1.9846990</td>
 <td align="left">mins</td>
 </tr>
 <tr class="even">
 <td align="left">ex4</td>
-<td align="right">0.6743221</td>
+<td align="right">0.5832839</td>
 <td align="left">secs</td>
 </tr>
 </tbody>
diff --git a/docs/articles/vignette-09-paper_files/figure-html/figure_ex1-1.png b/docs/articles/vignette-09-paper_files/figure-html/figure_ex1-1.png
index d8a34f00fba478f47179c7ee943325b61d8d7af9..e404dc885824fdcb80bcc926fc92f15b1bf986fe 100644
GIT binary patch
literal 60194
zcmeFZWl)@5*EQHklmH<?f+x5`aEAnK++6|$*WfNO5;VBGYvXPSA%x)W?lkVsoQCI}
z`o`ZMGgVVH)0Mlr>AvJ#=WJPf?RD<&3UU(Yj|m<_AP{t@q^J@Eg31knAoo4I55934
zZP5a6_YI{bM8P|For%=@fVW3>l9~<>$TLU8e@HVXq^=OiO9)i-jf!jH=Cq3^w$pX{
z-pRHJS7ZO%mlY2K*t5Q&V_>|Zqx(oBhON*Sn10H4g#Bi(M?|srezLniJJz~8=TjWM
zF_Zu)*@umf9ziAFa%y|~ppJ?@`%9J3W#uYB8tV4mq={|zj3J$B$Fq61^dSV2`qiyb
z57y5Cf%w*STR0-!ju0{|)^hjhBMH)j+piwFqY~YIc)5mUcl*INRpS5TAXVx)SEcas
zq2m}|6;udBPFhn_)1I4%divS3XY7t^3SRpi)cfv@;xRhbq~51#;Smu>1Jca+x?mFD
z!19s&We2#S$4ZlpestiQ#N>d$K>0diepdzQXa>a;K`(aurN60|JdYngj*N@TOH3rn
zkVb;MTvO$kTce^~z+Bamw$N*|2jiqOy}mrts<qGRHwJ%2Jj`40GNR}9Ny^MA7^4Oc
zE&t|?->=qIuXdYi?Th2BL0=@43Y%$O=)hF1gVWh^452C+7||;M0o}ptqs(i8MPulI
z*lP0P=v7Fcf^J_%&UW329*50T4VBcBZ~3T17*MV#-kH2v37bpR0k-gu9t|`Q2utFy
zmUC)ew~4ZnQh2iK!lSIrOv&((VqHs7b6_mNp+e`9Mz$}~9_jL04DVWElagec`M1IZ
z9|!85&$aJ3eHdkOzxaeX+~jpZkYDAV|IALOFr0&%qJ!MLBy%T(luNZkC8}tc865(d
z6P3#ute!2@WJ94xg0xHp1OzyC=7zb$|9SJ4>v>k5^A>nF@N(qlB)~|EVLqpdxPg?y
zs`uqoN9JmZhOUvJ=HbUNyj%@hIKTi@-t{I6xX<}L@NDOran0}>hCtrg6sW`0M#a$!
zh;1&j)ek6Y5B4j(E{=T<@yxF86V20Unl5sctFjVj9oBbW%ZHz(1G9OmLPk#B*+4d8
zSV}`n3l+LLjG)M~oozyxRO0XJtBYX<*2^Kh^E6GXpyQocpA9thIop;G9i$=6{MSyJ
zc@BHwT|BRVA)=yJPwKAVl=?WXCb`g2Wah)OP#|{h4-eN+3cmZZof6tjX<GQ?Wu`A)
zBv33zTP%*!0uvY4^*S1M-z;qpVnTsHVm}KZNd#p^Dz%qOZFx{M>TI|sdh36~hCo_u
z@7=oxjIybFagT<YTEb=4t9mCM8bL)%P2Ju?Ctah%)Sbkq-s+DzYF#~@JzStpsQU~8
z$szVZ4!QQ>)_c{VX;Gh}qMe!9GfWTGBC+TCG+raHEj>2&g=`XUW!IS#d}P=}b2(nn
ztB#)}bfq^1mXqSG)fvq=crxo_>!=CNQM-J^eXZ5g2#lX}T4gvLVZ#gdwQ}w{llBe{
zN9#E$8(#@+egIn%75yw8N>X$W4fll>9vDW#zv8Z81tk}u3ljQuCg>mUQLuy?K@AQ3
z)8I*rdz?1M`co~WHpfauzJLD?oUzFP3WZ9^%g0P^MR%(?k8Fd(RwsMAx>PhYG>V2F
zLxj!0FvTRFZg?~vcTy2EF$}qV4?db}n;tILRL+nt2>*AVNd0D>TTaRCxgK+P#mZ@u
z$iThu*w4UaFLCklHBI8B+-#L<H_JLBfBaBLFix6l@k6ngX>f~k2n!1f&rDAz6B8AU
zG$-V;Nd)t;Q=^z6&Y@2MNwuMK{#Wj1^rd@dnvXKHdWti2v_Pn_PDSX-UDDtc=P(-D
zs41TdygaGC((|vd?m(I5SyJ89zLj(gHE;T*v*!D)zK&kZXD_1AxusyRQk7+Y!tlf3
zOwXvopRb_R-+X}?sQ7`Wp4=)fX_QUi)T(ndXH-)u6>2ITfivDDK%oQz?iIE^U}fU@
zTx`%)??H0m4HBXGr^A{!RrQ=!(NomdABDpQj!B2)n@I+uinPu5Qt989YaCuRp=|7J
zJe$7gU)AC=5j{)d4P4SY3kewYQTMKIOYdvXIlE9qPG?1u2+yp_3Q6JXsI-ChO2^w>
zp5BL6Z=8&FAcMhuAGT2abRT`F*1>ezbupNL_5HWa<Nau)0$p@M&LKaP2mcc3X=r}5
z87jKEo<16#^}g6@x~iUWmmbW<KahP1nLCV5b`lMD^tz}j(_{$GwE0e%LB>VW6W$Jw
zIyyajrx@pJR88R{IE>QH1)`u7aOxSTgkY>{ATDEMSeT+oV{j6mOTKcBTp8*(*H?(|
zphQ8<3+O<2ur~W{7u)+vjf(5JS5T$SvDj2JuFQY(ion){c@`EH5JP6vEE8@0imq&9
zQ*6=|o7Q}N6v~{60%_szu^iAST3kF_iGD&UlaTg1SJJxUUo?e)S4YZ|;$n+SH(3T?
z@BdzZSP)%aS~6SfON(bQ`1mJ;xS+g;w{xuw0?GZ+ZP>4#@zuh!+5yJ^Q&IdFD`_^f
zbf(G1x7cj@-`aluXDmdb8eX#Fj*O4b9~x3Yq=wXibYcvF@tmCG&B4JnCFGB{2SFf1
zZ}hik3V4KUSn7hbMDFsLF!2x>r+6?KIr4;Q@?c)cQ*GgkC)mgs;Hto%QVA!jRZH88
z$g25XUJ=vL46U?|^VWMeA_CxlQsmH^;0!X3(?U*ytDcL0f@&&7O~)hx`S9y%9ZVg^
zWFk(7khynZig(w`Z9Zub?)9pcn_gv7S^WEO{OSpH(|EbO4NSdg!4TnX9O+(@1o;)`
z4*{qEuiT76+`#Sa88MfnQO%3cmCoh8Q>W%hc+}*gD%|};*^ol8-`oFs64P>xA$Hz>
zYpfR7mj9i!{ol&%|NlY%Jq+~!vUzH8H0epi|6vId-TEg)2#ozWxpi!vmzQT>t7F^~
zHB(kz?$pUx{X&-&K$ftl8EC-Teq`?M+Tq~gU4(2oUqE5<P-xzIIB@l+ujat7*(|0#
zL~i^f(1E{U<Qy*BI+{hxn4Gebk_s@rMvHQOfEVl$PSJub9`s^<Qoq5qR6Ll#_#J?o
zFULymgN)9Sbb9KFPAWLH25$OaSB*$Xfr$m;(lRsmt|@`9f4scC85_76NMec+RI>X;
zogm=Yu@?RWZ!)X>5~QlDJxNEG2fgqpbv?z|Ca+~XZn+wp`Cnh|tAN{d-NyXbp0DVH
zOZK&n>sc{Uba7?jDJiygs$h6s2;cHuZ@Ad57moqY_3_F7(px?bpfNYRb9@926^%^?
z0a|9g0M<0nD|EvE8})dr4FVmNOl<0h#XtB8_~4<)5z*ouuXGI4@Q+s_@5{%Roa^Py
zs>iL|SK>^c<URp&kE5)DxMf$efHrgUWf=gtm8QKU92P^8dN6g8Pzii|d?`7(k-49r
zf|;I!;~ye`I7^ak)0sE|i!tB^#<QD~g|?GJ%QP4eM-tjLq|qv8ztEus$34Fr08@;#
zbN#f-RIx4}4lZsq&x&dpz(eeJKVF?54XW#VE%OQPG#peoZ)webP*e=VAmKMdpwgDi
zyEzFW719%MKin-UfBpXNQ*<J31&|^JOjEqFkx=d<BJ_QAFfCo5t8>A;4XzIt#0Z`(
zJ?{pOrPT2-q@cJ?3gD#}ib{)NDKuOLmNHNTy<-i!UC*I<7u)UxE_tPnh4gaM-jOsB
z{|cja4CD9ja}8R*-d6|ZN!RCFL+$B`W9t;qZ}#@$u$|Gt4b1v=YsKGhfBZkQVyuzd
zJ3S>KDPCAGoc*vq`a7DMk}_++t6qnRQKLi{QG7J$!Q4-_i^9ky{>|H+POJi02H=?f
zhYueJUK@UPKU`B>x!$Nif%wMXRy;BWJ&8Q~uKeMd;0%lHsoFBFPExD*A3sv{8r@av
zc2<);<X*pi9nWSOGG$YpEt`-HW=joRv6;gB{BRwt^x);$fvSp1=J)5+bPNnK({)a0
zl6)YtyZvu{1?Pu%U_ij(*=mYqM>yr<M~@UGCBGqHU#U_1##pJ5J#W(a;h=n_)i~?8
z6(>MbW`}G2@m#jabx!c?si`T8N-*w;H;8;7vkGvaz<lB;eE4{468OhZh54Y=iPPz>
z0XUC_fuYNViJe_Vsbf;pd2=l7_ixczM;xe>l$4UYd)<_cOXZ|Br*@os)77!|thb4Y
zNn9Op&EeZ1ryS4YkWJ#g_G)XgS{~RhEiJ89rA1U`_z~BP_l32-!UtB_c=rBuvA}<c
z7i$@zk>TO#eSHeV+;+mi0GJ0!@bPK$-XBUhO_Gz5{i~S%&`#Wx#~nVDtC*qe4XS86
ziw3cKzzKR1Y@6~PJ*P4o$b^Px0yJ$toUa=H+UV=R;2_TP=NSM^PwaRn$>r#|E#n}@
zQa$TkzvC$>IX-*l+Bn17K5WgrwUS^v68qYS-K6Wq(dM`^4-b#G7r2Llpg1B!4!T|L
zwO6_ySyWn2s)biLt}6kPn>w)=(=0d1b6D+(=l)=rb#b}h$;{!fA`=X}1FU!Pjwir7
zP5|Drj^bit{{?jpKYzo-mJfg+8<RfpqUysbhhEilHsJCBw=*_)29UfXhz92&c+7f*
zz<K*!h`<vN^Es>3HGe_FvjFbtKK}(z?jf;Og;`oqP!KEJcyp>&If>UPyv;Dr8DTKQ
zG6>w2Jy>Qejf0IX84ijkSXY;fuJgEY(=h=4lvGq|`LM!5hQ_PoN%y5NiGRe3Sa^6j
zJxP2Ba{{%_zmAUnAbd8*R&*}ZOkx0tFz#jnDl0qOmnPx_hHkOH)B&mm-kZw<O6d6b
zc-_g2htpQ&D7bPkAs3yHkiN**Cp>A^KR@1kjEOn=`}=bak7FxzVqTRa_~<CGQKu12
z)3f3Nn_CkPN&s{KM~_FDDp%*cb3KRirUs&@QY%meQ7Gj6WQQ3FZicWrEv@*4#YM}^
znTh=oZCi6Nx$(SC`5?}|bL4^ezFh3?2|w6tCe#(z#0`8)nqu-dBknddx0{rk0}lnB
zrD9-^#ys^SbEo-6Z+(6J4XF*7{}#glP@k|Am6Vh`?{V4~>hP5U-;Tfj2r?pgp$hy5
z={-IC58x8+BcHqd3GyaXOdJW~n+IYk;<3>Q*jehZAxI?-t8$E5r}5h<EQ+6CwQtc0
zM_`AI^Ye-u!+c`ux=vArYToK-sSLMaSBj9+st;t#sOV@aC|Cec1z1m7fVtZk$O;Tw
zM#}y81x<hO>LggXIdI>}hHpLSekYCv?DFc%*?@z5L$;4VTp`YmSAJ}4Y&<=#aNk`X
z2H}$gJPUMG*CmO>bN3B83IEtAr&-@ir>zOqyHvtiA_ch~t8*%#n&Og^zhsoydp+P3
zqh6FE&GfH-;KaK>$t7r*{vaB)N)qxW_=7G;0gl-+ap~jbrj{U$5^&q4T5%@8tT%@8
z+`;Ozn-5T+(4)@5Y8_10@`YAu1tFi9iff1RFa<U8K7RbT;c`E<=PqxTgWv>Ge8aEf
zS0Z1{+5&JYHL!d)ChC?=IcygmN$PVJ3{m_{YxFp|pD6@!Lb*$HKYl!&jVD|#)hYt$
z<n{zGS!G(tqbb95Zv1|aWFSN6x8c$e<n-(qeec>|$G~ki{?Wq<+X4_PTKX43y(D!+
zP!|9&q1EVK`6mdUyn3>z(sI;dFk4ou<egtMvwksfn2b(flvKB<S%z9rP*1zsW*Eet
zA(*}pRB5zWml9f}T|IJs1Rn;T2A&!MF}W=kWN2T#dPA>XtRg2D;=DCMnKPIN!Y1%K
z0$l=d?^m}gZc$oV8W0@JQTGJGQMen+?vZzRR!xca?YL2+?%INuzv2=e7+7|66cAYH
zkyLbv<~t~y%ceB#_@&=C_Wtc;WC4VrgkT9)x)YY2>pL4wmj6Xn>frbWVB>5?wJWO_
z)KBVa7$mflPw$mfoTOrts6De|@Q^?MkimX6|9D4fuT;cdZHiH0RWr~ii0P-wf8fs9
zN&*49*$+UzMI|OOBm~d)A)CZhZ9HB!xl%CjIQE8`#>_@HWl&6U-{_d{Cxy3>O<%-T
zywc@OWE>8K#l<5HG2Sf!m3(p`rqHDVAc8o6M^wJMfl=yvIMU-czJmJP2+=kR2Lu<p
zH}5?Xz`^aWIKPKXWz1XwT^Cn3Ai28$D=X_zjy!2bC**SF$X!)KW9=lDyndPqZYwmN
z^3|I#ZExCrJv>nAm|B3+dYXyRL-BFmf;?jp?^J<!_kGk%@s^;5aZP*geV2?*7W6xO
zgNVx*q=JROJ_vLlE6~=v?5J%_FIIT2Pljj4njF#ll#kNG1~br@n-tiRy`YR*6}f42
z-S9#+C6)Ib$QeRbKdd-u`6LMX)B~Py)$u}pe!K%Cna~JCX$goI8up5h>NF6l54&@3
zl1Aak#6(1|8eN*+m(z^S%JJ)Yow|MPfjjw)6rJhGCM4?((!G6~c_R~cU7z|M+h1fL
z*D^Croj@1g_uZZ8&4W01)X3PjxVIL}HKPCs9`~b-ER*hd+RVh)8|5b5`CF`h&t=Y6
zMR*<EV)?HwUsUc%s0B7O!fD~L5+{O=DY!k}jCS|#e?R{)@Ov`27+IL*Hg;w2b}~2r
zUA8|90h7Q4{ny(&3UEmf9U6DmiJ%VcXrwsOdgpX~vfA{XqVPFuLNLD44q6a6U03`T
zIT&=ua{yqh0)3<BhfSQSj_<Fz%s}^8MxV-I?ni}3v!l1~IXx4TW&sMs+u@vP7yw)u
z%ayJ;q}uxrrEmlLafZHW`%6opArGp1`yE<F59Jb+e}oJP+~FO7l5C}AWurQm=^Tm=
zqmz<~v~6nJ*20laUuF3pj&lv1Tl8={@Mf8sNUNVj$Srcgk!c>>!e`pg{{96uHDfMT
z)syn>ULb48#-P6RB_q#lo~0Dn)2C(9vX~fa{z*qTC@pUN`HPenQ+RQ0#frtfr-`>d
zKnsAS1VG#2fKZl1<e`zr-U11CP;S`74T7BN_I!QTCuYD!vY8~h!Fy_wJ#EQy2Q4!;
zH$E3bxa4kjSqy(aoV{LX3k2|;2BiQAnI=cW{e%CZY!$Xi6jMHRuvEBVCux?-=7;r4
zubol*uwgv)#GHSHXXd4n)>_e4kf)M&ZZP9@qMIQNL2jitpXwd>NtTzaOLr)_L-J$b
zAZwlYUt?CEawehowMS7j{zGP>N!a_0K!}~z?I0j@3Y$*ru!NkX!+7N0(f~pqf$M*x
zSh87)-4rqE8#lof0>ZH8NRZQ~w>$ujo!hj#!Dv~8{t|6*O_5?QduipNTl8UzIPJ<(
z;3_gCKLlaykX%}!6)B&|m(xQs?wgz84k<+W&ea8Va_Tygj+GaWGB-o-cSI?7yR;w+
zbKG5-V|i|mmhMp+rG9e<)7eHZT;$&eIY&Z{l6F;C1mt1>H(Td+)buyg0K`Ke`7jr+
zJbjV%IO~BTAt_v_bw(yxsLC2AaJXFM7q|5t$khsGXJ;0Dseb)gkn3{otmY)eva6L#
z*ToGaU)~@DnbQ)<X*HH=GhIhZMg}r4=V6lVWK|J@V}A#j|CXS;kheKrE&%|PQUa&-
zVe`$Eyc&FGw(JQ3yQFTt^Ds!nPYR#XEWZm|Dqcw^$hucT58+N?xkH|jw94!N)&vvr
zD20Q9&eIxjW2)^T#2g?GGV0Xy<v#P974w(mVaoH%IQLqv&9s}$gP}kYeQyWn7|p1i
zJ>TlDJ|+mNs3MKhbX<C5Im35G>su&(>+mg<0kOh<B#7!AP^YG#5Y@A7TRbdME6`z5
zXH-K?{QZM!ppM9Vq$hlxjtS<IFi1{pFLmH7T~MAmFB^O<3V9lNSD2cPl^D2##s#7$
zVLyq-q?1bzYbc8@K&E?Y5tVE~%{WbeWhC9f{@iDinQnb?LUcXO8K6jCI<h&i7};!Y
z`|m5w`riUV5CM`w-C7AML+gnOD5&20_Ln=O;^NF5<63s!77d_7DG4DhPwz|@K($II
zxLPm)hZI1r5fljE5k6GsWLc!&tjDNb#a{9j0(r-IdvGLZ0#(>AOCdNukX=CyBPAoF
zVrZD&p9SFTb3+6-_D%{+?qC28JvJd>0R}N|E-0p@q@@SJBnC49ay|Cx9Zt~qxl%>c
z7C7oUiYu2KqdFP)(FxYJ>vsg#wbj7FulwHx?Zr@`Vo+cL+}7N6_VFXK*Y(+|JLn>?
z+s^QR$_J)y@|yr*Mz<>k?1zSoEWPUtpb$ijpzpc&g{gMw8F;#XsgD5!9A`nm#n>KD
z0b=9gsu_4J{t2#hJ6O>w1Vw8s@!ilmnd)UaW0+wjUSLKxW*R-fLV}Vugq99*pKQ6L
ze6}8Lk;V4V(cSsh;RZJc9wLmb*=BuG&s{$%+4xKxdbJn2Q~*a{-6gb+jt;GQ=c4-<
zM7isOIWgvFXlR4LppRCCZpKqL0OY~HE5<gaYRju^XNwOH9TBxOa2~+(^R6a308TWB
z2@?$P$?X6cL2H83dLkDP;Nmd|QsdU;V+U4Ke2Pc)eRKh2wy&RF)+@JP{-;&^7@UW5
z_h0}{nV;^?1J*%pvdTIx9Gok6SBv)uUGHHw-dtZuL7}e!5`a9VC=>_H3COMlpUbui
z;AMglJs8ygJQ1jTmG$(J5xqIzmv=eP6qI&#>)8oQj=aH4pyoq;bXUS346Ey}Z&lB#
zd3rXSA8!?a%-CTeuy?`=(jsdrFEAMKJ3|yaokO`bz6zfkZ!n#+a{+StAjHpl9t8gh
zC58R`^nm`=tDmq7a&q$F3<*r&z?om^P$8+PcMr#5Ir2gmnPLmHNn$YS#GAkR)59-z
zj*L`-tN_R#_HRs8h`!h!9K#|%9Ezj`<O_!iBV^9;mKyMVj*W*0?^V~y+iES2scuh^
z>LH%J{2&wlXC^-;=TbyATL|W{O+Y#)g#@9w^YwX9jz)Jb4>Y(PcwXeN+Rij+<!&BU
z7`8n&rsO#ILO8r&N3*ltY`&pe4d32#VGTfmD3IUfE~}Y_aXgF28UNQKMf$MB^2vL}
ztu@@WTNuPV_M&@xdwy^{i0|Uvv5|T%pE$UKRvl|85*Smxv^JeT(nExUdG@>zL6~ak
z$Q?L#o`A>k3q%X{Kgd-IFm4F9%nY|9r=%1E_6IWTKvs-iKIq(hqUR5;)fw>U%yKdp
zS4W2UV&2Uk*Q}@Vb1IoWfRZSn@R6dyeANPWqd!lwa&qXYsKn)kt}1@2CjUv4Tkpsl
zxS3(8dj^4cYJ9qNCxnxNwpxMpWL0=(mUK)$aH+cGXl)88SY;)^@Nt070Qvs_)>lEr
zKM-e{yw0ZwxJ8?WJ@MHOGBQDczc3Irrzxt@IA5_e_sFi;Kw00v9!Ry!daQ@yF|;+Z
zi=MFj?WQsst%$ce5kdLoQr!6%h*T<|C7hm~jxZl%6O)ayPNsv()#c^0kyT$mPV2}X
zI8&03-(im4z>x@rM6TnquRSyKvHOvtbI1~-5ae}^A%Zwnd3HA&p#OlVgh6Yj0Dxo$
z|JJmFULpPPNC0v?b{f@3^*oAxs@|{e4Ouel>rJ4uVxDSL|0bdE84#qmT7UqU%f1}A
zrDiDe`pBv49J{T}7sQo0{^M`JBZ?9A45*IFw2BY`?S1tmCG^pwN4F(3f$eT?h#^dU
zoJqY6&y_!Cn?mU3#0yrv1cyDKgh`a0Hb*zMTBvrGC^YXBK=r<uS%&ljLX_@~9}ej@
z8BFAHI0T*f4vWfLa;)j*+QVtB?~NKg^Cz~cQGKa#O-gcsfhKyM(L%SVU+7Q^eBAu*
z>3Y<XmJRwtY}y6#i~=wlcXiDFSaT~d`HK+nD@aL&J8v=YxU|w3St2XsY)#jzD`#Iz
zGhbcsK$QwOw@^l^L6fUMt)p(WSVh<SY#9lTS9u%KA*zGrF+;V1hqHKDHLmM_6qTB>
zGYXIuB&=r|G<~)w^^4qRi{K=^k!==H0{oost~>;6jCZb~c`ib!3C;RB{-g`4OFzFN
zvIFG$`?nv0npYQl%jz%GK>_wB_jPq%>zogLo>6yjv6-hDGlyBF6kGCUG5kKJO1<-z
z#ir&%ekvFUd~>31Zb}AzVAgyoXzA&t`JA_k`t26EUWGj5sDBqUb2_Wdam|2=i&!*>
z+~FVA^-m`#lR1-&<EO{ZeQyv2O77`zEe}0cE9I@}8_OaAOuwHu2H;@3^arHT6ZdTP
z1aYt>&m^*Pb6buc%~cqNWgRCQ=QWzXvb%L{1RJpds2cH@59hTt{*lLx73M}q7yIs6
z_{`84J-uc1zsL|fItEtY*D~gSrXt|6&m0&Spnw9RE_|rUS{uQG>ee~tDK&#WVi7K*
zR;FpP+aH&nWqWfVo^ZdpIspxU94vD2T!~N;Buh-GG=qWGP|2HEhNMwl3y19Y<!7T3
z;_lFzQn!eqk}$`1T@Gb>h>Fs-AEVaI%<~QRZllMcP?kC()E9S70m4i!fKPSXb@0Qv
zPnZ_Fb3dD^-h8^BXE|E@@9$p)d;2o|E5M_z>EE2k#B*9}ty~_i4<ZI}cCcEIn=AR8
zS`PH6V*$sB`{Ouuctys1MQ8n>H&dXVB6PI*!skrWBJ$<cvU8r7KLITkgsu%fT$<6m
zaiB?l1h29kTHk+!O62q(00B62v*XQi#H>(ss97wv4CN^gml{a`Pyt25V>&FD1^G}9
zWKX#`TISl?+W$kkRc-4y6%UsxOGm5HqkZbxm=;9HcvX6vTwL8x_>`l1A*k}DN2oof
zjjGvuuQqC@NDAVy^Dhk~)dUqA#ZNtIR;GR()(j%u&5&_tG=eL}slvo(6|J6jPSE$d
zK3xn2*#^-H*7MvGn>_<mA<gU8iWvT_E=)jGahI)!ZK|&fwAw)sUYwk*vQGH*>(}Ku
z5rl4Mm^rJ_8-xEN&oar6$BE0;<K%{w8)ry#g3&qiy(9`VusE%~-zh^Ct;SNdmS5hN
zu4Ttj`d~G=TPJdQ?HisEzq;#r4T8wBS5&J=QZP&x)2X_C9YV~PZ9bU2{sm93Z+p7F
zv*N!<T?ZysR;6$i@Y*{5_U#+1wabiK&VP0Jqppb{zR$-+GKN-<M~|-AXtfUhG4ht#
zW>_7w95p$m9r{<NX%t53OPl$G9JWQzh8neI$#o;Ea#$eWO%P}htS`stkJNlv{AaCS
zW0zLku>Vi$<Ye~dS{-z>01?StEkKdESuN`GL_L2}Zvuc>&#eS5+Yu1Z<n;I<bCK~C
z@Eo;if7*jdvG7dkZCB}#Z#5Q2wZ_fttAVqjlN3+w{tL+cHz=eosiR8sK`MY~b3IPB
zm#es}#uOf45Yadkr=|+y28zwk&kx(&WHetcGv77pdj(E(vIHMMZjQ{U#}`1Uu3EaK
zz8=_pJv~^~$<gGSVNkGtlMq@Ld<$RY0P3l-na&3Si9FB-3C}Fnt(OYds<tTt9BTov
zmZ%N5KdH9@;7kGhrGSF=1fa_h<W9DHN-=P08glZCE_Zwc2+NU6dS9xMKaJ7CO&l*3
z{BezI8=-vixjrDJq@>K#EN{UfdG;(FEG*rtSFSSz&j=RuTjm+M`tiyo=zn9qiYXUP
zxitNDK6^6$NLfXp`VMqTNlNzpi=vN;h{yo7RL8m_9k5NHGx0uRvl?RofV#`2!FrNY
z;AG0d829xSfT+<7n(1H}X}P%6*-X0nfUm1Vm2`A;DkqTURDiO}@uItWy`Z;gDEewW
z06n-r)j2BVs|%eNNlw#bvs{dh_xJ28)4I1JA|OX-dS358tWCUEcWIkq(TT?vwRhy1
z<-;Lg#7to9Xr3!}3k`=1jf`*z2?<NC3)Kn=N6KJmiQFI3O?s1yKw*&YeRWR2=Umvs
z3orDoL}}@ma=enYsI4v?C_a@`lkRTl8Cw2ezECH@Vbf(D&*S49%->2gIjA>UeAF4`
zdQ52nlEkp(2z0VZH+R^<Ms4sohK|3gw4X9P`nE1sKHqJ{M6_uTkN!)HPrX=Y-K818
z4Ss`ZGyn?#P{_*8j`omunnX$^1NG^^NM!>1#X!z`#ctW0hoxW5EMguEdx-@Ao)M>y
zmPOoh+^J-Tqa=Lk+qcm*{N-kI_{aG{r=Pn4`XV87YHWKiL3f6uE0ISTG-E*Fqzrm{
ziE^g3gb>vFiVF5C+WB+>bS`hB@-fE`iwB^{n;`H}OT}5)9&M6Md7*n>CgiSgCGpsN
zeFZv9AjNzF#kX)U9TSsUY@4EGFIxH6{F<7Ykz_oYu%r0Vjf2E{64=iOGz)q~k8auY
z*d9=<4TIhzp!FcJqAHCm0hR-DJQrNUm#qzY#@@}M->q+S4tn1T-?c<)m%^nh3|j9a
z4j-g<Bpwvd7tEwOm?*ef2t1rRzxlgN>vQ!z)ueQ-0iHb2&a78Zs4FxwW)u(@vH=%!
z=^(+)_+9>x`y={YYnA_a%xGi0JSW-b+RbTwAnTVpL6up5dNe%>!~ws`qEKz?S|&Vl
zv1>TfmUKAO-={xONXXXC*e_;i&tB8oxya49Fd&2kV-5Qmti6c*f|IUHk1grLeDH=B
z{B4KQ0X+R>$$v-%&}j0<ARYi!oV>fPn5{aJZ{0iFYj5KvlYBk1hKW&2dZ<?Ujka5Z
zjgz%&v3-)8R^<iTk?oB90rOete!#(yO~05{_+Iw)^uZ&suc9w5A0fk_^-UH7#jvTK
zx3#+M$P?A=TS^)#C?A7g-Z65kO`rxj8r6sKZP$qO_xG3Wc2oHdDg=>hje6Uahv1S9
zA7jBQ1Z-oJW{3;Dp^Zn)oh#->$GDZjmKh(2ZPaqR(-gG5bbs{8l5U;p=)rO><#hFG
z1-oO|gzBAzf>Lz!{4)b2%*Kf3%XofS!d{}@yTivqH<#sVFg@M*A#4aYvFJE?Wy#JH
z-J_W!g6eZYYr*hh7VTWU^@EQLWx1BE4w24V?M-k@V!lH8<KmH&XUs~G0{NG6)8QV2
zuWJMLE!UM*f^+=TExNU3=1A{u0AxpMii%iB4}5uZT-}PNW^u2*FO7;DiUV>ckIW~O
zN0VL)OS8>HQ!uV`X_tn?J}<tYw2WZV78KoD=p6`;)bwK}dWig=cHvS-gcE?oz)Z)0
zkl<x%0}qMYG42JYZIw1}oBdl9Exql@osqSv=My<M>>|^%YdHxMdN2+Yb(*^LenOm*
z&l>L^v5ivkDL<bVy6cN+mKuIR==MOXAzvM)2S9fCu=~P}u}SwrbtGh?O2;($dVhR6
z;Nc^6zXAm=@5uZUm=0B-6tb^Rwm??<sO(^V!A#$mN2{vq_^-S&-jfm${w*u$4tuq(
zoKYZLa@Wm?h>p$%Ocpd8DE=yeLwuq~h8Rs9b|lf+;aDu=NOQOxyk(T~5kaeB8vMpN
zYA$~*D9~gRPqDtgEb?!v+W6GP0(++ATR2kdZRi;Ut;^tkN{H|LOqWhtRK1|m4wb+V
z+a9u0p-WY?Nki69?~dwh=FItd%Sn@Epa2sJ{k<-~TVM>4F?&~V{g1T1(}f6+pkV6t
zq*wc=2-v{fQUQ=Luu$Kw*<*k<%5|7xKzCwwQYZ@4Y{PGjAqoxw0&R4T#B6m>QswR%
zK_zYnCi=OG?xc{7vM=i}%kFF2qtkm4mdkE7p_A3Zv<jnlp^~O?`#un0aO?699of3i
zuf|m#H!$+x6MR^4syyaT<xDzl0tynyr|v|z4clXS2Xsj6^p{7|6&$7sO}x9h!9wsw
z0sXbzQY$B$zS2J__g?tj#xiUK(JaKlp_(BP8r`OJ&9X5yr88~c$q3@Um37nu2nX43
zf^;`8y`d6SBaYnRs5+Lk1R_p}mUE#Mx&h|#gh*XSQF>?ZTw_{?*z}LQfw!sot*dJd
z=;?d4JJ(-jj^<=$4mm<03iQD13x-kD&3=xIRvNC-)gqt%8`VAYIO<4dKAdU`HScc0
z2`H1eS7_T}byuPPubF5y%d>Lx=g*!E0V-~S0TW`OQb;adUD@2R#_F5;A6owRH>{?s
zwm>{f&?&#K1}d44=z&3?C>eA+nRW?g0%RI0K4OW$>wmYMdAG*@7d_Aa&mM&K8YmpV
z6w+2}Pei{mCXUvs4%^^jsUw55m<Qd?%JpRQD}l?@3e#%rXJj~d(kQU7Kf%mA_aUj#
zdp@M#+|!&Ag}NOZT6*n$cWF=^Yh8z6Kc}B&|1);}?%O6&%*=&`)?o;{r*CQHSfSDZ
z)j+M!F@g3&U#&X<{*C8eTx1Njn}BgI9-cIsenfGDR6t1BTM^2@wpaqi0z@Go@cPKN
zF2yGbgVaqTSWC#tOrWhzj#2+A4A*Nnb;c9U4+yT`eeR&Me`^V+rmH>SVT8-H&Qc=9
zwBwkM!qU^Ta%pIJ&rcXPPFMJt%$Y6v6yneqaY#vFK)9?~ROnCoM6FOGt=3_6qdldH
zhKZ?X&HkV3R^=H1ziY`zp=RqW^&Gn}qEQ67VYgJ<`}?FowXQ~@<6&q6|N5FDExq(a
zBRq0?`$yy^{7?Sam_}P$8?B(=Om&oCaV{ehlbmzCz&lbr>XRX5`SmtDeRkcNS5R{C
zsnk(D?NmU=<(l>;i^rg(e!YbWEkXh}6j!Tl;k2|_682+T13wG~#+nmdLnj%RNf{sU
zOZk&Zgzl^^(;bh!*0rdn<>S)^q6J?2rEfsrRtZ}5wbM4uQb7R$`6@4z^t3@4$skjd
zR{@Y1f?^hrQA`6`FNNmIeKe#8b_BP9%h!jDEQ>lNno2$#gM=n7V8Ky}HU^qu%XCAM
zQ1l*YAa?9=z3jJ#qYK;A*QuLF?oL25q`#}kdr7lXb9fD?e^3N1jkY@g$#n@ir=Xg>
zb#%H&CQSl1j#8Oe%*4%>r{gzln+Um9tKVINLYvM&n-l{$(cF6o*b3RzPhiVZ$CK2q
zmh#7xww{oD6l}Q!bo*i;?FWQ}s2!|yUp3tToki4Rk!#lmj}sd}*LU<99(jymne&kZ
zC6W<+#|R?}-I<hYbwq_A-%6?=s?QB$dTdM`adH5GoqgA}KKHUsX48LddlDT|7h<~|
z4s**omgn}lgf|(5H`~&pS4RnHQLtuqonfL0Dg=Io8fTAv{_^GQ>dZnJG{AfK4`e{`
z*W%K%?zr*u0qGFvITn8Yj4H7ImmE>G!o2nt+Uh`8HZn1>5D3pyLBn;oQed|Q1))sQ
zu)TN3AVL)U0P9YBe7eF6%ls(x`K22Z?JVVT)ML^8>8y`fF3=0PoTYqj;#xfsd2T#h
zT<3vGak;8vNifI3MSKfj_JN>j+M!nf09B?zPN4Fkq@Y*>Iy!_PWM`%k2L~q$Y&oz5
z{dD)EVfCo6u(VGPF!SY81SiUj1A(p%L7jv0>X{9M=9UJ7<UYcUl2QI%0ezHqJq?dq
zI*ddYmkP$Ng<Z4YEaOJ=O`8B_a-!x#hJI-IxRt8Ak*ufZCGABrzgrm)P>?~5i-jE5
zejrf8uU~J+x17O0fxX|)5gRLpntgn{F1D%(fC3&=GQ(isMy2hn9@s3CuixC1?70uU
zx|qE=hO;5(ULev01X5lq`M6Rl4I}URd$$kVtC?Vo-kY&ta2+8*n6(Tg;&qe+dXIf~
zdZB76Uf%P}t((hLp`%`(Yvoz5lO(3bL#5fX6rXENAZTD>tIP&WD<}>kI+vBIZKl})
zy9(;bTQCd64u}W<N`>GTuYV1pB_Olhl6Vk>c5WgJ4p81Anrs((q3sz!W+v5Kx&nta
z7ppXLCKPpjFfqxVsJ7K#0o5GT458)(t+kDf0+<@0pMYYE-s^PUA8;A*9G20D5Y{Y&
z=UW$g8&Q2jf%08PClR3}2gI?mUP@%7O4ce6Z#^IBJGEZ=YYX=a0zIY-@!<VRm`3T_
zn-iZK>X^=~i=AeDKvxfy7__o`ojZQ{8ZCtW7k;z>s^UHgQZE$~cUdbdh74g}-aE=q
zxX7|U{R=A_TW4ZtcX!mV1*6MEAmHKKNqq`|dBDd_Q&d3}78w<l1<LF7d4JLoz)tIJ
zI{*p;bmX%Dri$XRE!6<+DedJ-35_{ymRkf1fm9ZY_+x<5JPCmc41NObM!KNadE|w7
zjJlqi*qsjWY=7B(F_;T95*@oaM%U8(N%C3dO0<gV&Nm92j>k>axu;DKow539?_p9b
zEX6FY5mDZvvQGLyi<Kgdmi4>zBU~g6q_)l8SBF4ThmdLUdY+bSHyp?TQQgjGOEgeJ
z17X5!JXwscbDW;fl~XW@fEMUTje=cHbe<uASOpCkr*>i&71!WVxT?3e_ZQ3HM60k7
z&^ZF^QWvauyJP}~kf4BDqHsaP2nYU6(bKGO;jIyl$E8kU#u9fAmcylGBH}0G(8%Fm
zL-7L*dZ7%Lal*HO3-V+iG&eSqJ+>7ACm$Xjez;lQyYUf?iF>{dK^Ya-Z|@V&I*sZg
zf`TI$hS}dAuiSw)H(%8sj7>!)643BNz$Kb@gq$~HnGd-&S;7HvUfAs)>Ge`tB9Ui8
z-%eQDf%Vcode~4E)INJ5W2~>BI_s&wu?lMXYLw<B5u?S1e~X&dI`Kxm^j`O(>k>#k
z@RwIMdoPgD>F2jGGYdpv2n}79G@h}saTE^-DcN{>u5Xwdc7H*4ynvJbBbwZDmQ4}V
z9n6u(bWKvl0!RD_CM*DQ-<*b#+PQB!DMUxa_54iBxU01FG!i^LI>fa96rKCAh(D~*
z<j|;(_Ye{D%tQn9C=p=q_&Y9xT&#BDgJU(nUZ{Sh_&7Pv>Niy@v!I30&}Vru{lZ6T
zT)fMU2|#$=d}C!1-@ZK5Bufh4?Qxo@-Xqs#-*2{X?okzwx%cU=;htM|+qQYDfL8i7
z7u1^@U;bQp!PNDL8GF>7`@@vx|JoWzhJ<kYpovONgkUezOvYs-rW0kYS}BK3Y7X>c
zKvNRW-O|$1sC6xu-dl!>N=U@iS4lZ(iV)S=FN~glILpv#zI<Ke?(N%Cvn#I4;u%kb
zb$t^@R&$$=A-TA)QEhUehWj5om&3bF*y@4^48p|vk7>jvGx8mM(%`Z7iM6~FjT&y#
zI^Y(WhkJl0M3808$w!lLfBK6*@zWBEksd61X;^#qqQ5%~i%wJ0IDV(~sNaZm^I;AF
z-idaxK)uqol}_9t@#|e2R2c_Ip-|uZOpKV<k)<xuzDT!T4Rm2axuN_R^O3TvWb7ui
zzf7M9KH3QI2GBX?4W$=mg_f6<RRA6e?0RfYHpL8#CMVNN3#0V&Ctq+q{cu@B@$kEP
zjGuCa!*qhwjfGx^SqPf=+DCYbDSNy5t^5Zf9EA{lBD%Jq8!lTd-U|9SIUtPYx$Mk{
z$0)6iZ8>w8^$7!U;O=d7Y-nhR`1&fL%G7dt#EJ(kG?zw5yXT`7&ax@|?+p5WO(6YM
zVGvAz5V6K})Xsk`$}e>3KLYRptkRdR9&qFpplnJS+>_DuX*C0F{h_{fI<QgAeW&q6
zrCG1WTjwTl#%=YAmrsmRL$gS&?av=NPEIv|XBbZ6VQiWBDba&-NB>lHXiUAU0v?ww
zGjK1U2_=4!SnBFk8c9xZvV&XTtdRVm5aY6N7@%d!<@?VB$F&*ui(m?5mJ=`z<rDxy
zAw`4&|AigVfyJUc(=`9@f}9yIUD=3UmO|ViNT3*Kczod|!^WCDJUP&V9ot+=3N;*j
z!?3z~8$NVeIJ3b&XTTv?Q2n|xedIrOQ<eE!0lnroect}y1B-v%qtTv~lV=}tdJfzx
zS7T0&TKPiP-u+Pz&!pNq-$<k<Ra#`bd2jb2*ZPtW^sph2hf+d5yA@5^=Uh-Fz_x8n
zAw>+Y1qD7Sek^`lMg!!YwR8X*^(!hE&bvXei1lphUv7Da2u-Qr9*yTq5%LwTdxp!)
z%XuLCBbv^hXqa`^DUE!gp#}LjLf5=p5m}7WW`o39QrfthO~P2HpC1P>Kp&gENhPyv
zJ~g^{{EsZqcyLEO?d5_R<F_((KNC91H|P&7`f;<$b}SQQ%<8z{iL<0BeD7J47$U5x
zIxfy!yU5v2W*m0n;y7zh9E4UK2ncJzwgZIf82_~)YEH5{oj;NNU+S&;Iz3TQ_dj!3
z&$$;aCCaAJJFp(x0U}+p`2-An|MzowJFr>f>37i@AmcmR0-M#K)Jdo<U}xf=Z&>6t
z;F@^_1)vFV%_$}(26(`7?U6umb}w4lG!hmTR#@`T(9myy3mll5P7Uv4lE}PygOrox
zOc2U%=yD0PG{RzH_kiA+l9@Raw9FmO*K-_z#W{DngOh<4Y3?bwhBqT_AXaEQKA-a@
zbtjcv;$N?m=?5+KJ58OZdyBL*G^g*cKpPlrVB<2gu-Mo?|IHHnx%q0#Hih5kpqq=9
zj&9@lgn^8V^(k%3D>5>&D=h|GqlpT0{(~;IDR3#PshTh_!h=Nnf3(;*IEtE@9U*+%
zW9i*sQw1cqpa8yq4u)2(;nFXL2ZkZ!CIGgZZM$XQ$7^W(jazK5YBpddQ2&v##biak
zu+Qya2(79w4NgyGnbaNo6Wh_wW;N6_zAcXS>8nhSGhP+erjBr?fn-&E|67)<)T?0k
z@mPc1as2UZF=-=z)A|SI*tp@<H-3n}dyKdG{<9it>gw%4V)F~>0Z~h+&!tfVV7Pk)
zbU>?e3Q!9WQ8GsP?+qu6dM-|0W#zqs7ZIzlrgJ*5iDKh4VO_i)o#c1AZnp9Am>0Au
zu*B?yzI-}Z&|qR_^nE>=*V`rEpLQ*C$6Q2n*|?J;6D#*IU+pg|mrVyJra}CHF&tWg
zU4mo}k>$AOQFUC^*E1vJ766zq1oBp(L{Qe$98d2C^@m;0YqP#oCYU_XvNy;lGi#Q8
z2I@t}QEeP7Ec`x|_bQ^IqUV)BSPxk*r`ta`h~sx-1v^csWMySj7<|C8k6FSR?RuQt
z8WH!zxW^5S&(tU6v8Mo#YYAYZ?7Y11+1XSG>2V16!b7m@_ysWLi_LOOOu^HaGae%_
z{oY?+(Bk-9n1o|SsSCBML%=Z@M`tNFm()P-@}}Aw17QG3HuMM+4?VEP$y&P?92`95
zb8~fhebP(~%rGN_Hy}b@@GJsfZUYXlT@2wn9Zb3Y4TQSf@WS#~9)}f#*EE5lOgm2o
zgTit591ing&FBoLl;+;9Sv;LNX=(>eqw9_)!J|BRggH&xHg|*b-51H90joa1zO`zS
zm@S9|lSauqq!=ChM_}^?XY*+rv+p5NNP<n>cUv#GcA`TM4}(q(OM)&gI8B*d6l|an
z&BROexjyTPef<gOV@iQ$reab>P0jx5XcW-_v8kE=s<)j69&R@KcIhOUX=rHP0pYjn
zPVHKH-~g^b<3@e~P;a4(>Nubq{R!%L%Pb~}x3L2@6y0e7M9-OSXI|93<8?bxtD*#P
z8(S}b>nBXXF~e!ld(o^g!vYXx6(Fp?N|<0s|MT(R9YVl*BVq;U<%6(*Viq?10Ync4
z1%)5W$pK(Z1#YeR+hICl>JEDNLV?IXRD#3})mODZZLf>1cejn%JIpAAhpJNm8$5FY
zdmuRN+aY&%`hb$GEIPQaDagJ<qoZt|r*=S9^c$e}F%XQ(rUgKt5QO3rf%PS8v|9Qb
zHf7g}4h}>q@xF5J5xVib#|37uPWbkSfyKz5rA5G$fIjsP-%dj-DBR9KkN5?T{rjqV
zN;0wmQ`%1=;CKqO{r&yZ@m_Cnz!rn9-wDnXh)C;X#<wvtRQdj`d$8aB2?qL2G9~AB
z+7X`*-R_?D|M9Q!I-DPXvA!2J`+NHVeEz@X<sN||h{Aw6gIa}5iiC-*kMi|B%yN9S
zvG-tR^~jlTr?Im{_eYlV8|F#}O^Z?Zte5HJ@_BO{Q|!Z`>b+c~Yt+Pc$c-16O%Qv3
z1VfUMJqbgp=1#4gSSAfiHLMwMJ(`|x%>=a>swSD8I5OZK%v@YPnViN8Y`feTF35R5
zww<lPutHBykJ(c)&1F~`3(L>vrHyb_RaK?_WB8ghu#KrFQ`gvd`Hew+0-L$IvN9Ww
zm5UG6R>MhH$+SMJ`Gco{ZFEMZW7<h()vUOI!!GH%?3detY*r_OyL^{0nF(+uT^w)D
zc9W4oxdhS4=8SMsLSF%oJbf}0EpbI7L;@q>MAf)Q6klP5zWcO<-(2%4(}l&3ldNc_
z@>gVf-_r5zZ<0P>%_X-_1$SMeFLq7mxS%jnWD>JQCrb-ohx$_e^-KJk4FAwqWQ{wV
zn>}W~_F7Qqq+9%2B<I#Mk6h9ZZLvC<evb9=`s*5@qsZM`m_PTN%Nbvf|J#CvSlFfK
zZeyg)R{wIAcfLG~)*7FHR~<*}xlTXHI#v}uZ`x~#gTi`|NTA~;#4LnpnIEr;Y45R0
z9rZc^nZq<WNvOP~QACS}P(xkad5z35ypa?c+hhXc#d4!5&<h$54-d=Y6WKl!rg+9v
zR##WoDgsi=AOu*wpti2=^!zJ~Jy%cpRid8Ss{)V;S-EGfAw9JMtY=mWd#A<ogXZ{E
z(hWJ^<0`c4ovjya;kk02CtrGaX1_bPHlDnpo(M+y+uJJyF6=HMF$pB+Nxyb>9l%YM
z!If;ll~BM{2b1M_R;xh0ZJ{Y~5|`#W;DBj`x{V$)i^n8H@&aAuM1fTFKEE!}H_0*)
z1gyaC3%9gjhqzRhP4jN{tW{)=UP<wjq>zNRhniV?Jw-)gVJ8Xo5sbLvUA>QV7FUJ#
z@g1tLh)65e%!RARBsH-ugYU`pyPc~)PcV4Kc_jErW>cc^1-6K7-CY|58c)Q)d}z5g
z;0sfwbl2>zhTwZ=izGQfjM&Uq6Q$qlhpzp0(TCK?RS<3@cGl=VMa^hD`MtQ;T0-`d
zm;<ZD!^ii;d-chAt;5)MbHfR__V|iXY7R0{k!JY*Ih$}r%*9JGva(=W4ODM|UBMDx
z>vpAW-z;f$lmb6egRpZIw$gtS+I~0hTxjvWb}9uw3_ZenisCJILG4lf<w~ZQ!ka{G
z*h6igjQ<@Q3zugSuV@z0sV*T*yZliNM$J=K9{xXzE#Br;TU{3@uJ>ymoEn~bvK?!B
z=b^Q`KCHPXO|L6+Kpazis4U96{U$zeKT8i=@)-`kb!AdOP}N&8@p@g0u>N11RbE`?
z^|mLTqoVhjhUcEH$2R&_mqzSt`RSItdIp1h`el9mu-ffKeFhtTeT|QhXR;m3F6Nxt
zJHEM>b2JQp!}4ETWSnxAJm<u?%Cj5Svp(bs(9H~$)>6Yg<~hUY!nW!%GvxPh!MLa)
zD0t!hZY)NuMaUfg-TLN*H21wzUOQy97P#kJbp?Z5Y{zS|mRP&17{8A;8#&(`yK*9u
zO+ELR$ig$!C<Ri&i{!pNK?Q-fqSwHv#V6n%bG$EOOLa;8c?xE=Ci3OePJ6T8Wt85?
z3~k8d-@T?WbqH-nKg$w3KUPRs8FMkrixfiMD2d$iDDC<(n>j8xU%R_1IN#e)l7czN
z+B!eCd%9D7gyc9t*uu{ck@@K2$#(mB%3lV=K#y@!9}_NWAvFZtqkhIJ@Q1c#)_0!W
zM{T-a9?M+cUhAw~6aMAfAtjqH9}q*huVe6Swt!Y4p23$%wtqzH1|1@fWxBSqnfl7U
zBAH%%UqU^Uqel(#Hu~+ZR0i;BJs|b;JWpanve=pA#<MZdK2s1rCPKSp?H0r{ibMIN
zy?)@%P)`w7^cr6}uz_HVBX8PjldO^aiwK*OY5dj9Cmbw>Y|9<ydCe;HCykFbaRqm}
zUUMSjz4J@y*%lGyrLGnKWFn3gcI=7cipA{&pqz-MrR8{;ai@mWw@pHg#y}mGSL<i-
zg-37SLlnXk!;WRqeVPXUaDIR5Vw<gbVsq-ijyne_Sp2aeriD>7_UiJCCK@S>K~c<$
zk0F5~Ngw3~T_@<C#<eecoH?@d12sc7%)=p#VXgrNX-~lnvz~>u&5KVfy_}lLN*6tz
zUvX$BosT@8hJ|QDHQxPL5zhMdgeU@_!xemTiIAndoMhe?uE8TWO(a3nGEgo{?N^*#
zhaWco$kiPb&n);}MSwzk9Rq9zQC85__F&}fRDn!20_udGj*iaaIUK&RVY54vf}(;v
zDw@&%U~()=vN`JsMi@|3Sg}`>%BHJ{7{xn%r`txaAXbsNLP;7sZn3-R2<6|zG|~RD
zK+?*}^%YsdD30m_L*ECQ67p!OGaXC5wRFd3{bxI+$;Hh=W1Mq8^4Z!YuafICQpg5g
z%To#nC0F$-dw<4x;eDUGh?<X*EDr4W(tW=*QK_P??y}@=#Lh`{nglkRp7jC6*}<4m
z=yZT?p-)n@*x+-UtgTrrj0<fq<FC3qHhhEj8vpiXT)?wk(7%&ZRP4tXtDXJ~1SQwJ
z=2v7pUM$|1r>_B!O}Wp+OiRm?9e9Pi9ufKT)j$(wYl=~+P{kv+G2Jf_W<A@n!Uug{
zKR33fM74O#9}xS-M{XTM-IBORiZv7mWEx`~FBg4DKf^jpHI>zjS~W&o3}=a73v`$1
zwx$%6wn!Uln$GA;I3BTW7JGe@Z0CP6g)K-LfFt2e(e3=aB>mSTyd%42(epCe7Vy6|
zTE#e-!9-$uuzg?9EN9yF-S&6t<Qio_0FtN)ZR!<{s9GY=G0c?#nfcC>xkL|;mx(ub
zeDb$Sn_<r(1Ox<R=;-M0MrkJvj=HIk)hR3SKXprvCcm8%wW6xEveRJUG(d@#oixbG
z%KGFN0hYKbP(0&E_zJ*Q2Xj5W89sn4mwo_k+MAXsu!j^QKs*CoSOion#yzz=2IDzo
zC;`YsWLw%_?y-j4*OS!phB%wf!8A$?*%UtBLk~|1k>%0~l{KovCnN}OfV~-$p_8X(
zEch))wo+#OsuS)T0VNfcG;1m2X!<mKBbg~G_(}G|hF|@J*{o^~gsZO%qaA9HQuR5s
zxS=^@ZNqd-_X3!=8ycED5oyykBPY%wX+#zNy-UIMBe>ZgoG?csJ6b#PzFnZe!!yLY
zoJA^r(DeNt<^T`j>FcLA2Prp+e?=Rgps{+Ab_RUQspx1&4o}J9DUV{w-RCHToHA%7
z-#k4wcxw;zJdMZ&s>dgG;}~OSVs%|1DjB5-N?XQW&eevS<-By};~Z8OwUr$nLj63|
z2EgSLBBHI4z}D=;nG1nL2V07-1rY94z3=Ud7hp#gy%Ny$Fx8(v2U+`p>yCu!0Ckv^
zU5_;}*sq^5F)4=xyolr7n}GV#IQHcBGM??0^!@Mik8%vfsBZ$ge}(yHsczpVil~4o
z`}vo=`@=T&G`e0==+7Z4LyNtZ>m?*%e9=NnE@v$M;+NOWbNcs>yZu5$O38EtjMX<O
zHpDkG7o^9;I(wj#jz?EcqAo-kT<5f8D>9p8W)#OBhtR^A9{rLuC;x0KeW|Z_PsmVy
zbCn8<lQ8=Kq(a8!GFYV)f|_&To#IboBr!hm(IdKUATz!SZo$gLRKb2XO6=3J5lz|S
zCT_?jIzjkn!SCj$i5yC&;FhEvO5qDPTU0L2!OMD@X&G8(gEtvL47zoPPF5z>@=MFh
zNf?)`1i>~y_cZ=}rBD~rZ~}(_F5bUSF1z^T`hI-iK;RO?&ebIQsJK-WuruQF7ijON
zLX2%1Ld2GRqAI$d&6SVoUTL;yZEi2VPvteP??tM9$7a{yYs|}FI{7h$+2AIu*+0pb
z-GKVRAXNW1o2u9;8&6w3QoX{dRn1MZJyW#vIj-}`(rbpAFr60_1v@bEy2TfbApN|C
z`|6g|SAK^qEbWF_q3qt9kc+{?&!s}1&DwIfdT7Vy9P%732EFaNCyRb6?tfzKOXH#Z
zqqk>_eV2Wgin3?VE<{>Xim{J1Yqo6JRb&q#*~*d#&5(T?2}uaqw`rBKW*J%kNA>$W
z&;P~q?wJ?A`ut?>`~KeF?>X1G&N<h)6lNc}6Z369bOJTwIO~sOd#`x=8_n)6wxxK1
zOz{_XHSJCf*8BN}G{dLSQwHvObVrW`GClt~<RGsT2uX@-ptSb>cW>PSbc*zmQAXgC
zuBewjdD5yQ$OIAEha{6<Azzz9OLlv-D?}+He0awWwoF1G_4CoGw6AC-pTHcU-O>Kc
zL9syghK7F5rKCN2-5<Q~?`U1IOs`t4mzy}Ix^JmG;k&V0xFL7gf?xSoT=le-aWe%|
z*Lm8^=@pH68%H*IULv`e)}p0YnE8l)U{d&^_PFGY91j(&R8WX)83XItv^RQ9LP4rD
zG*@5xHRdh*M<Y5PV=e0AqHek|e;Vz1vQ#r@#eXB6jvH%YzSu!hZShuCR9t*F*?ZB5
z@DAXx@S*O{E1M0PjPnmFeoG1Lfwguk;04Z?c(c$-u<i_!9-@UV9Z%OcGgg&T5Q{n*
zJh~q4at^s%TriUB?AFF|_oY~ZT22B3D&zjzj*lt1#CWn&(~rckHXIlGTm(=2C}Vcf
zVCheK{`V`^(iT^YjO1qm_fKJnaB()xSOLnt&rLJeLDuRw)C;>(rU$<b+!r<-)@0^n
zl)X@T5OC`44`_Ev+)x!qF~{?|Qiw<CTpd{X86$VL&#FD9vc~V75~25jy=Uo}VLAMn
z*(_iN1SUtK(Beryl}3xQuE7Oc14A#{+I~(dZlw!@%j(ZgGP?|y)_FmJb}Oi#Sn!ax
zCT4jXU(at59D7Q8@Wf0>dtY-$$wC6hnCS3t3r0le(2w)dzHwf_JBRq#Jv=LJ6f4#W
zP+%u|_@}WfW{kTSNslGG+ea!$n%6E@L+M`V?k;ggeJS~z#Ba{sqk)I{68XPmaWh40
z%crHKRikIKxUX4SHhV%&uM>J4*1kc%lIuNfL4L*A=F66rdqgAO69X^)GJAppEh>6j
z^_IoXPA2zus7n23@2q8p3O@e&VQvV0c*EN%$L-iPo>59MeUm*|`S_qbHtoEo&IoQv
zo67#7n*Q7l=t2+ul7;d=xSKh4SOn+}*-P_{e09h0bq>3lHn}X_<%5T2nzM2at8{FT
zIK#XwhcO(exK;YtH*i4$VN@j6Li;=;rF_#}RX+a5$rGLDHhHDUHfM?ri;Q}Kw1LuE
zRn`snP*ZvrCMKr?yKwN??L(m*UMoG8P*<Osot+I78(lsr#eLjhaPb6zQY@_hfI)fY
zb~%bnxc70uaH^dTJ_*l?*Nwj0Tc_kYoUe*~B2#99dS}%f^t%nnKQYJ#{!$3rp7US1
z!o>8K&C%W}o<XLBy$IR;X*F}u0JM(p!qelICLD<@X_q~BAQRd24U}yXQAM^|#vy}e
z8qN^dFaw~f^}-$$XjBr__cpJ(>W>lYw;Z&!$Md=V-T8<EW(At2f|aqqJE1L0i`Uhd
zVP|jL<%I<*&Jv4h-R0yNG{{oK@nI%wJ?R$Q!~_Q{{t#fWie2yAzi(B2=$h(>wRn11
z<j-0NW}KqRyz6jInBQt{S0Z*T^bZMq@UJI1EgS<KYSvF{t3CO<`IwU4nNPQG%=}2R
zyr)O=AKhZ-ph;1kfjIAKKXqYhl%0v{-|i3`2#yI;M=l}Uk$oh$E5{Beg`fU&Y+)BK
zuJRH9+nuDW<*z;bT=fvAz+b)+2)FGl8Wz;b@_J%89lqV<XVM$Q|GBzFqP=6mEmn^7
zwOg{wrZ3j??7{%e|6yN_dbc3+@af3(mpg|DrwA^1<M7bOCWhL#?*4n}f*FByn<SSK
zSy`u4&G*DSSUfnE0|=yq#If&>-)Q`|z=GMI!@V$^%Wkfwqp{)aVKSAYl1a$3|J`?a
zF<E&l2@eqisu3E7^2*1g-#gwr;LN}r%*M6<a4}CHQ&1z-$9@zwK<0K?LO{p&Z$oHH
z*N84FN%3&dHqy@XmOn>0Aalw2UXdIY9BBK?H3I+Ttxif$YDsOiPrWJUp|@Foa`Kx3
zegS`?X;M82sr#316BNwYrd`$__*sJs(=DkwUVC^z6tL&PzBYAonsdc?*86BO?u-CI
zYN=mSuIl~m-3j2;8^GGoU!pQ+jz7bHUTN{fDDRZr$eH|pZe?#3rmj;&q!n_J&Z>_-
zJz6jA_&?@emdk>K<*1}6U|b=}2yP<mT4Sj29w^=&9v+q;5$R0iSN{g~gr6l^TLj7`
z<;B5cUq&qLPWRTe<@bM=u?S%nOpcX6?U9+sUSMzF?U7RilJ74gi!3ZHb$v$M(_GqJ
zT%GMp^mE_`Xqg{g2g-pwEb(W!!y+pxD!Q}0XHa}}3KI(I>iffsKWgrM$@0Np#qhg+
z-`AHFFV!>pyWLPR#J$emx_wbe$xun0D`<t>c6M=jLD5X^>e)MGghu=SgkJ7J%ToU(
zqDODjW%y*Gzhwm;1BpTfu)2DXP|5dR4D0CVa5;DGTq1>h;F3GE6@1I<CBLS9wCEjq
z8p0E!BRtMep)ydWdg$*tw<9P*L2?m^)i{(WJS%7~FE3w|#J@OLmnXn?UNJT%wvCUE
zPlT<SjtyPJ$9B>Dhg;q~l6y|i!Ijv_f536qK?0wjHCj8Z6pgdFPWZL4Wny6w9Hhdh
zo>#TrTX&Bba^>2!p@oG7(RN1x^+Dg>x@<@k8XNCpOG-pi5IrG?fPjDzqifgP<{ghL
z?Y8!w%JOD6zIH7;t99j28>+Xoq-4vLeBtTcPSqzRCC<*|3mtdW8T!7+xH&p<{K!I$
zDx~-V;Ebjs@m+lFr~cuT@rP*Nl(5E&JDFo`fd+u?_Yb~&!BKf>_wdK--T^@6`H**J
za;mA;l$8AVZdCmns>U~!d8e)`v!zI{iEA;S$}_hF1rP5Q0STXk%zycE?c(fY=-cjY
z_lsh%p~fao5u+Ssm5|)Jxf~IPAaIN+=>BwER8*N+vA(e}=%URSxIk60L(!$~uC6?P
z0DzvGOyI`+6aa%*NYzTBYdx&B%yxm%R6wTR+{=#9<8o(bXTN?3?|e?xLoS;?`4{S@
zxX{4D-_wCQjSl?02ASLb6VIQLe1-~{pUG)t2p=Dx+voK4H@thr&z0UglTWtdzcVF3
z>8n51_ZN2*_FR4T;>Eteq3BCMd2*m<jI3@8)XMQC)+lk`_FjB3z3ulV=ttMF-n!Q5
zZAqX}+3{+?v^5dqdELG=?U#EmzIpoeX$4$35m>l=gm)A1`l;EyY|RG4u-GQz%*DN4
zrT2Z3ch(JWAQ!tHPg(ZXDZi4hv3uXQqVeHW&AY(;g#>x8-md{m>9%*p`+5ofcREYR
z-by<~HPvK;?y8g74Wt@(VrL5@spbV;?trBi9=*X-+Y(mY1pn(uE8_HiqPK*b^$gs^
z7VXAJPI~xY{rhIaN(G-9wAjWu*Lt(0tZz;G7)Wwhxw<R=F;{jP_c}jY-R#|6g*?!n
zVnBeUrKP<JU#Kcqe6OCEGVFOjnuy1F@Scub($p8Z#YnopGQA`Ner`uw{tP2&?yu>6
zs%7Xr=~SGRfG{!Y*tp{CQ<7`3?2^+$4Mz@iSkw?|Kpz4PjZcHRptz2`P8*!kQZSI1
zZq(aV1E)qDHqPodR3WCfuL0~^vU^`H?MnOs9Y0EwBWN*S$IZ2}^V7C)DtbQ-GDtfa
zLwOYkebKXb1r`@B^nB^)IMYa!%Xqmuty}W^Ia1!|ld>!652gMaGz8`T8>$*AeEUSG
zuCwKQ8H@B^vf_hgl@_J(@`JtXfga?}*r-a`YIi*Wxia>z01@mP_yrdrLTf%U@z2~u
zcrR+tPEHkfLr42|BJ{U1K)=>02WUx1<PBcy*SamRnL*rpHM{|E%F@KhXyY{$MwMr&
zhwl0OIqLxJD3`tukBq2;Fv?~FSUSlE>FMeLplF*w89@;4t$R)t)q3xq-2`+)$zN|^
z5S%*U7|INeIZ3S#S=mGky<ma!`1mSvYip}1%ljCZ&P>6iH|Ow@=@J%O8fMCqV39nh
zw?F*@aG{>P!AH2__joY}5oq{ScmVuHl_sOgY}U4YK)`RwTc3KO!Hr2Dqyg<P%l(co
zu~n%DFvOIwe>ahys_33-qLG7X^qG8jl=nax+)2`6(LY=O2<>m04Q>za&)}cCJow*N
z|0H(6ue;EbHItr4nP&Q~Kb!uU@CIM-FLk3g8%$Qq;7{XE_N?!QA26W*IZUM`WjuSh
zJ#hYSG>ZNJpyZ$A$o$GFswCI#$yv3UDJVvK2PrA?C)}xZcDjTvSHIKRf1S%@mGZIy
zhKBK+=^i!K-RRcQ0Xh2PnJNLl7NU54IVgB<zMl>M8Gd%*<ph=gGiZGFg>o$~C}3Nn
zC`C0h(9VrXoTJ!b=6JMl@!q{Z03tb4h(r?c0@aCHDez%HiG~%Md{5WA9#6VQyL00J
zEjL0Iu5oYcXMUKwbrttTm%^#$(fzxJOdZ^4krJeP$hhw~%X7Mul#^5}b`3`aQQ21}
zV5N_;TaaZ@?I;CJJ$SiDjbZh@?CB**JhzqGssI@^%YzKu`j4RoWLx<9!pmrc@FK30
zZ#IETE+#TEG9|1E-xvNvc7j5KWi;2VuZ5;9*AJ7SpLnc{n{<bd<NOwP%3bthcP9DU
zQFfiz!=xXRbC*=4rD=}csk~KbD2Ph>`(&}raZ}khdT!$l0*fUF)XDILcsbBm3k<Qg
zKD@rZUW6(_)sPq`hsei8qZw_|(PC+k3p&(Y{wl)V+mW*XL8BJL^~OD4JpeC>y+t1#
z`?(JNP1nq1<Sw2He^c!9*;_M7Xd_O1?ebwklgEEv`j@;jrIwm5pT4V{QR+kbA%#i6
z>SjUp;_20vVjAp|!&a~N#D!={_U;V5)}u3GxvSodQ{LJ?Z|ZmJ(ZV2Slzo60gB8Dy
zcciXaHAz%HLbsYHYk~gWy~}ie!$Klz_9br54&x-buCtYU3V-&crOXq?2FIr3_P<BM
zyI%5^QjW?uI1=eMp3^@}TC2~`K+tmxwYuqOZ;4O+J}81Pzsb8S;X!F+CBV>--z@4u
z?IGsNZc!l9d|{iD^)B8XPlAu??tQ!)7lU{h-4pjqv$nMQ_AWKnHSUgI7|Ln?X#s`W
z55mcU>vUw*6kOOxB6MR5u;=ss9GEU0IrN3_Bs}xzscvclABG(xIseJvs3Ft*jDar8
zpcnN>Lht0DM)&zRY9-$%W~3_VYg<gF^hr(^4lnytYzwK{QV>iqsEvj>t<uPD3;g$a
zpL2GY%YU+?UrO{#$y+_<KRrR9dDXO)p5Lqf*-*fhwbm%{)yL(E%*h)R->L?mA~%mL
zTMkr2g^gezTe@%x`;5+uZ-pkQ=3|$fIG0)HmBmOmIvj4_qTn81^Gw?kmc|_abCA8R
zF?rC++s}rErQ$=L4Ts(A5yDtd)VSpZJfuRe7|gt_?)rpp0ebje4|c#*{*2aW(FZ5(
zOzcNAzfE}D)%DSr1v5W1atmh7ZwFIUhEs&KSeq!TzmDKBj+(fq$M6@;dtE+hWJn;M
z=;kLW!PQe{Ox|Tb>nk~jBJbh~S4RMSHl@AFQeIY!T0~SAi0!KxDf%^ems`JotjWT{
zB767e9?87cmJ-*JwU+Z<qSWXi(r^k(e8uUvdg|F_YdOrg6Cq}+M@q3}_tvET*cv#5
z2~$+O$rV9T9>*_HXAMteCx?7i5<~PGVe^<a9o|-`+sG8WG&0EN^XIR9nY!LW(!uLb
z2)`aWt32~Db7toob*aZ96FTjTi#*8;IKnl6E;!(7uH)W?HH-%^3@B4OH5lnLzg1zL
zV=eiE)$6`@kA8Sm>^DY>Z~3nlpp_ffFqPXbA;0wFYslICOH@Jh&oiWGx#?3cMUjcT
z4Kp5i*~IA2UE6w!9m;Ip0xoiOb##1#%0s7hM9L>v9H=m=4ULXcE^qj?Y)=`>%=w3e
z1g5m;v1#LOX%=%{r=ma@5G@aE?B9&K1efN7p-XuywPeiA;g<01aM9S<SgTA~r$dnT
z-$rao$dh^Mh%TteYdg<_D9MZZZ;xDRp{^l^b|Q!~6RAg@<`U*b2*u5nMxiFGW#-K9
zkSTz{FJS+-S^D|0pJKk2mCnSb4LZf#>=44V#!5DChI=m}2?t)EjMuec1(Z&H%V;tp
z8gC9>P-vR4ZF*o%lrYjN<f#?2pJF4hlrBH}zg`rwk!q5BGK$r-l%fNvOQK*S17ysX
z2zIPk&K%m|(qW@SdsYz=(E*pge!zT<E$zoHrO~Q@iMWT*^6d2DgL``1D$Ayuj7egb
z{=|-6L#uce441F&O-A_(-&*=XO|h$78S7F!n4D9iE2AF$-!4fHAgnJR9(cqXb`A;|
z<k<!I5@#ZYoa$Kb*;m>QxG$Vut-AVal@F?CvOZQ-RlU&3@5gm@Loi{``c&YeubV7a
z6Rv3cX!74%V@hf+WkJ){LTxtto$d})I%3wg$~33v%1pt&Qf=z2YAdJ-CIEP8;2d?x
zf&A%zF_r2hLsHh?x`!_%TH%t`T33IWj(t0Ds$U*Ah0i2?%VQAzg}&GGX7xnrr7}GO
zMea>|1=kcKpyGX|x1To>?}u=G3SM1ZbzgYbth2@a?+gT!WvYG^C$L`i)*5Kwq#Pf6
zig=jF`jmA}qDylm*IG+ENOXc86{0DO;(_WqW0Y}vTG}FmN>2UsDL)OMv8n)2bfPkw
zhzbyVdG~`H>gR=p28cFt1t_bR&bMgaGK3FEu-H!Ybo@K${eEGRB;6HLOkE}u$lQEe
zWP&8>djA&LGrgzP1Jd5lo@6gHeD*Uj{1E}A*{nz{fc_|0Ull*7way2)lMLu)p4NPN
znm*fmhBzJC%)KQvkRKu7&;4+G+V>Abp(ioU<S#vueLqV+dgC@Xl@>1SO-Ihp1X`)<
z?{}7#M(}eQE}YD%cth3xXeeJY#Up`sW)YBwNTc!~(tu8khFhSO=tgeYNF<VLVjG3w
z8rhA+xyb8oTK{;+{nrxuBi9>_Bv;DYVcmE}2Ml?WkaIZ&v%*L}`{zfBp|nXISTv~4
zqkWn^^8^Zk>MGOgj}3?_ajiqO5!2iS1rk4-HX7{#6S|<$RyX7OLQPf$nIAB6w{KaT
zJ==8>)>gbiE6dw3yaZaqt{)<ue@^YxL|H|v@waHte2w!lZG(sZA@sSewWS01o@}VX
zM%j>Bjk1o%uQv@G;d~Z}=tQl4|7v{+oVo!BOzf4Hmd?|_yEK4)VBC=ifz4>|MdvNQ
zo(v~^G`?3ueAyU7g*}m(ojm`y^5#q45nRLKKC%PZfC?-Kc!S^YCR{enwJ$<ueiR&c
zQB4m2PK#k>^Y}UR^Pqj*ts_Le7g*ggx(LFl*-6Rn?(We3%-!$yhB%Uv!SF=f{(d3;
z_jmHWu%b`|&=<+E^u4|m1qSzQWeS35YTuqx0GT5xpe|wqmqt*=2%5@qEl$dE(i{p2
zM02Elb|FoD_GId9$S|hj<qhkB%cu!IWijT*@2%p?yM1w>;sV|hKaKV<CqJHkfDZ26
zduhv<Mq-)0kg%2pA%xuWOj(&<?*#-`6Y*Nc{KzSV;Xr=Ox~TN!I6Bntd$a@7O~eUh
zC~9zdQ%c_9fl*!xes^lb)b|Vx#Egre-pdc&`}GST+}mUgeHc?5q`b`;YLpEg<_6>^
z7N@sKX3afEg;FAF1JDiwMy{LJq7-MBrwv|9<k)iZFkn<RrpmgoQylX)zw&7(;z?q#
zp)O@T63M-kn0v(Lx2Hy?KRoY>Bx!3T@+;3e54sVWLfs)oqp-*JC79}W-r(T}p?^L{
z5`r*L7my)w8w0f|uk55b{O!?8Py+}kCaXeOJ~O81tIoP8zS(5%a>BiKn<^Z7{`bX~
zfii5d_AD$^6RNzi?Tth=^)_E(yUocKkYGB&S>14gE%x`@w{PR518i`k{~wBhe2%|)
zFd)GdLTu-(-gesL-7^+iC@Co^r&iCjmHAn`c9MRx6?zIxnwFZkhf%93d_)pmyRz%a
z%zxttuYPR;8No-47s?ry+tAd%K8UTI(=HI#3h%;`3gm}9zIl&AWK`bPK^F3Su}xEl
zyTtz;{(5RfX70*8O7SQKb7WQw>l!mt7nw4mkAxdL&d6`06}|XMqPcKeg^9G)K`hhu
z*5+~2kh9;kum!o7lP<#=cIjA=OSszZ{zb=vf^M8-jpSmeW|GDz3y-CEGUu7r3MIAP
zEBjlSTY5)-oitLIy;z#o<6>u-erD_NKhr8V;CpgSQ7MrL%H-(eQ~DN8bc=@Zlgv=o
z)IIZS7+Pw_ak>D@`iGKDj$Neab+X;<8huA5C^+UTxEw1#ynioTF5X3#6pxX8UUXCD
ze_>QCWu}q06~>srLO^4RO4l!H7nmGQ^BI05<jN5{dUH@0ncGQSQI6!x3+NC_lK2M`
z^SDYqCiMtZ^I7&}9t>C-EW3QUh4#^jyiuJ%7kWg_Vqx<a=_uBFYOFL4S8aS`C0?l=
zNs{<y2TJtk^Q;x+wKS1fr66#>F)m>>Vf+}f`!laA<b@ErQQZvPrb}L(3HXnr*RQcl
zn9;?x##E~5cAsd~k^WmL`K5mSqZWieO*HeU2dMPI(uoMV=i6rPej>3vQY762XoJ=F
zDJviAwLS6$CM6`z6=HHSsE$z*)p?%1a;WV0co?Gp&-OCB-wV)JU8YXk+GVL<6&4&@
zB{5=mRKm&_rGE~_ux>OsKYmD^#kNQ4&$}tdMmnzTeYu|sed!tFwngr<$~Aul;`Kn_
zi$gx*xN+8rEo9COztPF4fBWw8F%J?C9^Z(1@`lxF!vQbs&Dlb{%;=N`Q<Ccn+83XR
zZ^0qvHK=YVmQt>*qdh+FE}x{UW$Etqr`e<&<sIeUZgj>H)+q4in3@~&hI}4EMl+vz
z#YQW*j+4EuUr|if`VXfV6BzJP1aEu*egvl+WE=K%l()9+m75WQ-J!5ZP*94wA*`I-
zvR&evf_3W%r^&prI|rG0*oD+UMmGE?vOCS6o8}6C4Bg1@Q}U|wdYDVKH{^QO73dc_
zLN2scXj?LDCo|9@<!xm=>3($Bo=teqcwi#BbkY&JHo|$uHh5P;B7O;L6Jm@Rq@Dg?
z#>>J)wLrCls%4;g*_=3~mqVQuPNw3EwCpyb2)v}`A+f<JYt1PZkH$P<@@Jq~I~rNq
z7)~eZ`s#<b>#K;<o?-~;X7253)wE*+q;WLcvXtHOCrA_U(uhL-|8}6NV*}Qaw&C1E
zEY-QYY)o@*G}w=<MTkV?eyg*WN(7nhthJWHR{m6&fpu#p9U7l6A2iQ(Lxz&u_u;Y#
zmbKffIIk{b;F8$q{D!Nxt#J!DYY`-tZIn_LBU3X!8Q+=0^Ix>tf{c2J3nBMS&XOYY
zZx&WQDmV~x8~f@_nTfXd)Z@KZ)?;kjC>`|uPfZ@brRJ5qk4v$2=b>{xY?z1*+DaP<
zA3+FNYbL4t+Van|OK&^ags1m>?6XF%7#k4jqmy_Z|2w#Q6f|gaa&25g=!MbjTCK1N
zTtgn_1Oa^4Mu>urwm+w+*W~EiEtS$<?a-67-RM)4+u4oW8+$XaLJSOAr)AI%Q+_xl
z4yvJ1k5_}ZC!$C;p5Jmtt}h#2hb!Y+INPnQGQ-ke?XjSk8&>?2QUC26l4X<9``;M`
zN^|+kQMD)<f0*HPOGt2FKpbC&-?yC*h(eWj#JFrw&?~ZvF7J@bD|b<rl6TUI(R`*W
zUH+~-y4>1Wu%R=`f{RBJjwCHyjwj=<pHLEO=&){0>Cv#Ly`e0I72N&DgFN4hDHLCd
zyTD*}&)%!?PPVKtOMlQSk>CkhLNcd}LlD{HCr>^aqi^je;yhOc>l_1Jl?GC<v2*OW
zhzvul5`!U5V?n`ie7!yIjcwR%oWL2a+WDnXSK~ok7?_ho-L9O`7rM7}34`cwog3B-
zL@dI`%U9RHpKDV^KJXkJ+dN0q?$o9?1Haj=jDP$^$nET+thnr*Y9EqP!kK*!a`_&+
zMxfu-&E@I3hwB&{Gy+a2TQEBf?%T8-Cbw>rxFUyXTfDN~j_5otrt$v%OIMT;+DCTT
z{gKQud_@0e9OwDoHj>mAQQM=)O>z&qwAOyZVa#2OK~WvYx--ET-TFi8Rr3eT*lqrG
zj8#F@UrMmfk<vrO=(?8Q;&#iCFe{tl9|0$RQsWvrLSH-BJG%JtCb@DXFVL^$y?#)B
z_a&`_!kt7SZa(#n_2|1}6xKSbwjaF`?LFpW?0r?n^P3`#u}@Ch;s|fKN-2M&qTlB;
zHONm#&<Wovj}In8?*}t%7%QEvd<LLKPeKXx+$;2mk`*WAjiQC6Qw-;P4E)&5kEYYM
zJ5XeV^^Eq<yvADd8sa{NV=_eB!a}XvsF@Y|qpUgkFN<l}Xg%6>qnlWZBo}M=vb+2{
zdX(|aNSK0CPV<J?qp1F)nQ^=H^}OBVC)!9<lePX*>2=@rUD;A;!d@T^NY<UImFpCg
zkEz;-__Eii-W}cCbqTCi7R;_dd8nrq-=81BA6uJejGO*~BXV~BxS<VM%0q+`VvfsJ
z=(N{H1PQ=aeE5*jHEdVR_vs8nr4q)>Pa7yx2b-(5RocEC(guf~B4+3(g5&HLvZ&`N
zcSyHYj#B*1P}Mp5&8vKcWFaHD5!a?h1v6Wt$aj2olTVTHQRaNCSeQNfaz7DE^RWr{
zDSrv$w{Ix9xLqQSCfHDIOh3ns>EoR(SZrk;TpciH+&b%e{O$2a>QQH{8$#FgUzmyD
zrQ1w1>k5y5T~Eg?tt}MT{|xU#u%hV>bs_(O3lXRoE*Nor{CT#(Y<xz$V8j9yM8jQw
zzm#P2aL-*?$sbcU){iH$el0y~!q-W{oqD3%iaaF7-uePVdhorTyhHg!w^t+v$^%{{
zqk>}JlN!Y$dX^uI_dzvFp$I<d$iD-wK-oS0UGl{AzyJ*z_eyiYb>z;|n-3;$3d-Mk
zBB%aJ;^DSoSia0&OizV0QuN1frm;-2ZWNIu-KOo!uuW?`6T0WVUrM#xtH_C$Mus}!
zBRqM3Nw|_J+{(}GyDuy@S6K-x7<+l8V^ie`-x?;}Sm(X*`|J9N+R}#iEU-0Cskk2{
zc>@qM2t(a+-iOYQGP86A|7C)|wY7CkkCGS60=V!TGEtC<5v{MtJ(y+k7;*!zp%cSG
z(dPBNjy8H_*UzO8!)Xg9#irCWKyT<7wXN^@N1&Jb3On#K=z~%{5SDDM8nJBXgmDAh
z1N@rx{ZT;u+b_8NrHa4gMUQ6hw&UsWOUc8kw59kMeACdjMGBu3|3flhd<1lvJy^#$
zpGywp9QR;Z-ZmQM4`UsFcM4R7jUivT@y1mWdZEV4L8WVw3ToXxXM5AZ922VmP>oH`
z_3YrBb1>}EgMLvRL&L!bkS@uWajLQJX7htC);Kf_JGqx3bG=*@RO5!NZuVD4%U!r2
ze$Lw!l-zDZ+CDV<H*{3Y-=%Bw)MWes#Rs38p&(M+V9K)IN`-If(qZHOxTd?Tv)ewK
zO26k<s2>&G9}*MU)#8;u(zcGjtI4Ph8d(wJ<EDx6Ut1{sr80S0oZ95j*ma80mO%yF
zj0tmCe%se5`!YW_=*7`?ri)*!ppl>QJF*e*#eMwc+@P_!XFEZ(y!O*|^o3Za(fy@s
zuFB`q+G4_<_QZ|3x9oBu)j>nJgm&vCuMz3SB>{Js7(oa11nbg-!;h;`3okmUI~T)-
zmi+@jjh;V&TS4Nsgr32F(Rz9~%>;U1zzBPBE77#{(&rsB_onp(s4B^=bsj3IyaDZY
z4ewdvk|(CyD-z#^cV8KHYmNM{TUMi-O2$OcAQ<A=aT)qXn01LCxC{q+$`NJ>KQwY+
z0;e%6OJKw(lzwY>!Mi(|Fz3tg$d~>6h?Ww$Oq{Lt*lpq+)wSIWsh1Tyb1Uti)#q|5
zuS>eflij&yoCys?Ux>&Ar612m;3kB~6bdB$k(s~zmE?oU8U1tD#;A5Llhz_=JY3tI
zkB8e^@1VA`v!jCL`H_8hexk%4da@dwzZF_)d6stjJn?mf=8eZ+-w9=KMUh>cF8IAY
z8q-tYbyU95F_?Wr@|57ewvT~Ye#N5rbz#&wpv?sZc|YRfZg+<1VwHDMhij3`CStWK
zTx_z>IwC#i5$CCEIK%BPH0FMg;Jcx~P2>6{Y5^ryd-*O|Y{HVw{+X!l^Sf%*m<*wO
zsPFD#s$V@w=huB#5<B*r5yRv=>kjx{Y-$@lf1~)GUiB_{Lpj$Qx$q4V`2lmuIXBnQ
zi_CynnmH(-<cG7%mooGWRv?K&9)nDbMrS8jU=cdkW$K3`w|f}25mM5VOQf#P-Y(%x
z>I)eD3+%bXh60Z+llt~nkknErAD*IV=oS`CcGJRsO4X{IN)mh9MnXyDj&eub#vwXo
z#Sp)o-}>J9sSug?<w^8~(d(v+<v-ubRz45^YOVeYhv}ZhJ`SMW&h$Fwafv<Q@TtFT
zK*`TVaK-FVwtkz`*pv@LWAi}8#mHc3X1T-ntubuv7I#@0cPVMKU*ovAm<VS}zjbhr
zmP&V8Yu*(f#ipK_Ck9{JY@`%b;z7!z+VRr$Y=nEZm0s=Pw{dZJN;VIZXjWccvMjcZ
zKtuL~V<!LE40mgp!p2zE5W^J(rmkzoBz?9z@-2}T?~9%{_2))mRp^4ay6>P9?QPDt
zN8Uc)O2U?kw;|Y;VrMt-QRtpzW~?c9pKWW@Ni;Se@eIMY6scBD5fVWrw?R)-=gTzt
zs*ZNMKZovNq`Nkk|Ic9GL@fscRXE!=dKPz2uY8QRp<Lhm4g(cG>MXAZv;4g6i{+#!
zx@%O}(&*i={kDpeIrzw!qa>RANN=1#r`ag2aA-$lYNH;}zWs}BMz~fO1J+vVHg?R6
ztIh4vfI@1s#=p>~4I8s#)wWo>OXAO_9Bvs|D1ZG`>VpSCL#O=Y(jtRKpkY(e;vh>N
zNN09HOGZFv2zMiK7UL)M0up%)Lc_q}<&f-%6pmH(JFa!~@)>G-T{&B=eE-OgFm;3^
zG7!@mUG}lQ3U4PC`}LM<?OCt1ZZE<Ny0woSL)X);(eSEY<x9J*Oy&7k5!_nX(gbm|
zl!MUn)!K_&$N!nTA#34B2FO*ca;SIsHa2Cb9iP~?z91qluF9<V9EbC4Ihex10HtO*
z(A;O9knQNF@8-53_}FV_bH#QVIt4Y8^a}C{3hV|43eWMAEl)2kWGSm~#j6dD*+K)D
zj7X3yM(&vW^%`YWkq9y>`S%rS9SLYW#^vpXOIUXJf}YjqJhtxXH?J`R6pcsF@?HJt
z>_ia^mAsfx!qC(vt-Pu$YQwJ`=~p!8XA}AvN1HOVwzO3BpUty4;_K&r$_T%-kE}s@
zJDPZ^mw_A837c5`^=kvl1-KDFlQb_h?JH-~dOQHemi!wSSkFtcTy6+1_>qKnZ!m+u
z?4;|=&d!dRa%CvfoVSAwa@6TWcP?2kspZdgzV$3kjT1%<HjHd4o8rqnY~LLPf;Hc(
zxNhn~k$lF(4=_}TCt8~t!il(~K%BJ%l`z8URZNDFb$f{<zL^0Vm#l6(#&hJ=x*3j(
zXMGg=@O9wBh~OyYs22((L|f!WIg45<-@i~uDe3}UAdiPc+Bb7LvG01H4{JrvQ<5gC
z{CXqoM-Bv?Rr}M&WSIY|XF4Na|F7zqg!YZ;rH@&G`Jmyjk#bRq2#r89?w}Tw1kCA~
z*-0eS#E(E7As>V5#`KILR37O-Wskd`*zbc5Gg3{dC-E0ZFb6L7sUa5|@^`kQ0<8dB
zZ1B^H3h9*ClzvF|?}~_uPTW#2J~Xii8U3-3G?cTGEK91rb)1~dot)O{(5H5zZ+Z64
z`EfZpxp<n~1n5Z<W9#<r{n*5xHs|*}tD88@?(?aEpX8uti0XcNdQB+wtku(wA5+V0
zg6=j7zn4kG5ZzfG=(MVV7xd`n1gYjFV{`LH!Q0I@Kjh%mv53yQXxv+bXhi_I(FZT~
zE#3QV$RboFmgI>GZfT1I$_4#gs~?D?g}OV>E(lp8BTbduSv=~9&5Se>SFOpkgpngQ
zhH;6=gw~j=!WGKSIDEI7b*@m<&zaXhZl8bEoO|UlEjotmWTZ{_deX=3UmsV!u^+|z
z{6046G5pI(Tvn48`+AfhHh<Ln2>eX}-+gnrH^cjeE=C7d<?QtKPaN|XWmxrZl9rAB
z%!D65e(aI#+=4U1?Ck6;2n3~iX<1nlo#Knssl_`}AGhxiLwGejJy-6-?j8Dn6X}+@
zN0Q$B#vfa7l~<*VxIQr^RMuPV-K+Z5yy2|O1T>fJyIoIUCv7#58JpfV*gI<ufA`7Y
z>pSHKNu7xE6G$tn2!bf@lwxtNE~%!6_EQS)-i5DO-QShZ^T^QbJA}++gv#{RS@}GL
zRxWi%+w}?b`zH<J_{jHDDHS_p1g3eL&0F(C6RR2-Hb(#3-?%taKlU->#1ay5TUyr#
zRm_DitE`1f@a=nG9(YT5&YOzq;Y2++HUbhm%e@Vw%0FN<M;fs8CuFc;;}`MNu0pAb
zY?KnMPn&sJZ>;LS4T;3GS+Nk_ocu2}5SDExk0Cp=gui%Hro4ckw3Tsb_j)7o9IUL`
zft*Epdb*zl$A{4xFhG(^yRRK(w~oA8`vFqR?kF4-Zka!nh&@6{qN|T=jEp2_kJcW2
z>6F4Cqs)T`YlXCd5l(8#e2FcJP#N=F`O@T?(>%S0X>T1(6VU~?@+m`+In#ILI%l;q
z)5Gss+fRsUDN6*7`QCNGdF6#?S~$cA>g9Q5_r7Ml_4KxGw)YE%_czErN{M?;v*uAt
z%YjS6inABJddrrl)w5eyB9Ty6WVt@AKE617c-EZI)bi#O@d<ISb;IoC@=R|X`O+>g
zT#B{f_>Ohtf`e`Ao!q46r<$hI+efmz1FiR~p&CXRf69;1H+<VybfR?O`=0boMqk{-
zBOSD_R;U#H_S7%}YGW_e!VrWYR!Dyx)W(Fio*Ma*^_)TwYGrOW5t9Tcf7}*$_#;cX
zF+RA#V>@Wp+>nvfEZaMN#lP`=Airm?oV>g|Tj%E$u<LdYy(QcMj;nABxUOcvfGH^Y
z_P-n)yjvhZc|}IYT!l{(`XsWDubYT>4*+{3Q4?y5I-b@Y3zEE>r1xbI0R{&W7m7!5
zi!cYzq5`C6<MFSTE8k0{{ukxsZHFm`xAeEKjKd9bM+i8M0}Kwi&%p;J-@dnrdjp^V
z6tWHF6vg;jc&+sX4oD4X#Uy~V&2E-=_aV`|en(9xeNv!|$i6}^M|)2ls-L(A*Ylg3
z_ipPQJq^`H7_q9lgCeD`8UZdpp#gUP_I}@Vc4w#05OF^gv@d<#ERUM`X1(^cIn)^*
zdFQEy;Q*W7;WU#J>>7940~Y;mBFzJSG7`FTrp;v34<yM3PM>&Rafj{x`O~kiQ0Qlp
zX`N2)SZRJ8E1PD2?oq!Oae8(8^Sjp-f+k$;$sNN7@)K?OflGD7{3P+d-FK&Q8MY@T
zCT2V<R^wzAk7=yRc36;$S-o`ZZ#Wa|rIhNKjUR+ukKXm`vx`a{!92CHF?oGrCKYyX
zWhlJrCMP^|KZO)FH3h%TWg)JwTe!s4jdQFvZES3u5DzB#zOWFSaPpgy%fb{?Bc}A!
z2E4-K@uxHyCoyHgk8^VGRt>8MZ7%Ou4=+CI=v%HD9^Zk>Q;9w7s9g=erpXxexU$mw
zZFG#e1o0GMZg||_=FOWbQbV4>!NCe257F(9PEYUc-m+08<7+=gFT7>J+Q#}+g_-tb
z$wt!2j&ny3b?Yuz$!=k0UM%@S0z=g)bC$PnUFc9W^G1A~Jbo-^Hc7wmqHjh!qJty3
z>q|jr{4QV6OycK|>4r;Q`5%w}NxjE^k1zV6Jv{D)IPiNZ-x_9Df7>W!FUIFNni&}M
zJsRJDIiOpo`&Q+B{QMY@nI7&AxrBh}w&C$}FpQwX-%UPpHh|s4C6r_olD93R9IJ~t
zIXSaK9UQhs$Y2Nu%~WMLwF_2)4{frPBYuq@jY3W{1X)+#esp-kvr55PP7Hyrbhxnj
zO`#>=aaq{{rAD3k#WR<`Sm-kjt^HRE;L%EEW@c(C<o2@(25_;~nU^2Q?4dVvxqw5T
zXP~E-plo{&uk`YZrR~)4@bLD+D<^n$F27rC9Dk>j(q$J35C4XY-5gt0R_4Zm@%LA3
z0yUe|!Ygyp=XK>XdDUR<pI_0HV&U;Yn04tLU}ZIvd|oJpg0f9YUQuz>NF-Hy?&d8;
zwpcP4s4;d`<l5#&O-uF`bmk8`l)1ICo@-hyZprd0Iq`QvKb@;33XLS*Qp@-$tW4OL
za995@G~^SAL+`ADipH;lu7;Bo417WS%U?nJMl=+(L#9JPApRW)l36uEn5=BMmd7p^
zm_rP3S(p$~y6)%aw*(`;Li}MCDRC-jwTJ&;Kp}fT4l%=21x8*i+DZR*hoO7^ntp!7
z#Pa5?AM$;xd(#}NA#2gOZ!#PXOPmPgy<swRq;`v8mgL1N@uA-w%d@r}!9g3RRqTJ&
z+UvR_htV#dhZd@<s@9`Fl}BsmT|LfHQD=VRqa9U(6m1hM>fPe2&PQVemmp(JBbur&
zNEtzj@6Xp|KbF#!Bt;u+8z<L+M%czx4L>^l<b9)ES2C*V3%Lex2X8>dXlE3U*YK`0
zPY_P&`V~P&6AAPQ)wy|?O*ica)7pnmKe->@`8gFN#M7=6Ke$*l*p}at>s6N(lXsjR
zQ+5H7+gkM~l{}yHK=D8}rpR-rsq6c!fr&`!Eqhp24R3GnMNm;dF5m6#?d3+^Xj&hB
zd(GHleN9_%B?+pKCp^f$Ojp9nrbu{l@bCmVd3&!4%fncl@7sHyJ08}+0?(j4V6x!X
zhWh$plF-4Sp(>_Pm$=_Q7Z(>pm9EXC+>?z<{PY7==*fGgqlBR38Rw%Uq=;u--#Lf5
zAf2H?<6k%%kZStrNe6KkV`q0%b*$@^)LLm($7-V^h^HQ-jPMs@qlpl>hw{&1XN_$y
zts1`CLJ#J>_z___@Axn0?GAHwS9f=HCyHD<X%iEezfirqP4we|8SUbfBrSDy0Rq_<
z)sr4LIXiQSuuxwXOgfGngLHq;*7yE=KZCD#lgEnwt<LPl<((4^3)K^E@fU_VQzipu
zvvoZNyZ7x%W=q))1&cD@GzB63F*0*QKdhiveOIp!!drl0#~CKo1it+$9aqwQnVB>^
z_hON-xKyUz(FnL7vGndPcrP*SHk#rlS1ttIvF7N=J03Gsz1_@@kcWqZ+qou^x_MU8
z>rj_nMFr3wyT5~)_tRT;?BT6nWTn0<)VG><Qz`q!G3B_Ehc!e}AKmPVf0s@5+#3{o
z4nF6@c`cUm%q4_X`(h7@>eW+QuV?!VhaZn=gx%#*7M=G-rUTu#Drg%Q-)>_KtJx;C
zW{X23M&ZYZ8KNlov$hm-n=5T?Z5eQ{^brD!+aZ`Z9NfoOSR<Nqg;(mY6x$)ex3Q@V
zxpcj|9PzhTLi=eCarZQQA@$wU)t8RjsrJz*gRCHY*cL}|pG%vYo5?<lA1Mayk$wV{
ztyY7jIL%1vX&dL}<aBCaa#Dpc(prqR4K0+ysb6@7%%!aOs>oT}xC@0>j=-6*g9=l_
z#Sb!kE@fjNg?JtQ)vIF$HXR(?I!Wn$vu4k`F*XR$)~;vbn9w=Lbtml^tcG6xsNU|<
zg*}WA2_&wIWjgNk`FF_h^~|oG2%DasZijzwYC3ft)>bz(biW^B#d$un@0AXqwW8G*
z^i77}ABwI<NHqz7Ec97)`0n>PRJLg{zILgU2lb2lU=lMkGbx~3XC762wZyed4#r-%
z!M-{%Lu><b@MAx(;l-y{McTW+dP#$!6+RU0rIzbnV$nz3)dzzq-63ntU_kN$rFJjo
zi~`TGKVmtTT~RhevlsB(VbUHH<)Yyf<|K`Xn!W88-vrKp29GZ2PMVsUQb#fH-&j~$
zvdYpgJfqSVi|@Iw2=j*{Kv0lQ9D0aejj@fnR@Z`9dne3k9#vCOnPf@q6a~2S3wl#?
z4U4ZH!lA`VCG5}IOJ{sEeRY&FViacm_Nl>_Fn8+)%`hk_D>IO564<Hl!hVyGfPmm}
zaq-+)A5HifY_8HOp(MQraVHO6b(>Tiw{6}y5lINJq#xnDepssTn4)yxgFhNMXI3^Y
z_qpSn0)kjC@pzlFZHh?D0TD&_1uS4s)fQ)!8@xlfwS%S0KHka{`NUX^jk$R#3e=24
z8mkvTzPthM&R;r1>`L}YCQ{8H4Ce7JNEh82hnZg+40LpIbhNZ$@4$GQ9v%P-I0gOh
zIz#r7A&jKafRoNZW%NuS=$<X`^6_2dSM<P~eRAo|R<qSga0?8ROokI7$`+e(JjQ0|
zt91XHH+l}|Yndf#lIjZo8iKqURNu$0E;^4&g+0bv?(Z2rkF~WsRwM|@wHw8&Azi1-
zG`?$ZcT9wU>(PNh4rrs|tDibJCdS8a9gOmw;wRNCy(&@&$AL3-^4V39F1QlBJWEip
z8UzPB^x=c7;#U2EtBtfSMT<G$tJ351j#q<5+%<5Mb6~wmjSFO90l-J=TxQ|s<a8?W
z@hYw=C}ZU(gYPp&5I^rpCX?q2<^#PqaL>s5mqOwZyo|#MHm0PMI<lJAi-=`QLeRGr
z;T<^KzolqxZEGkrL73qsZT!^q$Vd&QFGqD2A)nIq#jyCpn*(R$I#InaiR(0VX)+DJ
zY#X;8+>#92Ouuk|nJRX?Dnekntjdxj^5hJ$J7@Opkb1tk#E1zzH5}8|#9e^!>6K&i
z9vqP&BB^f`<m7l=+}u<LhKGG?eS40_IH#ag9+Pd_-c+CRSVJ8Yl$^QAznqscA1c4R
zEY>b4=p|i-U;0FSZ2uURHGSTnq1>pgaiG8d<a;m?in6jFOu?mo^*Ut&f6U#ie?K;s
zfuf8MFp@|Q4ntAUi<Jp+iV)H&?W8k;;H-(T#YVfzD_<3taCX)?^PJbjRXT%bMMW~E
z^YhoFq@?+Wm(H{3SXPzIQ{^4Mb%o@q(e1X?#^c!+*^|5MPQr?SZ|UM-9&(7<Y~HuR
zqwrtxg9cNg!bK5(?aPr{>>iRH%*#u)nPg#=VG@KR_|;h<S*9fBI7DViq5_H?DM>LM
zDH_>Wvo)&_W;@vQ1_sX!!n?28SX&e0tW>=$lB#A?7R>yfCg^cd(X82EQ?H?%tn7mm
z>D|!+lsll9v!NpS<NJ3-e@G}SNu+nLz_jXIZxBan2SWt6cM44AxpUq9E4|=44gn|{
z+ry(6?7&V7#Ry8q<UKsu(P4u-@TSVJ+smqN#;58RUWFhCV4t|j$$B<DA6L7$XvDAI
zL5ARCOk6I&&#tB8vZSa4;I{z^I_JGs=!iogZ*;voRcjX|OZnGOF-=jJQw8y+zvCzi
z6IH8s^3^)TlyQ}Er+#ZwV40fGjh}s<DETp&SNVrx(cF(8rdQ!fk*_-!!GKwj-cp7S
z7@YKt;Lt$;Y*7j~CjGY!w~d6@BcC9#!&fBIo8Z|;;Ve7jU%q_FgFK$9-v#?D1g4e`
z2dT}N|2%&vwmkv>5D%$l-j(7_=%{@cuUa!3psAG2R|j?T3QOZvF6}UPj}2^?&;g_x
z2ZvUAOf_gmQunf5Y`e|CYNH4ARs?$2O(w&|ijS!xq%+=0-S3@Uy$}ZO@yW>!dIkoC
zz#R4iKnDO>9{fPX$qpkC)|u7D_YZ~3l%(UVg!RAOKXlMZIq+<;qh}Ot7<@Ers+K$I
z*W}$}g3^Z>K+y#>0X%$twiClb@jg*oWXtq~+^1fzKo&RXjcH&?6&nV#)(rTn9nVis
zM^_kl>z0UyJ(T}1$#?9?Nu^NHqjtW5cGu=3m}9;I!$JI1W>59iZJo$^QMSe5V!}JC
z_u1j<`Ij%A+9$7`uWs=@Pc|xj*VdlxACMM%)fE0AQ;*5%s*|~Dq%mp)bz*X<-2F54
z<-IuOSfPE<We&sh9yk5`TKN4#jg3bt!<2QGDu>HuI$UR9@H?{%RoqFlyyNw2Ffn;e
z!EJo4U<<~NgUpsi;T5NBeV+Va0*b%37IyE~E<5Q`zxlaxST4?t^WWaUtVPe+BG6E)
z^oaJ)u+EuH9iNy8Jvrv7e;~dEpYt;(CkK+GZ=8BN;vsiO=#f(zMKbfea>hzBm1wx`
z8F%b3ADN@Od!O{#o<u2H^|=5`i}koyQU#ZeV`1<eKn+!`K1%n%iHVTafWAOJQKQn$
z+jR-{Ib-cCF&iR(0tryA3g&O?8QzrpEPS|w+VkAC9O=GQ+8f2MnMG1#R)TXsezaD;
zYZ%1skIivW+gWbwQQ!H<V1*oYmFGTsD7p|f)=PG)AWxRqcEK<((BS+4YU&pzTa{&4
z=N*@>GUXsBJw>}*fQ@}9S{nhd;4N6fR~f=k`lhC+_^_0XDfm-hT;7}4Fzr42qH)F8
z{Mq!AFv@Qfgj)wzRm@^rO<i6|$%1wM3Q-ZKvICTd&Vj7%Q5gO5TN0j0@oJxokU@6U
z#+$1NhX3%E-BvI8p8DjWR1;R6e!jj{qlB?Z5^cXp&(8ODpCMFlk&5rOBx^({awK-T
znVC%tK(PHb)@tWd-pP&}n6PBFy}iwzreFA=YM6!k?o(~S%P^+tY?!mc0}hN!nK4*!
z0z??}iE(ju619YgGB-E3Hf9SITCJsnF7jX)BUc%Ja(K0oQ6N=6ezV!CZuu1rzl)Po
z#!ct09MwQ227nzV66q=6+1$$3KdtyX0ji3EC#tKds`6Q@5S-qgCYDd&Y&+kY-KhII
z2y(~!yMz8yc5HtG>Vwy2La?vR%S#bry?Cr|m9Ki((hGvT4oD9U>-$XT43Fm$#vk_H
z@ikW-H@av}VvA^S9jllGvDIxbW^YYhAW#6ImY<Vg=v@JrGDmBlq}5W7$wODIm$*R$
zz{APeB`YUKC#Rs0D|?CwhFrn{bL+N~z6(0v5)5rpY+zyFjhN$X+CLkfy@7V=fo9d<
zHypNp@Vz1_U9=E~*#KyVP-G2^nQ^&U%_0tfNI?eA!012tlio3<!ISe-;{Y^}6I2Ln
zgna^aTyG47MAKjac9!2i(OPN%Wlr^*AfG7lOJzx>)}~0Nkl&vtouEjAH~opME-NdW
zA?`9_Zst7!4Au{tC`Wm?xSV?cjRK-m0dQx2oaNE;GW2J0aq&(gSmPJf@9d?Y8<mZR
zLK=OYjruN&hEU3h;j}sgsV3vkpFhp+_sz;cd<t$^{^5BcQ5Z$FaPXDL<7}~jyXBHE
z7SSM5^^aKsP}I`WasyK1o<-BG^+lY_sRG<T!{RKy%t{#eF;e9-VHP%mXvvB_fJXlO
z%c6w7un|C&3C1X!Gig{_div&<b#=d)VX7A2TJXN7?YoJFhK8k#XqRiMVVQMFG9B|M
zi8U2Ty`sV2U%S)!dE7OmGfGQgJk$hK1sKEayE2l7z$*k)wFRwcw;W+F-3EY*Hmi`n
z7OgF)cl3~fY44TeEXgot)a(En0J?<}r|dz%S2#~490#mg(kZvyY;{&Br5!NU!LyRx
z%X`BOe@=jtc?8XNOXiLKD`#V?^gmzvbH=#(z06yiC&CgD`|{2+eqlX$XUyz;l#rlT
zL{;r}i7Cy9>$<1@`r6t~vS&<+k=g6!s;!4~6ybeH1w>C(`~mKq0o)?h{}}dAGjupq
zgV8J~M2iVMuM~;A_*MEoAe3>Kvt0qPQ7_ySU|21HG=QRI0p~pGo6TgGv<!Mf+(igv
zoKJvL`VAZ)C#h!Rqn+a<LfIE*A#kgOkq4|Warfa+p{{iBAG4}=q@K$E$PIFetkDdD
zDqgF0j!<d&!9d$CZrz&~&ofFp|He+b1!>M4+kOTHe%-t|uL)2B0-CFt83%^|423YS
zAA~V?zkniW?~AJ<D?ezH9)sz-EpQJGFE3S_ltUcuAP%BRIKa&jMPbi-E{m~MLeP3&
z@2Gyxtj^ujhco+T4P}@rvu^#nsawaEt5fFp*%q0#nUd-@bn)sr5w;f<3iS)U=ruSZ
z=Rkjb^A%vVMhJVGQuW(sYTRck;ONH92K0j^N`NlrPw&5TaTbk(>Uuh$34e%74onwX
zjEw{++-UIbzf>eks&11Vhd^73cGq9U5||I8=SOVfqT%tr2TU_pX`HFaPwja658H=*
zmQ(0penLDcDD7vHpA*WrDb4IB?u6V6ml1Bok=C-+mY7Rekz(Sar8CZ}<@PD4LVXF?
z>bN0i3+>Bapypj1ikgdQBx*6US7YxDy8ugZ0`PDANqP7$gna|VU1|*QJSwmO&m2<&
zWtA60UmO7sOT(eHDU*Y3N(*-8PYjz`Viovj!EV`tzMNqEcW!)USl>nL{t&)nj7C@c
z3+WM!pbsM{LrJwR<hB_X-0)Ti-dgQWMhzC76$tQwYwP>!bq+wh9uPAp5kM6jj}bYD
zals3XjzggF68NOat{*<<K``$H_%m4zh(f6&1o&W2Rns1$ZG)*-I>A-P!~kD&>LiU!
zP6mB|P*o|r7a-JAaCrzA4WEt3IPc&OLlImcO#gMYLsUkK>Yzi;Nt>T!1JObk#6)zO
zlL@Es;JsQ-Zq#dXQ&4)Q3HdCzWS^C=etwd#n!PrV&E^rEF-bPw*W7c6q;{MYBApC)
z^*q2fa;&~~0Z@G9_U+ql)*M@xSpp7z2*Ug~7eEbs72FUCJhI9yk@^Q9tm<(Pyofh1
z8ADVGJ@H#nQBh`Zzj~E`wLxFMe%)PHw-ZkwU>ty~gqZi!HgU<yQY3ZY?&&ARTfzQS
zCcF^0-|umVa<)K62%wj*+CIqmM*!s^2|<M1aDJm2xAMI@UVkNQ;6~daCM3b?jA{!#
zvB6mljiUXO#>5tO-c1*H3?S{wSXqI$J7C|ORb#<<+&$nqVbz}XJXZAj@+L1xEp!<K
zVL5Bs=dZ%XEm+#N2Is*KGljxu1K|A7K+xd{QxBAZ(93zw3&!q|yPI1N0Civ|;J0V4
zm{nhioq{(4fS<Yj5+ISdlOO#@(mL|~?&lIq0z;S)luYBRgtIm>F-%FEQkLwjs4SWX
z;k%VwJ@1rAKBPIjw)a$^%nv<=$Q`^qrtIPMvNRnzO-6_xw86{*{3p|7Oziy3+PC@&
z_ylnaD6Ydn0`_zh5@HYcu5c7Z+Sk?J-@sy34BvtXu*=Tt3kdd4D=F2T9Ea6@3j7lo
zmJenY9iM(T^TS5F3ZWQo_>pNIx%n@YK#btfXGhG|N5Fr2`O4%fv^&&#$nIHr9;PJW
z-$xRJt?0_XbzCHP5-uhPlU?*9es)e2do6zLECTIf3)|v=gK~ymoo%p=Sr89*!3Az|
zL^f9q%PJ`F13hS&DSp`=Y@=hD+yVD`z?l|bjY;&L^8D&mcR&bAzUZYqdri4u6iy91
z#Uu=!80>Ia0gG!04`_7?tgkP`89E0$4~$3x1dD)=$lt-UnT3Xic0y1w4m&N=Wx-C@
z&p4^$AER3G+pA4Lu&5@TgrDq2aY!)LauDGRL1i*I0UAOaizZ)qy46ckNED1Oyyam2
z-E2kv?qDe)<^1(vgg|F6l#opwgfWGKb3s6N3|~AH`DAH3gs=@%MtCl`1LD$5%O_~o
zRo;iGDUFUWqvTv!)i)qg#-<wlBqw0X=_%OZ?hFx_n?$Yy<_+54v0%5Ic`t$ebBx1*
zxdGc)43h6zGw^{ZD>&>*G0O~izyLt2(6w=ThIkvctRf6Xeg#wk!NDmmE>0GrLKdBA
zfU7PnI?*P#;=i7Ff9Rmp+ey2MJ2LaRRbsRC=2@Q^$ca6QC1`7(`Asu<C0=~!#S7%d
zC(|~P$*|0DU!k4m3cM;FiD6$B3)3$=v`!>aTxK~oFn4nfAkb`9;9}hbytFRUK!Nn`
zY##ua)o^R@IzMiQoC5X?n4V(?tRt!BeMcuJ&R@YT|3lZ8$3xkMZQo<3vX`BbrLvQd
ztz}Rl3E3t~Q&baX$i9u1$Ic^Vk4lljP?8ZbQG<w7h-7I*_AUGNoznBZ-yh%SkBWZo
zx$o<~u5&rg<2+8@f1IzuT#6wmxa{4w?e;M>3R6^RyRl5-eXvg71mf!$i2wi{V%%Jz
zQd@~quBjdWf<u7*__FKJVXpw@1wtS;vcn0$*Pnqi-h^ksn*c5ghIAW1&#@}+;Ct{j
zQbYiz6B2k#IW>3uE%;p4`>q^OKRkW`qZ^erN-T<74&c^wJ-0W_b{mnGjEg~iEyY={
z(j_+haNeO7b1j!-wTyT#i<h{K&D6%RLy66fhHn%>*%`V?6v>@>+3aO}a|=rqZVjq%
z=1%zB;f~QraCX{HzCFifIPNjnbRe$#){5ADwpzS17=#FwoQpC0ZE^IfNr{L@PZd&4
z+##pE3E$1>C_UW~_h=tjx}@9AUxf|K4@EuNY)(41=i|O8yBg&&5Mt2Amlz_&?KIRT
zDjunlqs4Y3v}I_v+PTn;qZY5i%|NpowXVoS(^8^Ta+q$lcU@F(=I5r-r|Srj!b;m~
z+x>JxKH8;6Jlcng5fT!ykWBdc`OBB<e*ZXFYcc<geP5(6d2gPGq87$pCU<7NTca7C
z9^SnSeROKskjx7UGP9ooUeHF=pmlR1EFZ;|SGrGtLbf)--!GQOvn|ROdt}U>XQ^cT
zN_&cC!C#pR>J{+{+I6`C()-YDHgh2N>;S#>WL5{JjN8VZZwVWrWACIdF6Yw)`qw9I
z!t9%8vf3t}2rYzuFe0cKxxX!yDvEjKE=z8HPtL16qpI=uJ`ZNO3!(>p&o5Bb!X|&X
zFkUCD(kKGff=i#u*~jnNZ<9xbuwaOC2fM<I_bpPGj*D8eEepD{Hy-FSrG{&*ze426
zaYE$G1(n@M7indoKPiddC>E4Ud94Rt3_5o|r&Tj&n#YbBfU`c=w~$YeBF>L!1!JDm
zQz^F$*for&oH5T`cHo)o5*2q*V(E=CPmcZ#vHxx|1*RUPl6hf;q;s|R5+!|5?Mdln
zkqH-lYOtey*g4E|r~MOwp~ybV_%h0b^-#Zy>DXIokxZ^i4(sUb_Vo%Ojae+);`O;8
zYes(z!=oL7RBwVmJ%u9KZIW=DGVneW35`qlCom=7x>C1R!1=@Fk!W7#iaY#o&qz5#
z-!^J!zYA9q%r_RtJkL3Fv)HHdAopJo=tHgu3y+o>e_iMNsBQcO7ctWMFk#i#i&F!a
z&6WIXUyP7NI<vW)UVzF}0Op2M@vFkJ!}DI$!4WMT%&~z^N^0OjU($ihGey|#7O$HR
z0h=?Q&}h+2niwt0l#5R<f5x`R*d24X%KVuj^=2C@g{dhpEZ1t`HU<5@Olx{WHWlka
zW`yQ?fZQM&JCBX=??<o1s(2uAcDQPbp}eBn?-4<T&4kggC$tnLTiobL`8CDjF1r;|
zY|}?eR$H8s1`V_Q_nErf$Fvmh_O9NVQr;e0ql|oFZ>_@3fTwYq$%5EdF)CcPxXJrL
zx9I*37W@W9B-&zPUYr`uuf0n{7|Usk(>U#oX~V^;gq6;x&uxF+UezR6f-IiZrN^i!
zy=c#jT>+wf>;^h>8|HH1TAWHw=%?)XIyiQNXR%H9+q<T(>N*wbcDJ@pXCA(g8h6mW
zotkPFp~_jgE&1`+=!$a3Tj@O08~Dg*UiTaWQ<kz911*Gx$g&^2ZGp6mz#UT_ZFMKl
z!$6#BmUBP~DQ};9Yl#hDCL`Vv_U~s3M`i_nyz<zA?4<+<yV8>i7byb`Zj@W^e%bBr
zb4lrD%GmZLDl>;rqFoiWqAR`QnA%n$B8YvVXJFa9Mb>LA#_*t#PD|jvs7L*g^CDOW
z*Fs&F!nK8fW?YP|N6bkhbIT6Rmf4U-yNWM%E9Y@;%1aTlWM92v;l;4hmzl%e+rpx=
z_r3ZiIIC|V#&3{NNDGjd`h1w|ds}62f#}$gcu<vwVY_Z%{z5Ee26we^oNC$<TG^cC
zyj?@Ry1FOp^^kuNQ`F~0?b(VWxYu-IE~9I>^I~wzY+Rd)yX~C!+Otxh_a{kLePfJk
zl~+BmO~G_AwU=_V{V#Vda?~X^x9OeeVW?#BEeV!++Ly###j$9Lxv`ryh=szH+~min
z$fFuzJb%6K4(m=#aLg>><Uh}@MurEx4HH0$-j?oC`e7HbK)-XTkr>~1Co-I1Y<iA<
zHT=sR5gx_;Gu6Zml(>*rY;OOZ2ie;m#v!cu5`FhNY4gXZzkcG!712-+)#ati62Ivm
z3K@@6&MPKI6W_$e#c7Xdc_=2!K~(cua3S+5Mqim`DWe_og))k&QSScI*;!r@N2qoT
z4lsi-wq;b%_vqX(<6KZY@ESg}<=T%3b$t64*fgsXqf!B|TCR?}lT%;DZZ?Fo-Afl&
z7z)$P3Q8G2FA61vC0vOSVH!Jlv-!fK|F{6OHFjd@e4Ut>n5_L-WuNYUMm}fx{ts^=
znZYk}Avz#rS8ly$_u0GO@EV`9U3UkumJM7`pZjp&>O~{M?FZFS#!kPzR#`&%+*#_3
z2_=W(P>wWdJ-tZd0|hl7q854F^G=v1Ejwn9;?>g*<4aDyr%F@Y6~Ey(l2>K$`ef~n
z1KlR^@^y*wtR%vHR-I^BS<PL2ydqdvp8<lgV;Z3-@*vTFg?1D)O@y1hWc?H+Kh?X}
zd?;8tH8<LtHCWABHa5HqqZ^@iGLw&wzt83Ps;Kn?wVU}*rjPa6VXW^ucbi1bQ6^-3
z&(&qzwvT(sy0D`ER)ICOpog}N5E)&P|18F`sDd0YeyF=CZiQfc*`4Zbqe2muC_TLU
zB!=gcHytCZF|ycd{M(LT?7&hdwATy$H`ghm?F|0blBloF$8$xhh^fct&RTEH&Xx=_
zdBy&|{yRdiDFZB&$5bxjP(LpRBjw57*VGDv@iTe%@&;?d`O`EksV)w$lX}#<`znS(
zIyPP5Q`Bh8Oc~{pv>4^|K0QUPJ8wRWEQ)oHC5p$%lEiC2d|cYmy*<%$&Mq2X*km>J
zD$LJIj0h2Qy=K*|l@~=4i`p1e{a|7oH&3bvO4g%?TP0NA#ywB(s=+^Mlx2AV`c=<W
z0YIL^DZR7dcw^(2bY4k~zNC95Z!IM%2qnrtC<6v8$-e}tZTiE=x?r;Ak_Tqy5{f);
z=rt!=OSwqP$2TOW*?s$()$fv4@J9M{)!VlX0;zjv=OohLHV(d&>6XP;@2F&=Y+E?_
z{W#@HpD#oFj~>6Mh^8kVeH^$DpBrmMEF<FqUO9pVdT9hT2{Kih^;DZ%ORcne7rsQ<
zuq;J4%Gd-`vskJn*?lL;i@+Xt4m0Ddr5MK3e@CQLu)U8XBSg%`T&#>=gHej)fE6Iu
zPY3krYwF#^Bi5KX$9%$qxufsyRO!En0|&t(sqBZI^eo=PFe|zuVNq~m$CYbwKlpg^
zmbfPjSIWYaeuOCL@^z>Ts8ic0NsngDZ`n@}7>fnVFc*<z^w((n<>?hlCm&sUdlaQh
zkgt5_A|K|*qwP8luW4&-L)jIjHr?0mWc71oR*57Z5uz%D2T_D4@!o-He0+*81IB*s
zqI`1tUmS2BCd^(W*uZ;b&1+XeL_L9#A~F_AQz%6`w3xfkM(5I91e>h&-?Wu2=8tr2
z!_O&+zm?{`+Ial1zo&U?!{Pif26tL%uv_g;<zO+z29?9*<`t9UKNnT+#Uqoo-USC@
z55Ig8kePZ-PG0T5ZCZWPK7P%!+islAmp<i&QK>GZn2o||I;uqWyCmnnbezGLFBGqM
z^C~xLO1FD7@vSu1Ooq*oWruE)I4^?nIe^I<95DCe5*z4UHSLeX%FeZfG9{1MgpE!-
zcl$!%>Ua4!Bi{HC6Gf_lq5CZVPKw@+zmdx?BgzFr?;F$Ukh^^LN>LF9@9+bEMjK!7
z%1zfj=g0<zx$Vf+ajfIp$c9*%?#cN;LdlH7O=C?oHJGw2Q;Rh_6d6bgb%>Uew0o<N
zxWp;zqqDz%BnJ~e`tOhasD~j54T%RnJ@~u0zlSJ4_bJsDI#x4UKc1|{XHl&N6*tRp
zFX*p4<on*bAJ~X9{u09LC%Z#)Zy$``&Wd~V%eI722m^j;>w_NaE}byQF*D0V#_h>j
zmCT_|>&hQH!yT7DC5L4DbWcqc5BZBjJ>wY0C{eb&{~wWaL!RZ9z;C)0$%a#aYVN$f
zy?rfOlE6bgS4XLXIb-{v;ywmxRsqa8Q=+w5C9wSP?+F8yg=){|^#5+!w=FzA{dthq
z)P+6^ihAcpZ2HV<=+Hhro`ke&q$<+*LQE1WGgj{I>Ta^3(Q)9wCPAlDTHfr!+}t~u
zP_t0j2@u&G$jtjevnptE6DpnjUa^pIQB~`D=cQ!b<@eTAGq;3A05eI<EGsKpsB;#~
zECEh{5y0^w(2G;O_&ZmB%@Hc?dJjv^0eA2255T$&bbFJUfH9JDFxq0v+FeAEPxSiq
zRZxcCN-$-{R}eM<(!nmdu+ogL1jg>kA~*tjH&R$k@E~*X!h^-Em#I>y!<>aEvG&WT
zj@UxyJ*ESNm7t|YNAf~tBg~Qd0oqL6qIyAoOee%)HteWvOTam=x$Yer7KSeI8XYLG
zC=oHWh%qsrSdH@OYj@tLe6_*i)Yg4#6Mj8J^B~xHvH5k$zx#B;fdTW+%G!C{iW^ac
zP}F(-wP;6F**WwJfn=s?2FkDf8P?56+7_kksYF?(Shq-r5<O}`_tvZ2h!fZ)6#3kl
z1*Vm(IJp$9tVPPLTGVNI?qD`+^z|K|15I_D3RPqTOJ1K#9v1EEzXI3d{%c&8SteZi
z#ruAV)@_!oC*OQHyQtkaN@>rz<g|b4Sl=$pj)zN9w{h&+%5T=k4z}RJo|m?p6g#p{
z`JevlLg?oz1bD6-(I$AyRBfwtYpRl*GVIfL(SE9x#5#{^B-R<Z#anOaD1Vr>!z2uM
z#CP`GNPjo|^<ayf>whKK;GGM4Kij;-zzy)Pa&$}&>o}?ZOZ)y+zkW5t%876r-OF4+
zRnRu+zbfZe(?(Ar$I?<H$q~a>F}iWcqOkjXOcVC+^SozJs@<6_l18^dqjPE)zLB`X
zAPplsMCQk;#Msw(v6>ZrKB*@jKSVr-am(o+DfP*`ez$4cf8Z|TMB+z#;sztyOC&C$
z{^T0x6i4@D;a8@*)i-0eKP!tf>us098=p9Bwaq$uCA5Du_g9eTKmXzED|m~ljqQG7
zQ+sJygN4%alByjJw<}tE*snQ<D4tAh!c@wr^f5eO)chyEB_Zpk=EeU?o0k<Pg5Te7
zD=IOm9;u3`H(56BJJ3fi=R{|IjI#gX$L{_z{7oBrNwgTpNxviXAL<J*I1v`?er)<B
znY#OotKt)-om{JTbI;klaOhx2{xS0KR-2m~;bc75k5}L7g--l=Uyxx8uP}d%v1i$X
zS0VeHWw1uBtCI~IhVTjwu6Dcp4i$&f8+JhDS9pIuaU=D%=0(o`0>J+zGP)j=<Vk4w
z&alcy-c=V=RL#Tc5Rd!z%Hzwt0d<2xyhd8r<t*R7OsB9fzl~)3|5K#owe{gw_|52x
zfr+gx258~V5)iRP-1FY^&?Z-}*Ir{$H%dAu@vZLKBx^e&hre&-ayE4C8Mmg+)j(dT
zAZgZiBx^m^*c_VmX<w%H8!sSS%pY^|n)4rL39TYY`)oIJXUC=9PwGWYDfB5&P9I=A
z!t#qLgc_-4Jg|KW*~%dIa0K~~Z*nL2o3VG&1HxWYJx`lR$-nuKlTtWP(?W=hvql>6
z<IdHb-ER{4lssAfA{lKZVcrxohPtaLy(p-Wdj_N^v3DZY@dIL%C;#|&k3~x2<<4ln
z>%bK`V8{Fw@LYH$vW!s}-?q5-LZ|*7la9U|!Enlc8um6t<0HW0N8k=#GK8}0uy!c=
z=#vBA({q2V%0xYiQISCf@mhbK)9VwVwyM<GIsM{4Z+)8aza6@FN~SNG;{Cb0#*}b~
zdaC>INN9S=FEl#N;@EscK-gr{$8sh18&2jr;;dh9<BS5O$xnaP5#Mj$yhaa(brX_T
zjcW`m`k@!BO`Tql-}RNfQj~JV{TLsQ)=`B$B6V$<ziOqix=!j7zDc~#bDJCS?@lQw
z`X9WJuKt{pav63Y%Y!YPRx*oE|MrT+Zr#Drxe{Rr1s3oAdtdjYlkLMZUk~emseC$K
zdaEvI!fEf{Ei@cXQ6E=dU~Cbc^XC1P(iQ=Aw%l7|WBVCz!bH8rcVPM}>c-S`%2~pj
z)c!M2+I{+3dW9?YRNpXbR14#)Fb_Z>B_(Cg9P^pC=^p)#dTi+Ch#J5*NTt^ObxGGX
zrYb1&p?&Ux0mLH4bo^*nua;~=WJJaZ_Rh|^<9$Au*NL;l*)S9)Iqi*BjIl8B)l;z4
zC$hUNgNZ9&CSs`-MET;PSElHT<Tl>t9t9z~hxcqFf3Cv~?k1~QF)wmV@nP;@1+7OP
zD<>a4iT!H;zCRktzA<|GqvIqyG?O_n2KhG)asff6TlM<`a`|8yI?r-BqIBMrk`y8F
zYj}96-Na$`U}E{&{5OHz&sRbKWXswymN`m9yL^;^l{DS8@M<D`^vL69f0u7WwOjVV
zP;BSASNDh7_Poq`yOr(Orpe9H-B2C}*+xEu=hyhy*w|M9f4dgeOh-G4Bziz1b6m~)
z!)G+#R0ma}x_^hqRB&%DMzJ^6USwleU#^$%{ra(CS7CNro>V;;u=9t>q~e~outtr}
zchqO}6Poy^U&YB(Y$(u0N4%Fnx&h-Y?6$A=fBg8-w5%rp_Y+1G3Ld0Yew(f2`tU{F
zjPC=$qv=FHeggo|JLb<vq@N8L7kwuqDe0>DVe9(8%cv3z8>%iVMTB=NJxoD@6!!zN
zYBY>E(;HUc+&Zuf6qc4@d^#^VRzgL2FA+yBmw`?u{q2>U`vkG;im4_cBQL+XZ>!VJ
zD)qUq;<5soV^P`UXVOuRjDb^qDimhOyt&w+b39mBgD#`htjGA4Sj4aqo0Als!E47U
z>zR={ifl?_(-K(i!klibi7VGEYJ8HV%)FPg`0(Mw-uA8`NBH;+!k=1p@J1l$KL!M>
zGbjVEn$l~xau1lrEx^}X1lIWH_O4SvjkmAU9s&kXuTzZwN%+<n_YG!Be!NyFX7c2$
z8&K2_HmIACii##*XAZp{mI}|fnSavL>A6>|OQCL~Aoi(Rxl`eJj>j=7fhiA?jyO6x
zjy;N2v~wPBFy8T&PFQ)>@!EYV8{!b^NpczYWfZ3{_LTRnuCrq4YP11a-Q~r(=SM13
z2HCu?I~TXhgq?OOKI}sZ!j0P==@CrK$7Vch&t&9}RS6#vWF^PuFpC#?b*tjsDwj*L
z;4jvHQS(ArnkoNjWp^HTnU5%}iT`uDQ7n$Ji|BxDs%NSj!+CgY+|SFLz5Xm@ZFvJ1
zGn^XU*pJRcZ(%Gvx*Sp|ns0a*IZyvYx7qGuaRQAzq&&HDKsuD)pmi@az0+5?L^X@s
z?J6pPV53G6p`HSYPGDF!@W^IgPP-8NGF|cP|HxV%*hyc}1J<t3b6iirXyLtQL!uu9
z`O4aJ%$`f&3RzA3nwaR@CuGWd2oCCmgtUEO1n7~GqEAy4k{e~$kixThDOqb{uC_S0
zX{qqGo@^O_`z+t(hsqR`&NbcLti8~`ETpxQhdP$hXjenq)tc89M<jfMD9*xk9bDnV
z=4zZ}wGqyZBW0$g#%@#ut4xzBEPd8sq~rX%H#IWuWE-LCbAuYFpXCl7W=k;qYYHEk
zF=T-xr>v~j_NkD*ZJ%}h%P}F{3&AWbEUMH4PFLVzf2A^rAIY<PU*K514shA|QZZd!
z-N=1W^M09So=sQmiN7`vQd*6A{C)W<C|+|hNMu|N1z?3C{$*L2qvvMJiaYDa?w+2i
z7tI8HX0i=XkU6kT=!R4fU+C8>)z7N~v9cq9fh$_F<vn#_zJ}&6oS}zZzi4lB86ya2
zt<eFk@c@VqLjF=(iU)%RMAIANmUVOU^Lll&m065g9v&W{LzZ>0wKMpb-njP#LMKa6
z+KA>K9HyTG?*UieX(7h?Z{yE^!qO$UO)KxXl_x4L-w&=g|3*e@l`EfacCnOLLc7cm
zQNzgm_bPlSYtCUZ?0p$*Q+U`Wvwb5(_gUtJ57c8AhJ$q}=wBv&2O9g~H-s@wk~<W(
z*UoJAv!;z{{zYdM-g=)Z=UkMTVmpI0GdtWiJ{7cmaaM({r#FuVBNzHiW$-&JW=p=C
z8m)j~*&X=qZ#FmIYjb|L(cxkLN0)~;2Ln7D@=t$NQ4x+Wl;F#58UBB_tu&Cez6(5g
zGw$3pdY!kzMU^{2tC(StWhz7ETQx%^pQa9<!0Syjam0(?pto}Dd1DX8p0mt<cCd~5
zxUrwu!qAcx`4xQ{_z^F+!Kn6=|K*X{lq=&``sV0kMaYJ0#<<xj{N2W%6~N7`cKce~
z&;0%Scm2&_=-+f&Rsyk2E2s)6a<2gUD8&}$okwr-3hSbw6U3!uR?nj__#65SG~g__
z!>@)zX*{6!gGTS_Lzv9$n=s^`_2$Ecl;3G{)uOq#H(355cypi5xrN)1^USfig|iP9
zP()P-SWEm8qM*m!G2b&RdQ|806vIo}edw>wMjkwfZMwU@_Sfa?3nZPs{{L%*V4%#k
z+sj#>_ip>l7x~r*e4aZa3vmIa*iJt-7G+^&b?oZyUX6s|>|D!L8x)nnad^rFlb>m>
zPgN%pN0gMO1Dg=*zTb@{{E79~7Lh5%zKz(<D)tWBU*m1q^@VFp7A>9>*gh>yh9;6F
z27e+cEjswSF$Qw#<&FEm|6;P9q?_dgG>^#UdKBu7R1FG^7*oth9yK2<h0aSozcso%
zb7RCSLG4{B<);_l(suWAb^FcmO`m__CHY9X2TC_HwEBzdync#PuLlPQ&&nk3NjP!s
z-Ohi`ho_U3JY`O*66$1$++{I8@5&`N`Z89RLm!<0GB}c7<Nz}rn}2U!=<F7m?R04%
z(R!@Ad{!F-5}l!<Z`~hNvujXOF(LIM<LaA}^NyUW*JTugHrF$(lp1^f{CR_x*$Fk;
zDi+?Qe$(4IwDA2uj9j=WlU`QheTr018BvU4E7C|5uqxB{sA)B8)j7{8tE%|tWPk_D
z@!-O<f80?S*i^+ZC5%bl!;|X#Cvt*I=%@pb&M?H7Pw>{7M{)HZ-X2*mzZI*X(F1uU
z)b<g`e~D2sTT>b$>aUpnvFC8vky+{P*I(wcGtCB<g<?;j{{&W0p2++BSu;dQeYrfM
zY0uhm7>v64vz9-Yrq>A=VDRTFXEOeDtBT|d|JL4@lsudD^=aJ@L5+^()7gb=C~G)J
zH7VIWf>0E`&SvrK)*BgpFICs8eiJkS?)_n*!hg*9GuFA*hBEu<a+~6>Dd897-%6=Y
z{_izIh2~Zd_hk(o;O5af{Xd`WvBn?gKN@4SM{+A$mW2Dz6|rGWA<UN|63lwq`8<(L
z$NyU*lU{Xd8Iij~{Zaw9UtNKrl&aa4$gSn5|NWZBkB^-H-tz&MMq(}FIY#ZUKcV&@
z&z~@cCsktCIsgx*Lf9W5i!jAcyGTr4AI@4ICYi|8R<Dm%SM6LlITx0?-E!cdm)6u-
z>W_7|FM8;glmS{}SURz+>Ew0E*@&&Db$wYvY(YfGujlQsF)9?Zlu)8v5CwOGpv1Jk
z^YSa`-{EbyD<}=!0frAD=Me8t|1uEH;<7XhUeiA!EGRTL7XYs)m9Vv_o`Z)aU`U2~
z-)-m1H+CZ$!RyS_o0OUe?3yB8yO`~Sq4~vvQe=Pkhwcw;;Sp3Wf`K7!f;CfXWK`;-
z%I5l5Y^_g5s!pHw5jFB5s*RpI`za0C=8;%N$4jdhBGuWu@2+e+eeyBz1gaU=6Y1WE
zs+fZW{B}<1bXPz76QR#W|9!NaUe<Q}jC6wv#{4bW{2Q4m$)tZ3HMM+&CgUjs6D>yf
zJFYE<JYZcsT+D8L`S*P<-qd=Xt&FWcb@)`yxxIVO>AoBk%CIqWl4Qpa40s5UF{Ak{
z3gnXa7y3joH}-01I7c=q9Z=mI+v4eR_0sB-{n&T-(nIKr0aZ~IMFDdu<7@DsO}tw#
zPD^c^Im$wld<%EUb{3a2zsE+1;vr0k%((KL5+L#caESV#E&5D-YMw~P+(lJFUto4s
z50hwSji$un&~h!W7QGU+ImPn3V`S5Q1@p~gprz*aMtZkHEY3nJlXoiEY4)kxC5%Ft
zgfc<)p6V2#FLZ0Q7Esj;>wiFk?&^eRG(~5g1PhZtmMuZ8!=Ctr5_bSQO1K*qh}XD?
zU1R%-tZw@k_+UE?cqr0Jgl^e3iOh#72B$S}gas?#bJfk~BqUWEw~nWl*Uqr~FkF~E
zFe+PEfaq1twHSzD*qh-rHcsW}+o=?~)VOgV)kJXxIu_T$&gT6##RPNl5ZYV3k<a%U
zf9kPd4tEeDOy0Z!zus}$XNwOm>H3ll{9i0^GbWKVpyY<)qOjp%q3a+s@5@*`hI3r$
z%8gO6xUxr1HLK0VV_@iD)cz@*$<OBemj(}WCN$h~)_k(D)kd?qOx`Sb<MJj9t7<BY
zjn>ucP>IMmMVgV@GzIXLtNWj2Ll3;jmHsn`XGcz1nQ!Tkznn~$N6d6VscAW8;{9Q>
z`-f<6b5+PYA5>RwK)9v&6pLXcPOenLlQb7%yEzbSlQa-#t}P_9+u~Z(XihP<_CpJl
zmjSQa=3*e*xLEkkuHsQ9!JJ<qPjCFGD+V|*1p(jNudHt8qov@5>1l|?q$ajW5~q+k
zq^ptKC{~F5^?YQp(X$-$ZL{WTeXj0&M&nOfVQ{~{0|7W^3txDmBQjjJ-o_Q>%%Ss~
zm41)b+Pegc*%q@8n|%jnSht|XB;`rlXcdq3i|eFw)*F>^%N&Xhn6;8^DOZnL#;6ED
zj2hY^SQ7A@jw>h~62wM4`c)RAQfMJ2U|?RM!qPI^Zi~xZO|!$5wE0=Lsl{D2WyOu3
z9ogv2#ZRt{ij@g!j8>2d?Oi1(Xe+=KmOeTBkAuzH>?MlbE{h8>w$`sFtf;T1<02pJ
z3&+Lkd~T*6d|JJ!sRO_O;K3M`<i@qD!gQYj&+@7(F)LRTlS|sW_P69bjavzTe?zaT
zPFXWS0;nYrmM)Zg#{~w$kD-{?(WbT;<aJ>bp%txL`%hhOC1ZqQ_C@VaNpFPW@kOl}
z=onC(&m<s_6TzB}XV55)z`>e1NSm_^ogas25;NzzC^8oV^xlfR<u1&5$H$VMGF$TX
zf6m$XmW8e#u0;fb7b<$^fWRk$S1b-Zy*bqU$`-O2ks+?gy3S7NYgr#_f#w)j#tLFZ
zm<EHll0KxkT8WqcSost5K)tCfDl-x2j}VS^tx$#-su?m5S-*jK4+j~xwxY1T+LT*&
z)3@ovYKup50%Cf;MIL-vGISV8P)l9s+w_4~t`#xgNff2iyKw9CP2qbbEiTgh$KHOw
znHO77zkSQ&K7Mwd{c#Q+!(4^|^s82xJJ$a#v3MDSKCSZK$29WC&k}P|w}bojL{0Tl
zmNix4WHm1Q(lfbe_)8)QJo*W2Aist-R&`G?_8%;?0)As%6`$hiJ8(Lq!FfnoF+nY9
z%jU+8+f&SfWa*5(S^YNX$9e{Rhr6%E#jBQvHcG0Ru5ZCWlCs`PQ<Fyo_yj~0=z7u<
z_X{7O2g`ermwmi~iO;CAVT}Z_c=W}8F8>kFhousm;@jQ#HgCN2!fi*yJ@qC9J(!=x
zpb-NUt^G0O0)F3`Zta!aN{H%*A0U_=l?;pPz+)c=Ue<f6MwlF4!*t~RZq>jlu2tq@
zlnqDF757>mo-Gbm!spw}ufir+#r+<z>2yJ>dzV(+$%|D1+uVo0E*QG4u=dqaU7R#5
z<T&LebN^Z<h!M1+M*sc9Q1C`UzkF)LW3s1~PI|u<7TT+uc=6Mr{ZE*MC)hlP9_)C8
z`SS2~(#iR*lL0C8*z_CIRI{Tz%x@X#%uK)WIuDG-QOs}XT5&HvI3(U`Kl`X#{@=4a
zN1667oGG_V`($z_b!~k;Mc126t^CDrmfn`x-&ZOY0v7B7TeuHT7R)IkuE{TknLpmH
z@l4Xcq3#*+Qd7}<>qK=O97S~4`q+`W>?!7%Dx?ePX6&;>s!>y)4hc)NR&coDH(4J%
zJZ_mcYkOyrO!iO(7Q7yZcCKh~EY09IxkgBIE2y$>W<j_-Mvrp87Eyk)SxVmg($?Af
zpswix1LyG#^y|19{^iIKu9b|Z8c0E5US8_QT`wh7ACPc+6Hd6_AlxLqYA0kojC`vV
zr}mhHGmz3wpq!ApvyILo!z;ja6mcqys8mUye6-_th>u*>=1^moxhi&Oyws7)sgN|n
zL_dVsh%mjm)sx$gD{Ci+PGF90-N*m+r+wzF>dQ=e{C^C}BQZ;0z>ABDuD&FN?4k-I
z$si3-T3BbW)0DATpQDlH!&1ygOs>^8WB$n6w;yRkWEfN8YChFehYyLEe=RjabawP5
z_b&yTKYeDq&54b=o7hp)>xlZ-;34{*JMvF0c8vM3<9}QL6fpp)cu>smh8#Jr*&J4U
z{=DDos$yN_%dKiN2K+vvR=@fo#BV1x5fN7ssul0o)fevhb}LEZG!yRcTNr-;T`jRl
zI}q2$mZ9yj4TQ7BJG?twRa41sx`zT1?T}=~%SL<hxKU^9vSp4er3$@!eB<saS^8PW
zt<a=9UQb|-*wgCYOdOH7bvRDO<?WL)ztH~-xy^Kv|A>{Q-dzk`8qs+4?^{h4YQjD#
z^{X|9cvi7|iJw+HPp#`Q^>HGuw>tvA5m;FNY>HOSdO<ZtYUA44S}EkWwg|XYDWrEg
zc;l&?_F>`v!${WgJ^s5T8R|O6OiiOBZ?@`n2z=w`6HBa_xAOCneEg+;bZ4TB7Si%a
zIUg&$p1>sJz5OWn?g9;rNz--I<=R~-4R8M2VA{K_bTQ~mc^M7l2znjb(Xcw=@qk{&
zETd*%DRdpsImz0ajX7&>J(d2Kd~tRZe1G$1b4Gc0`9kH8?r5{lY*Y0;ZQqTx1+(`T
zRV$LUH7>os`>-fvF25y7yUe+kWr&nd4lrA)KsIY<xiY3{mOf=UfbO|Wf?9+%Tz;lO
zc0mjj*Dv|1fnZ*`Pw+@y?nvW(?SX%fw}<@T%knGjE|)e=`qB$GRFNbeIN#}->+R!{
zSxB-X9YB<?e3B9kntF53{YoWG&H3S1nHx2#DW>J8r-S&9yz6*)2iha(prZKAPmD+F
z3Me^c>Hd0f(?R}aUSZ*f^PUaTX35bK;^Ijl5_n-@w5U2{*d|!z?_qm^?c$<lo#J{r
z(u;CRKJSwnQ=>7z6O>hg#kh2TpXua!4W?^1hqiEa8>ptlNvuzWUup>+vB;_!&dqX|
z-wZC^Q5d;&xCI>ioI9acn|whO2^uzbgl2QOz~+Tu9nem>)FKwTI=ypplRi!PnPmSy
zv!d_Dwga@+fDB~El|F@u0ZoS{`C)>>lhB=3g|;uRb%sKD_8R@o>-ny20_U%rclueq
ztGT_+YhTl0p`CnxZHM;mIs((}WB-cm;NmhqDEtm^N+U&Wzi1dww=r?$bI!f{hRX1!
zsvn$r@mnFsEUT&oCR!{g`nm(Q7Q}u#fXl<&%1YgBeylSb-YF>b??3|NwyUc`nv0do
zU?i@baW1b-FcR0wfB4YZv)l{DrG*}Q$Tny{Rz8%;+LWnV@vyu6yiey(TcqW#wub-A
zz;~CH%wVVCXUPr$c@r(6>$4_S1_TzPCrl06=86l0$%g3PFt8)UWoDq^_&coj_zR0&
zRK}f(dp74NKis4lrvyLy-$Ua4T`P``IdA6yR&Gov)-dZ?yWw<u7sm@XQIr!q$2(nS
zqNSCUGc88QK6~=h!V&R}deq%rh>(ZF0~<WQB=6I!O!TkMp`u(`@1j9|be5mji50u8
zQeW%lO>nxGS)r}Jd!OV6Khv8T=x==9yE|#mw4aV@dq9w6*E7wy*$7fq_3PKKuR=Rd
zZU3`cOK+`|?ys3TaWy_<*N0~wSo{+n-rCV&E#G6iDkdGOsCy{Q^R}T;)jWrj7*XTQ
z(^~WHH_~Sx2A790^1qB5Kkh7$a(_DhzGKE6%BSgMhy2Bb4uo=PImjOW=aF0Ej$%4q
zA4P8`CZj~zc3&!deWBn;%WU%*tIr=lPDTC*e=vPQCNbgKp%o=$kimw%&k#5MkLl|$
z!f$JBW#y@96B#c)nW^|*d`4E>y9@fp(5<APUUil{8q!3_h_)&fr+xq({0a~`aWbBB
zjalL*dSp)(y1GI>QJJ5Qi@sLEhCnOV3hU7+hu#H`tP>dM%lnN{^xa8xv06){Yl3dT
zWaqJ3DI{~~J7?cQiOYxOL1=Q%G0JWYL^9k9*VbN&i@a`dN16?t7ykWj{`G@YbdBnR
z>PP=7YwGB5@+Wnp%}<E)hoWw~*>z-3sd;jP_Dv~CYGNn<9iF-VEKs^*BUXn`OR#`Z
z7jhOkx@TIM&1rJV&+25>+P<;OPdl7cIKI`cZzF~tb58&2is;;`IY+Um!A*o8JNoJ5
zz9vyy0h)twTxhh@jjb2#6hlB^nENi{#)&F}bi|!m`J3I!x7G`_`$t<)6gP&GoGP?#
zc}b(ESGkB%#2ugqxCqJ1z3+P-Hl^ru^y%=V7LfE`w);5UeQO|@@HE~q+bn-1UM;oh
zNp`71TThQeTSv!&Ah7}h^3L=-17OA}yV=sL=F0Bhzi(_b3Z;`J(z1l4=R?wI<|ze)
z%MSO!Z^i5U!fQ#-14P+P*(4y-KRo58+oFBKYI0XXBT``A)`)xoVtnYS+|Ts&2hgxI
ze_eDa777!k*wQ(j$-Sgy5oF_9;fI)#Z7Q@?acV8{3DIMR;KPd+5WfqW1torHHoKyl
zV4|7;_Vq3U=-XU@y{uJ7m}mK@0v*`j&gOlocvKGpPw{Y|*R+Xhh>nt^<9u+nmwkNn
z`zYk{kv*tlmLtQb<<0r;?6x>_BIq|NWqY3#vch1qtxsd0lrRJGY|3t_*iD^y-h7{Y
z#jbSy<Bv4AVGT9O-iX8Y6*)J0jOI(zEff2K5r2N+6i+8G=iR4MhWxKs1%I$xDQ~?1
zWs8SX<LAYxKc=O$YPi~bQC~akDjq`nWiGU$a1k%Hcg;4w+w87}K?Tq%F@cSvS9TCG
z6nT~w@JY2x$Tgsu4`(afQ=!SDl&!6*)e@w_RR(hry&Ziykuq8_Dvk)AhxijY&4Mp1
zohO-hDX};emTEAFOa|T~iGyU&ghYO|vOivg_|0J8vQ_XGyKje1g*q}8n@xLj@*Rzj
ztoIG--FRcOcDv!OJ#vtN7}2e)qZ4EfLNc8SM^K<7U<5U@_UtKtXUs`XY||Sgq~q&X
z&5{(1GHq3qn%UmfNf#Vak-T6qkV_IqG+B@M9zR}tCHu8wdRiK13|i7RZ=^`H4hBnZ
z?4<6M{4}@{vSG8v+o~O|=EJucXLWJvp!y#B=~NMv-tJ}tkb)PHoKfJh9nfsnM``9j
zQkD!jeztpAE-9SR?-@nv()5paq1J7aD9vYQO}Lptt=cZbJX6>Ywn)|Z%DQ@cS5da|
zy``JLMYcW{<pIhj{8V-%L`HI`f*i<9o4(x*BWy&@q|C2{BHh;x1s+#Tjfvr5c`$WU
zoO%g44OC$$<sF5vrI(e%*zv%v@u)j?KETlo<xdC~t<$=w6HFSBhV5QD{{nP_#X0**
zu)oW;oqB1sosTctYVPCoZ&}HWZGEB8*C+#_s<Bl(j32biG?B@*j7n9#^gh?z&*M8o
z-%C78tIgUN2LgkWe$ZC%qC-OnZ4QGT$R`CvPOdnXwmX12qzh{C<K}3k@2x)`wf^!)
zLNXdIRzWw)DRdl@@^P!sLzy9wzwX~;-{clN8pj;C=LyqMLs3q}(9if{koIO@K_Cm#
z1T7dd^_mEL#tpdTkb>-NdMj){1|qiZlxIYJFv6)yyBjiFg}Oc5atWB!5htQaWLA6}
zsmj6Ay)@!Q-pJ7A+msh5BB`BuDmNo!<tFZN2aCt!4--X%MMU)XOL`$G(=u8zW|^4q
z8qWxI1VkaqU*_h@j{5iHLBmjoC0m6M^%9&@HDj8dI@^uuUBTZa`sdT3Z&L7Od)1Fu
z6A92vnvcQqENKxf&UfhM41Be;{F?LJ5{DbDLgry8R;Bx<k8)O(A(PZQug%Yx9~(tp
z*x*_W_n;2C+`$eur7oZr1EKs#@{?z~L5NkmzMGgE6>1ADrH->pLHQ9Gd@|MV&$cZ@
z)zdV4!7oiA8Dp;|G6?gg=t;!nPZ;f12MSwv!RW2bG%B$oF?7w2@VdQ{GdxH&GsoOc
z&|o}QP7XYHCMC7l^XGd<XvH*jgzJBPw|^ZF@lc?<A)?m%8xl@Pjesue((?!mXWyyS
zy>FC_tuz$hDAUI7pOEs|AFFWEQ=Ts&8k)`ArZXm2xESkU+KomV=`s%Y4dyNVJkz@n
z>RR}k7sL=u;v*=MGw)8F<ZPK8(P_~pm6Xi#;Kr@b=2`Zwtk;qbvl2BjrHp_%X2)Z@
zGZJm6Sg}+>3OST;;tN05<FosrHWm+^r6NeTSghfKQk(>}2fB&5ZnfwQ`u9jfyXdQ9
zWxk{l|0fHU_Vz2*NF9Y7c=5_{Y?D!NZz$VshHR0Ryb%>7>Yetx*~(dCK1Bm~ogPR^
z<%M@q-=ph-dMB;hN|20>uC7;~Bi8jm;sy&eJv%^kptZ$eN1tcRodIQ<=qe1wHuDu7
zI>@8-I0eIqkBPV@uR8L|y<1SNtXmyDa2oLlP=qDSN5AE|L6sBWqs=`X*^mn^EG%td
zSoA;&@BN7QbwB*?M(>TZ$UW`$qMZzyIMbgI7>W5$FoG^9jSn|KPcq?;KVw1bh;zYM
zx*)xc)&zlBXn5vNx*^&Do~#j{N;8Bhm|h>!)Fq~ZH!9%YLj9_OJdx3$(z$Ph?LjUK
zu(>iYXM9Igi4zprW^{_k1HUiwla@UIW+7X&ncWo{x{nwddp@?D>rNoR({e2Zg>;{c
z+s&qE*#>0hz*7zVb+ymjP4K_x=QAIX^uf)609svC#m#})B|54A^XmGri1VAe+lUJc
z94j1~*~7EMLZrv4h$I?0Kl?aja53N+Py_UlM%ay(xN!~e#_=es_jg@n4BCSyjGTS%
z%URCJTLl|`Pk=7=WUILjkWCDiKFNA(GBL#J`|HQ$$S;^~tM4-4U6Ur&d*r87-3Blh
z6AvFgd~UHyzxpwC$jEZex1U0nykqE`99{-prM<fnC`0~^`*q1A6`o-^0Y(H)*<nNt
zIuuLy`Nv1pKP8}lH`g+4=69_=@~eG;$)qH?PlCX0{GOZA%~pMd6?uF%?+7{x*=OI~
zYPaIbe@FCB7Pji7bGK#4mGR?d;BjtA{{&hb<iEN=jH#QwF96~&GJb7`<R@OHqi8D$
z0(ep^jJE$;os;82xMI)aznksGDLRHsGDJ6fxVvYn|M<;_3_fTYi<?T{;6uHD7D{wE
zP5@Mtw9y06k~7t`6p4UL-{+(rq;KQYHUi7hsolimjZJsI?x1?T8~6uc0#L9uJ9i*v
zcx63W@)JYfEG-3EoIlV~<j_)|dV4o2E>yt0Xjh~_Oi*y~#Fa`?GgC1%dvn37y&XXT
zT=q27yc&NlHxA7@@+4#j1O|HSZ~n@b>|PMMvC_LS^vRgfkNaM)nl;0s43YVXn%6Nd
z%ZP9A>Cz3c@wT|+<cbi=cwsKdQG+Ibi1q?boH~cStD-zCxpA^$@NqN-kx-*`LzOoO
z+w_FgMIv45&4RUmd)^Xa`iPfm5W;FhBjt;TN|pyARA<t16$Hj-pP7RFhntaLR_3Gn
z7L;0tOB9DbB$@LJ^Cskx0&+@ynl9X>rJMxG&Du@7>vMiVPV$am`7HE+_*<=|MLb3@
z9))Z1(-><jSju2Y&(^7_sUpw`==hMOU+sMP@~@{4A3ppE?UE2BNlR*5cg@wzR(=oa
zbxZ{J=2Fth2P(5-lDrK+J(XCG;*Wq)vdDfHtKM8ZZcbYEMkZ0k7x8#yZy1Klsk7L3
zb9Q!VbrLS0wV*ReurdXF#g@0DfYgEV3o00(MYxt|5JMU91ra(ajXxD33<;3I)J)$f
zoo~ep+_5ewbv!_WTloM%M$4^=SWyi-&+MlVS#5x8<0IA@3>1-GNf4Kyp$K-v-3S<~
zZCzCFQZhsuse|)G$kxf-ymLPH;VvkKPKXN%Ufz{3X^&elIN5Ix9EW}U_3DDs<>m(<
zFSCWQ*m<N7Zt5CB-kzKhLjw850LUr!VIKXKts{^@9mwx`fu>}@(BG(BI}z-B9*Lyd
z&sxd{U=Ws#8jDN_3JbsT0#-v2M6fN<Xi1~xs?Z&XRANIOBO{}`&E;7%5_mg#Rfu>=
z($lD%YX!P`L8NR=fyjuyxS!ZY3UH9$m7a7umo;iUM#UP~qr`2r#O{5O>5oYr@N}WU
zFwZU+c`G=J<$vjIWwmzRMqt1ZL`FP>u_B!w3>5Aor4ilvq)Cf5+|1IdFRVOkop#PN
zlxXru`Ygq2M16)72B%Ak$txxVA0vp1;ArVb)MM$U?v{{822tpmLdXi@w~LyOT!q09
zF5xYlh*L~CJGEe1O5;?-=qm{Z<?NR)=bNt|ey!KNUM4~fM+$aKPTSNG;>KW&w(s;A
zJ}Jek-g^1^9wHMFa6UxY2<@^`)oTG&jTgwGG~%n%utW0Ehc{bhX5KVLWs-FYqS9=V
z5t4F1QqF4b3qJ2;Ck5s>EEq~A?A-Sy@hjJDbi>{_dJq``o1Ra!{+OsvNzX?S8Ba-p
z41R~?cfB$}C%Kk%&Os`$0RdEq1G&~uNBsUFx2`Wdjdlmpw0$QAU}ko|a)OsIN25@C
z-qrXBdB&-9&xPa@QdeOfp-MmRT{F~I0V)emxt!>vSk+`v%|82IAW*uz-y&ywvJCzJ
zyK7^mQjlUUvWO6Mh${vIoL1-0oj<?!<M;6AEb`s%6Hm@IZzhuQYC?N5hsGuJl3n=0
z!c()5?md#e5%u7dk9V?5X&C<W?{Oy`wzo3lPxy>N8w@P{m}P=nX?O-ajd6^H1s_*k
zz+n#d2i18aGe2-oN$TK$B0bkmPnB1mgk+W+y-Q*F8=4N@GJ4Bfia}l&f|UJUSlGBc
z@+Hp*WKqAv0Vt~DDPyD3L21`9urG|YXHv8IVSAUo%KPw8{yPU282oIldbtzI5ubBy
zOCUF3UQXOxGYDW65C;$$a{PL0rD8kv8c+GIBgd?U7#5RC=W|qkG#_&|tXtaGBy%F7
z{zKJdXzp2)t)#Myk<w<j;>aiDs#$yC9$ixwZi9}NLiMFWN3n&Gu{-!mMTEPwJyR=T
z9mw%7$m9uiMBu@{Iwz2?>_<vU=kJ{&&{9eaF5ZKH>Gb1j<(I5fM$+=rh>Vz;39VdE
z8`F>3HRRt@h|G`zDk3rpND?UgxhJV2{a#Bv#FOn^pJh}HTV~IamJ<>hpGN%pu5kFj
z3`AZaVGyKVM^W9{E~9rek?}mskqgM{{-S65h}uY{Qes1pWU7fFutK<s+w~mC1UMSo
zw{P#3l+>PESO^VZz=~)!8(Vs(mbPa(b?<;Qt`)rU6Qt^(9TuWC%Q+tenioS4wjTBI
z@mX&WoF$Rk0haZ6P+U@fS<1w{(ie}PQqXRWc>)nLpcZ08#)47-#>I6TXoR?*O%li(
zVdooHdLep52X2M=_D@rKoJCu7VX5nbV);u4d1W*>K`SeJSD09X_xDDtxx1GSz8_%W
zdCl(o2g>hKk+*K+gEhVOh=6dKN%01{AL|yyHa!MeL9VhS1-bjDStXwrR99CQ!slzh
z2|<wlq}-+AU3C;F6f)YJcdzL3gwU4-wF<q3{xN#dPk61`ejepM7+<DKtYBqjO;3X1
z<TRPzto(OO=_5rT{h&))zJ$o$YH_Klk=`$HS|-s#mKW4GKnSk~PA+3zE5=};2S!e_
z`Em$r)DgsE4YNzib#6x(Uf6;)>hNVG2h~<4u=eJN<K)Cdh_fUCpdCGgm+8%(BM-Ur
zhWuGAFO%E5L^RaPQ3xPDY*Y^v`p}-Ba*{g+N}=tCUY|WC3trn6=RwVp);KgTD=D%P
zT}N--ya7-mGA*aIccmZNTaL0i&$%05C#+sNKoGgW<#J9>>qdxaX=$uTC1_K~-tGGZ
z-iL(>1H(|8x;Q%$#V#ekTd4~5DjzTer;F(bE>dwhdz*4FQ3P$txrJSR<H0sfgC^^j
zy~XYBp!d=d$LpJ-GFYja`^63t(Wm7qGe<xu?M&Mse5ymsY$c#xcS-B5vz6C33lAM1
zEI+olzr>Z(VLzFGE~uiJkn{vEZrlu^&Q`WSxw8$hoFek5L+ppATCOpv=8AbBDI}4h
z7rb2VoL&xE@hwJ@33b`6zdn;kGpq8}-qELR%5C|K#Js;qz)#2E^@HO@uqL@GIq>lk
zFCd$d=nP?~XgE>snsz!j99DwW@LQVq-Hr=Bzydtb;XMQ%K^C0+0j!DjlZ(s<g<N4Y
zU8&-7&ckCeQF2BAQF=uVJbK{X57%9i!UF5Xo0RRY(}LJI*5Z$Zj4Bn8YOM)nC?P5L
z-`|Jnhz-p26?rek0FrNFnL}bvNZUl~u80W>cOAE~2lnA1t%m+rWC$kYfF38}@BZTc
zLD1cd`|#5<R9-_ZTX5*rM4jba5Kk&|pC8W>X~@4<&esFb2ytWrSOi{18Fwxp3|Irv
z=Kw4$C^KG-_!;0RDP%v-lCzDn_c1sTpotnnIFzm%I#3SuO?kT)zuGRU3bAkV>lxQ<
zzJmCOh>Q1xJ__DBSt~euAERjfl195+1mRL8K!$<kZ;x-rfwYvH1_Q})^o%`{6(~&>
zV&UBZ?u1E(soGhM=z5Lil@%Q7MRdJ^^A0}uxFcxn-wm$KEweoYPzza*XiYNRZE5L^
zB7z`b-GQ-18S*I2V8EKRob^<>_PCW2GW8mWWP$nl`DO?VO%^TQ4@gXmoFwGw`I9Sa
ze1N|k<(6l`+riz;-h<ks*RzPA)`c}$EALE_qD7m9Y#P9J9=<%~d{hRU+d{-=J1eVu
z{0Z65I}4bh5xe3boDwM9^SUPoM2^0yiz%xHtqluD$7&lZ$I;2j>gQ)GU~i%jp5BG(
z&#kS`nikeB$|RQe3YKK4xX}Do7=f<fi!{NTOZzw>4Ncw;im|4+1NS%%9MtBb-hi9G
zBsp`G$CmMd_|7|`u$>*vZQ3uOOdu~C+N=%X#*5Gcnn)pebiENxJBgk={WlwDt+|9M
zkxA)B5D<(lF_O}t%{e=uyAYtYRu}pa+J?{rrO<r!8d*iCxe(Hb@@O4O|FGWW{oO47
z<*L&yYxHCPNA}5w<7U(7nprQ}6~zkglk^HnMIu-|DammwD=P^QT$r_-i>9S)9HeDv
zW&Y%ZkkTWfUZb!Sc2me^K-GHrprT|T>T4)SY9~ey!Dd~P^n4N#gsWWGtPx?jnaaQm
zWc7UhoS6Sbk?n>v@)@X+cX@IZS@FYuE8EsK1LwC;`KdgVu-htF_%=h%&g7zqdN0y)
znu^Omq;)_mK)F3dPLlP}zXO^j*Y|Ix7q4K4sS@pw3ct*bGdcn?`HE~=vaST~)~Acd
zZ)_s_a-g_12T#yzVP%G%fb~fR=c0Egj*<rXt4H2Q9<u&Rv-uJrS8|5YkW{V2(RIdF
z*3C40yCLg846j@SyDo-+qq!9f6#@wsWyC42bbcoS(jJ|Sgp3^!NCbFGuK8v99sx0Y
zW4`a^*-bX#IxQ72LE;;%LMS5y7JA5M^uREnN@>j7c5QEgG~uY#Y`R5(=lbdrI6|_K
z)g&NlH}ShnKoF@vj^6ED_`UEeNHf*Jh_L4!P|dN9PkJl(H*>K@{HU7{DrNMQ7kYpN
zHy-k2JXR$UdZ<&Jj>5^b4gBgHz9UZcIMxzCK`TgN>V>fDJE&FdT`jlh<-3R|AoqI8
z#%9Dq{5=9TZX9vg;XBZ9Pc-y3xHlg!j?Pco_FxQ}T}p}`OwIlZbJNWnKfhT60L`Gg
z5g9jXwv&3J>VH;E>c1$gd#R<lRYj45i7@)-+EOA`jk%eWG~>L+H0f1^r0~f4I+PAK
z)0>QE7XaftW90$pxLUgQ1m2f#XTo{ZwSx3Lk}ISUAy|k`yOmQQ(D6>t%bJ_11z02M
znL}$2F!y5<bXxKu*~=kuL5MnZz9ZIS(B>gYqz-rl4$Mwvpc!&^azz&A=T7){co&I^
zi-)KtH|n~^X!*Xt#pPU-sbi<5q(k$gFT?IQdXqpxI8R6_=xhmr{l}pPl;E#;AuMnD
zCr3Ds=mAkA3(~$32<4cOI)X5wV{xtH)2ErE*rqk4N10Z@38Pn*-sPwaI$@h;t{tj)
z2sG;!L9d=1{zVLk5GmIz#6;Y!g66m>E!_B?pGm^^ng&uFtgT&#TC#c(1Oj_KVt7jT
z15YrnUj$2h52e56k=D<C`g}8b1;SA#cY-%yt1BR|+H|JPF@NMeoL|CP^_9SPTg6;^
z8;BQSh*1&5hryc-6$iw#Hsf=UmI?;+LcEY47vm2qUP<6Idwx6vwijtAD|EdWid$Ac
zFJL;>?nZP+)r2>d^v0!%f)<~|_8WbQi2m_@<L7%(-5h<!0F>O?%~q8|Vnh_p+x8j^
z7$KV??qjVkmw?TF#>#PWa#BARra*95=^9z30|M8Q6W2CdX<;?TVzp+*qs2t+EiBF5
z-24|11wfB3`S~p}@hmy=EZvw~(f|5o{gjL67_-w6^tt>I6GQ`^V-!4QEpmESf-KCm
za73@0qPv27tNksOw*eVSl*-NV?Rm_(NErm-a6gGYYAHUc8Q0O;N+=rg_a@n4+Ml7(
zmBqC8T!p0q)~{T)(y}_}|7rXg0tG{ThQ4Y<{>rs$I$jnoUbob2#o0zHD>?ffudlD~
zrKP<&tC~pPhu}~JU?L=neuw1x76Dm!0m>yEJmK+!+v8eyx?0%cxX=yTi>8JM*`+>?
z*Lu!%ecQf8?{IU5+|(Z#(QNK{Qir>G^)yWVrm56~8w{9kPIs7v?C6K5eG2fYtRw5o
zP(hAXEtYM}9(5*Q_2YI*dNL5tLo4YE7!I<IP(MLxo=a*bhz<qz`@Dp~)X(Us6Nl0Z
zZK453xM`{01C2Qj_V)QCVb<ak=))SWGU1aeWl{CA1_NwJXym&KLkzd2#Sul$8+~e<
zc}{vn6JxQBRnff8<ZODEV}nWD5P}(9SvN-pQJUHSFLCr~w0GUBnq1fnw;2~xNb8#j
zL9gjB#Pq)b7QJaX=gDJxr}1ac=7KAMG^4Val|(a$^A!Uq4c4-@CU*#Qnlo&M(_kHp
z7zLlVJ9K>izV^Iy{*aZfII#j?VD_Do3L|4<K~fmn!IS75hReb&GZ}YMMY$Mkt`+Lu
zYKZma0ys5R<#^^$u>afX#u$~+QyUe-@Eh;#<l60EHHr)MzK++UGewAIc_W2_32&jW
zbqEOrfQ)^oLQm%3+UdEzR}*uRGhy&z{RpJsMbo*iAiZj%;ikB={V_LN0N=|7J+4(>
zC#=Fh$uO+N{%}dQ)Zp><uARu2*-ACbIU1fub!iz8+%MV~kG#mg>Lrsnd>|9&?ymQi
zkYNQNW|rrI2B??Zhfrg31dhRTNWu>AbiP|yQ~Lur+=<9YY#DNdhsDhJ#C8alFi`7I
zK?wB;`1D`kcKk_Qs6Jk##SXXOYC#rj9!LR?lj!Uij<Vr=Py|!K&JD}6VVlehq=r^P
zR$O{5**m<yp$F{SyY`-$=?B}m|NpdircX^?Q5e4=AQg}j7};U$E5)KEOe|VZQ&6du
zwWX5~DVsDEkc6<hu*6a=5+s01i&#MzwHAUv6bVKNiDCnk3hE3=KnNfz3Wl&sB?3LK
z`akqb-_M!M+?(9@p7T4;d5+Pb>PXd&{s-7;i2E|}K^-M`%U2GaQ}l6#ze*nj!kq{6
zWuyVlGr6_BZV8$4(kR025ISAck~Dm$1svbGz?9-iua9eW)kYLU;$ai*V;lRCbWU{b
z>Fub(UOZDWnywKnSde;K=NGM-Zo9HKcMZJ_Q;lE&-=c%C_AqNjytk8i?<(2Nj^cIF
zNS8X0Y3i`)U4VQfdx|ST(iw|RL3&WIjBxr1866X7_mbmq?t>(oyS<6ZfWlJ%=<lmk
zhfT8$u#|`w2Y-(y^+>lArY4D1cqjnArs^mLqsbyW6<!Vo2qJeYq^cA!;3GqiJRvoD
z70uGXLP`XY_Xh^3&vh*ZlSg@3Tgo&-UrTbhiV?s;;2!cdiXq0q^p-_GAf=&K1YfbI
z3utRFfAsRn&52RJqMEUKM0<1mmPjPBlI26V5amY29dTC(8E_koHLilD5+y}EH34k9
zYlRRGRYsxFr(<+xKdMpIhRtgds3J}8;qzgTsu;ye`28W_w7G4_4)Q?-`(CoNy!8F<
zf!*9}ypy>nK{gB2{TIqJC6>~}%W{Q4gLVQsSvtVTx!YiJ0gxg{RN+_u2q|5|ZwTvE
z9I$L8$VScc?vmHmN7qVx>0Y-WP6Q}Ig@Tcmh?ED4^o3Bvj|z3!3}070=8kY%xs3>_
zCS2N_BTM6`=o=sGx>{n1xmwED9FBUW?4(v_2u!#^-L>R*+TQBtob60lSa44WnwOMw
za~2cO<g(8b-7w1)4S95Jqen!#kipQ?q0N&)cXOAA)54c%5fM~|PX-P^B?HWl1x{Jg
zax%6K?i?~PiDss+zC7VQ%eh+J4bES?K-XX@i^itKkX8+@C#v<{Z;YQMoXIpMbA&Qn
zQhzAF-M2-%9Czi0(D_aWf-<@z$@9<eGDG*{sV8(SieTJo<b*9SomlK)onS7pO~R)I
zq*TR2+MeESJ#&Hg1~9nocx$qAak`$jRrWr4sXu+rxh>g}otJEx*WQSZJWG%KE7{)8
zQqzx}@HHiQHq_N|0wF<TVb?Ov^jD7;vtxdAKtYTi7M~)Kmu{ucu>kePB}1q<FA!|N
zDqzM!ku|e~*RDtMO!Vg$$GI!|a!>{;TZ1X_AoE&5?)9Ce&r|U?hts)x5K32om^W)s
zdZ5N6%{Y41+iIi(ZBOgtUG&9#UK+t;r_T!RpjT{ILG>l)HbYxm+b^s+Hd`)>Q%8e9
z$6<Bx8@mxj?PL*qYMZt}ka0q=#93UJV<fyt%x(S6{!rD}830ypq93BrNdV2JCMMTm
zJo1^&ee$iwYtIHKfINqwO=zQ{k0E-wJ5Rm-UDf)$Pju!mnUydTrwKiFac^^XEeKuk
zKE_PcMZ90!`7~%l6L+TY_<Kx7yJ6h%2kk&1FU!nZU)G6DOW?L@yX0FOl8|35hyPBf
zIOQp&MLXEgs8g$Rau`uD53}Ra<fWY`izw?sWOI^oez7aK@3Y%KO;MYY#;0!)?15Hz
zNy~G}nl?Fm;Ot&a0E22So!;Xj$XJ`5O8}8iE$>sq$@IbbL#hi<V3`1y3W8ukKieMw
zz8y0Ixr2B7S-=}yv9+^@e7T=bSte9cE4sB&`wi<IR=<<~Nr+$`UVje8X9S{1<D)}0
za6CSzQ{6t$>jV&_`Q#4PoKI+m+jUZlXmqcYdsBcbcWic4A7nLmyF#rH{R6yI7v#@D
zWp71evxVRDUpER~!3lC>k6$$-1;U9oMs*gaM*Po+ZJ1>J$q3TxDv+#iu$Zqqf19En
zjp3g)sI>A%l@HD317dz5zFd(f^KY@dwC7O<F6RmhOAGrDcg*3bLpp<1i+CRM|MTDM
dK%8!NnJ}epHMM&g+=Jlp{W8d>ZpYrE{{riu-pBv|

literal 60017
zcmeFZby$?`yEZzAfRZYuq!J>nbb}(@-6hi99R{GJv~+j3G>S-fH%JUacMp8`z`NG^
z*4n?dj=hh)kNwZy&jB(o%-qi%SDfd0UC;YfURDeXod_KQfnZ69zg2`l?s7pOcLwgE
zg5Nlew`hVNsPCl2-h!XtIuow<0YC2BifcGPAdego-$--D<gO6NQ;5Xd*UGMmo3n1N
z$}5Y<dmdw(#1Gy)E&M1i5+zbe=SPuR6}a$=Db-ih&TLfeA)c1StY(3JSu;h#m#2?K
zlpo;Zvx}<`poqkLmU-|i;^lpbj`uMqVGG~I@7&ku<6gX|Ev{vO@|h&pqA5i3LiyUC
zLm<AI1emCwL`}eyjuK|q5HkeVsK`&Gn->gWACYffsFR_f-MlE}L&v^(q3?!Aa`U2P
z{{>huxO`U>Y;Rsbx(j4~?gdlX<h5I|N8E!z62r?YDps89<<7+=C5IM%emd%;NqO-Y
zlZdo-O5k!=xXNr;^!lvdCph#D7$wR2=i)au-H3vbrK@;!6bPhGx~;8^#*2i@Rw#&o
zQE_Lkk==e-ygQZz1qEfQ%rJnDENoB#0x4}6X5^}q2)rC^s*cc9W(M;VTIosPl<SUV
z=Hvm#@$x>@tjXFE<-#WFaMP^sJ_PbWQC>dv%a<<$)$=>!O<orx%|1Rwdd<2KY$kmq
z@I@S-mvj^mNX)nJ@HE<MSJCU_s|YGs1V&%U(8jERLEULNSsnE1Vl-1Ca_WmZrEJcE
ztyEp@$gwrte)UfZHC!s9s-F-nz`tU;JK1#ImMbbEB0VR?TN8eM6iHR2Rnzyy6arEI
z(@yS{Sw`t^@h0}5^4*$~<d=JD!P7cf+~`4G=L^&9b(I!`jqslVJSQ^_7PMAJiLyWK
z@+8L=zGeDmZcux9dDh2CGO0|e!ogq=ntk`iG-z)EC+ADBz*SXc<xIN_QS;7GFDK_u
zpUrzmn+!hku$!l9`D2Y<H=U_+Boz58jwKFjd|EnI?1xHe(wq6@c?lOpgL6y0<8iV*
zDvZOAH`N5oPQa8#hWOTI<>VAtEd4f*B^KP6q!77k;M<z1tq@%;FjHaVqEIgH`_pfn
zw`?CZbi#2FNp+yY4S}>Mrl+TO%}|UPSI|;Zix6{KM^PQYh68|c`9@AO`!vUki6ox4
z(3z4E7Qin~QqF~tm8%whT&hj$kzB{Cn(KX}a?t$-ERmsg=OngPWV2+}X<WyiN6w`Y
zI8YxZyVvoD+21*=hz}Kv9E?-rllnAiC|c*e@=m$~ftZh0jq4WBB)k4vdkiK+m9J>~
zTXJ2^+`~8ziX_1tJw;ZcM?eA2DYyLN$2%yWehyGRxYx0?w>IaogGI_}HbmvdM~18v
zvwP+l&uS9tk4>6CZPTm+gTX|_?m!lTm7+f`<?C57M=moS{U{kxkIL*D4FdMo_rAGF
zrm)vo$<neg*>k^><zl=<PcOEti`{8m8RZTVI3&4#fs4YSp`l9SUZM;`B#5nOp5)Ix
zsJ-q6-O*`Pf*dCvFk%r;fq99V4NYt7V%1{pyoZ#c9gDUvSXdM@#KOid&knG9Q6MdG
z+2!;O)RODEZME%^bz%`^^G?7e-6@%w<#70Rh2p%RNiNB+%_KLcs-!wQJMHUd)4XVD
zX}dZ)zRmB%wkjOw&z3Y_@f>=#hmxY>;FMGpLLgD9YX>mcsKw#>NrKalQhCDXoP({g
z&T3(+y}_Z*nsi#W+7|^|t`aLcyY62zRoYYPsdX5>2FW8qgrY@6M51e@vl@8wwB1&s
zB_$>EyVYqry1EpB3wLb>1qMb?;p5}evapo*Oq}j5;orOaW;%qJvl#vj;(MMqQP2`Y
zPDg^hJ;{jjaU~Sx<7@bK^mx_Y!ZSZ!Qu7C7X>u+*gE{plrm$3`x(1bleVU^){E6f=
zA?f#LOXZz&WlH#h;=Q?+KNZ})LdI3RYoAHzNp`7q5|Ko($sz*R0^2Fv`wV`C-acJr
zku$F6(}<0$vN>K;fQ8z9ZLAm}8b(Iy-sH5Q+8f836QhU(S+8Fh%%Jl&GWJ_)lF_2N
z{@VQXTV~Y-j_J}*;kr623r3PtF>XtaqRH5-Ze5Mc3&C^gjdo0CA4Qw+%kFuFJ^zw!
z@|sD!<B_O*?<gXLxQ~<~iUEq#{VB`p&6$K$GF(mkZ-$D_SDed%6X&#9*AysZh^_<A
zvaW+Y$0A}AXHs+P_8l$y{rjzvkx}%dn!Nnegx+iryl_LCmqK~p^v5!2q|+(|pFMv7
zu{G}Q5TKACa<)bTbCOW3S&pcn@Zm}<5_<XKVh^qgZf_WtneUk&5Gn|6M_2Uv6Dke=
z!Oud?YRkNyo*tGBtIfGatt1|&h~9VekMBVGNQ#yWsU?EEC-RX@QmE^Dv$4>yB_cH2
z+BA!2zR~%7MdZJ(hveiXeM!6)gXtnfEbqVAJ2<d=ojbzM`kR}ONI@{NGZ=nup75Zg
zrrMa!tjm#FOw7!}Ig|=E+3jOuTQ{r}J^dTFYbpsYjcio=(jZ6_>;}K?hLG~AZcW#4
z@(@5E`1TW)n*8Ra+uP0`D0F0vR!Ig$WD?f4h%Yw6th9(pcPedBL8203ubs8d%Fd=^
zXCGaA0f8uM22U56xStHLHhMfJ%@VzN6PWSQ0HKHX(54ON6Vi-6`P=Lzk5d&Kq_ULT
zL6pSWzoGJDSRun<4DX+;x1RP8>MU>9QbN((EV|08ET795;Bc7Wk7fS8DPHt8YYOp?
zQ1M8P*VyC??lhDn8j6g@Z{%c%IW7Ioa$k$USaJS-fQe1;1jHM#*5q@ku%M&M@r_Zh
zzB}w+1EUH8SANHOoOF_7<~w;Nk~LM<`=?B&#YCu*|C51_qGpYU8)xg+k}-HKt_MBO
z?tGwYoFm8VBEOe!^~rV8oYqh#Z2>Vj7~NsQ(wqR-$}w28jp)Bm*BQ5Ohs22e-^}~}
z|3EkU{y$=)|M&MupNG}BCpX*><i5C12tfD!<NOW|zoL@T%E4LJCOy2Pvvcr=-BNHU
zCIlk=+YIDX2t*?A%NKb&J3G_q75k*aQ*ZeBs52D=qW=sLJzMaM`;&`=o<5Zz;<nFR
z?TMqTZ@i3@iHwLCsW9oU3kEJI9eQ)8U%QK~!u^Cy+PNT{J^2o@qB@oZ!qJ8$#Ki{&
z(u6}o0n}vpeDh{mZ|^2uk1h6<m6a!?z(3UQB5eL;(qx6nM#DkhFu+-FHtfNN6ESbc
zIJ=m?E(Q^2!;esRpqsh5P{8ZFW6L?kcIG1r2BtXpaL<>Uk^8b`5|$6n9NX~K#OZ3^
zQws{tFDQd`^dPXSFRbBWyIxcb3{24apWMWkS|)*`cMU+HmbX9Nf*E}uS|2Sqnsu3{
zeT<p>GA1FR;6Cni)63I6R4TBq#%~e+BMm(o*BeiL{b|{jYx(3t?y(CO$@zHmm5~q&
z*gsY6bP(gDO~wi}++kU<%G%n6jUFfYc1wTIQ9q&Iz0=jxqnyBD<p@6-Cnm)Ocf5DI
z2h7cvik!2Kk(L|@@f=q9h`leLv@{>i4w>^jkm0{NT|&*d3#RbqW(w`lgU%Iuzv_N&
zC<#6RLCoZ;D#$CTY=|(X&*OD*JPOt!JWZ(Unl@IXML5MoM>q5V87=q=I3Vf3+bsjD
zEb3;eXQ=G^0~W)k-vVhop0YSU9F`jeFd2)CZ(<<|Y>f1eAQ1dtCMENGtVhZT3>Ios
zP(>&HwCiObH(eV@i^v2wu(-NB6BPy1d*1p4VLnmrN9$R&4p2!t4Q)rT6FiP<-<-Fm
zvU+-?sh>Y5N_+s;egAgdPgsqk>YEK({mlo{{VX}O{Mx~RvEttR5v6&a+&wcpy9;^%
z65@GyWMl@|$0wvu5ZiWRAxa=W-dJ({q1C(Y;S#-O$^NZ_Gd7<qPp{L(0I=m?A7VvA
zNoZ@PNr_T8ZIb7FuD#mY+bvFa7wb05ySo~ic$UD<9p7*prSJS6)ptJlQkgAxgd?!v
z=JCE9hm548Vt+C}J^#r;zhKG4gnDl>zm|8io>I2-phG`@4jQ?qd;}Z7e28)98r<l>
zX8-<*i%&*YY&}~y+5t>3*X!1;{Bg+1+?#9{--Hmc=Zmf~f_Ivvcx8(O5p=>FyI_vy
zhiijGyv~K8q`c2d#tWk9)#w-*Wv~N;!^6Xg_}wd4`;rR4n6#XnHJXTNTi+zLVYic6
z7r6;SOw=j|s4Pg0>=t8%Lk^?@p6cLp0PW#yVPIfHD%-on=+ryYa&vP#)&^2#WoEvh
zq!jjUZk}4g4#<#BE@?WayTOM>_DB%l;hSV3ZT2VliTiT6Sf$Op0LrJ=M;qgE^s0s5
zl*_OF0FX>dM<<(YHk|FgRW;tVd9jueHd3M|=;!B0OHVJws9E`fl?)%BI&XP1Y&}#=
zR5a#E{XrkEaO_`k#ic(Xa%bQmTk`A(@exLA(|8o~T>5*eMC5ach_MpA5`219)AK_Z
zK(O)%fbDfWN!Pz|<bJf@&1^B><W+|NMHfkZ<;@58Q9l_~7x3y%G1knazcTz|+VlO@
z`d|hij&jigz;9oWo7{vXxAnZ7sO02gY*HTa2wfkzQ<do;g`>qp9IHvb1GF!m%kEtk
z^zv^IGn?Tb%vrCK*}8>~7;i3cadD%gqg7TX0DG9U<PdrLR@vin9E4Zl8%zdG@o5@L
z0f9j|_|~!hgb+I&PnrZ&Bf?^WbscRbDKT*Y*m)KACKL<;(^h{h;NnyFvxK{27^CCj
zG>jXAXX{-GlyYQ&%a4+Qu;%nh3WP6TSf%->Bw)c2nSeG~0OVe`-$gIH<mTokDJxq%
zqCGh|>3*?Qb2MQX7VRM_;_2DI<FZ`{tO<m#uBAQ1_7S~$_ckOL5V5+SH?HXj=FRT9
zt3T_$Q54VTRz720!&J9MA)zQM8@k<m?LBMTiWMEd5kT%O527_H7Lntx3BW<{A3jX`
zic4b-oSBHvwUmg%GPlgIJs(7xw55Dr5OvJSp)!EoYZCHJ<|nM|UceD%e3Ln5i(Tg2
z#ME@0qNzH!oaq5Uv=}c|dGqE?(Q*;JR1CwQFB0<Ma-^L3ug~{I+mRsZ(l2SW=5nsy
zIR=+oa`^GHw|%QweRN|Zz!5<HQ8h3~L&v}nUh$_Y(x?!9^5jVjGbaFry?<}srMbAc
zNVB;yk{5|20rqgx4uKfK^*>4*WgP%UzO4H7>&*%jih$;#Vo37H&O38O;{*-)Xf)Cz
zN8@mXuV1m~;Zt9}m>zA8%8hF~glsx{t6`)v+=N|85Y~q}!YQkqw=^XbUp|haa#|aB
zeE{uOi)GRY4n|vGV^H7CC-c;AIM<CR!RQO`Gp|%ODDG-*da37r1yef;Gnq1-^q|Kl
zCf4#WKvKuP4UmZJ0W70#?+@|OZY$2l`9=v&2o?#KJi?%ijkBjJ%?$M=Au457(;PCc
z`VGlfK4jFtB{{Da4_9k1ABE}q?0C#NLDd`iToii!@9LW+j)O!e4c%)e9|y6#Qvc^2
ztj8~fumdRBj=?FckE-b*1P@H<d#U-WJ;}uu-YGesiMU{Y>*yQK7e1}fy?Bv(^&Ya0
zb(?F<%*=>*98LX}KA~a_l^b=hzkKMsK2?3hTV+a=8(vG|oM#rHB?v!M3h6;h^-f&?
z2a&zx$RliF9o24K(+xcsHK2fu;$xy3)r`(~I@jaf8<94+^Kb8Dec$n*@o-2AKa^R$
zR7g$F%Zc<kNUz526d53NI6jw4E{tK+qT}b+G5ix84n&?hr;SRt1LNKlK|QHhrcn?_
zgF}H%BK;2ldGqFDwbgVsINv;*`KE}>Vx4-)h^cC;0xV*VT$KV<)aQsCD}0k<0ehAJ
z(o!s7Dq(_GFF-h6UVcwaLqkl81@g~>o71WK@q-J9S5HWhg&67ip7*-zoNpFHpuD}%
zro*aN4ZKHxefl?@1a32#HB~XEaD~+J<Y^Cami2kY;4A4{s5V5X=b2Vn$!0ZKE&@XF
zjA0D3o^E>{L^(5efd=J{2!7}V+z&q#<<3x2n4}=-H;P|U=%JM-nKnM(vIR633zR)K
zzct{?VT^bj|5eYR{EmKWH8peYQpjl72&gfVEkn#ra_LX-vc8J`DtBhnhK{6vk{QOi
z5@xhXN6X1nClv1)`@TZrMDUqa*EejaI2Nus4Xf|D-)%Ts?u<NI3?R>!OA$ag1==w=
z!<oBAO+*DJ6^z*@Vjm;($M@kB&S!Z6zbF$_1fS|F7QYFIGH^scxIOL2xVXHzMh{}J
zN2F(Gm?0&c;<0*)`<_FBSysGRm4wt#v#XJzx+S>zOoLCz`@20azDF|d(D=OK@i)6o
zZ?zx{JD%<uAg}`|iV&&kyk=l1oreBlf=oo~vlZ9OE3eGldl|z>rRep82xZgTYA91e
zQbtBaL7@$U;7+b!qAcv1EY@B7g<BAHvmTLCQWQ_m+IvXhXI2D0q@4l5Fs(l8?j5AY
zv(<#a;9zBt??)Tl9Jsr=#k@cib0WnRVl;cqEYr}sh6gFU4fMZG=6#q``U$7D=Njv&
zXaxoJ5;$!*=m>m?_Jy-YHr{0@L2~<V04QytW>pt}ndS;(&&>M0Vd+HiB|m<=FT#eL
zKDkA@ajy(TLr9QY_^e6kNEvhEYQeUCnp9l!XJ;k%J^yfX68|l#Peni1dG2+dRhbsA
zmgv?v)NCv<ejL4d$%dSY$D{x@B>Mpf-+%x99jNYJ;eqV|d9q+j9AYl7oOu}Qk(qN%
z-_|bP8!yhGwW4<|EZV6-yl9}fBbB*PoZDCkDg1hK_g|Zn6+p$IlAsg|9ZBSNcwe$G
zULheNu@2C?&L3oXWKFi{D>xyoVXWhBDPG|(iv;Ru&n>4iJYpSU7(|`)#xDI9TLS`;
znYDk&XNZQdyu^|}XyUEYnP0VH<q<9V%G{@9G<I1zndw|4Nd}RR{7>NBd?w;k3T!d>
znus)GOq#b^&ng%jX(9NQlq;ultqrHtNauyujxDP?ILSa-ByQDe{9P1y{IOyD;426u
zS_t9T5cQtGVkP{Y;8#4NPq@jkqWW7k=pey2P7nTe_G)XS1K-9mn5D_~GyVA514y6g
z%De~gd2qG7xtUWQz})MA+=#Y=-#z~aC=F-JtdA8XP*74vfB*h?YCcUE4Iu%yyayi(
zzFo@ZY`q!~T@>q`w<=xt-c4>bdv>$fE(keoO{q-R*l@ml29gl*jdqZIaJm!*V4Mbm
zfl+lX+J626G)9RekW1s)%pwRF)b*9$0^G!scoVE>08A7Pz%H`&97uf)R344nIC(p@
zc9<jGV@y>p>zZl62tkAfcj3Ja-@^WwgHp_m%8RFfQ8^xss*Qti4AFNkycyMGf4S3q
zvOH^Oc$nR|=kXcb8>~ncSUEzLta5}YZcUW6u6+j!e0#e<Dk>_|>1rz?cJrw7<1KoL
z^{J`?sywUd>ZwwLPnuQcM*45S=o~j-7>MN&m7ZtzKzxORwE|rmfErbM`|`m-;FI^u
z5NZ2mmSKCS*<`s<U{H_}z(1o|QiMf1^{N3l<QzOi;P20Fmw|>urVRG~s9o@?tmkT@
zxLy_j?4ib!Igg=wAawe&;@pPCoMd;&PiQ!*-rnAEZ~&~%H!J!N`*?Ruw`Xcmvj9{y
z{*54rEshAy)@=B>2*9k9JB~b~8DcnIS7*?=%l&S4^N|;y9)T=Wv50v?x5Pq86kJ_h
z{g!0n*#<#6`cE6Irt6j!!=$6Rc<PTuJOYwAk2FXzJ)yVwj_+(al4ql^!Vv&?3&5rj
zBnTj~<DK~u9D!4z$mcSMocf<q7>MYV*0Vf_oDE<BBP#)*85$MFu#@KN#@8^g)+e`A
z2B9rFZ_nff2V)u;8|Rmn%7cW)bA5RL3N<02U~i@UZ~Qt22-v_LM~ZcLGeA;>L}o02
zf3puKO;TVXoOAM`qW+iOKDF%(dQA<XO2B_A(r?~yFlA9<hL|#-sDR(KP}2g{HR{$3
zE{{rlMlW`nOMt9!%k)3r7N44{Jx;7aX`uZ^slRn5nB&Hn41lyZ&O~s_Z%sh_t^=WX
z<;Xn*6pbJZAK7n+e4%cmrcE=v`q86DBRO*9?4|=k&KF;sW8=6TvXTB{@xx~epKwO1
zEYu{1uHY9{KG)u#a1k@VU0Q|tC|w36SY(g^7C1m~>AT%}(PE62b$CB?c&64txL?i)
z7=_a}pqZ#6-3Bj4t?JK`Aw&IOD7xG1{RZa5ss0JrjWYgC9GAL>`Uz38I^LeOaKr-x
zzqvJa<DTzn06j=huZT8#!x0%7u&=d1`eJznh47#E0o{uGhskXe)vy6k6qu4kM5Q;}
zy-2<6&2WyK8aRe`hK5JLUWpPPA?WjM-uPFP<_+LhkyO2j+)7EjE@Oa+h6Miy1nT>5
z*~ycKSSzlMP;0LwkCC;<Rlk@^20@^RzSTT(05}53Pu|h75@DRAyv}P~^d*=f(3KUx
zF{tm}*fQy;$0WCCdoFdGC@?PNN4E|)GNKHW1a$yM3IJGRsF^MbVyY95hyqj|*oPkH
zjP0kNv_)IPlw??!)(#;(P9r-d;ckF1W8XTq%ZyzB5EychkWo9VYc@dm3-4ZuprWRx
zMpT*rkznV_s=`P6)U|gO9ID_yZLL<Y`!d5Y>S2`9G@ob=m<a04L`VZ#(-1O~;PqM9
zrgQy_wTr1`Z#?_@L|MkBvo)&Pd?h;0aLl$27(~<{RXm*h38*Z|Ke;)*%3rOmauefu
zN;Jq3z*s7l)W~nF=3iAtZ!8S_^}k51-!%be0kreIBQPg@`&**`9553>=0T<B>({So
za8e$}0f1j+J>~!tmjtoPCeBMBMb2ThXAtBXCqVS*`S}|W6iI07A4_)G(RErMk=UB5
z8vffBWK?s#&)D7a0eKXdzSVqFW7h!z9^M-H^=ZJJ4?=Np5HD1oJ{V|lJ2(V(gisbh
zg2^khgFxgj5)r2cIjw3Z_tCB}?nMw8YKea*1^=jXfGm8Z>%I7e!Mt{P(9i79DfLgy
z3IJGc5^NAH5dua$uS+plhw7Fc<WFa3?4?++EK&x0aQIB95;@w&?aU)%VzTp;a#Xnl
zOb4FMdS7g*#2)}XA-dOi7>H4afP?!V@BFJ$R>`p|%w9C22?XMS15h@g=i_UTOhFb}
zY25caR2o>rdpSi3&l>xhBY-{jzh%F|)0{{3vrH-dYOpgv3bJXHvP}VVnyN67iz_QE
zWB^J)dQVes?h6S;1qFqwc1lXh?qEa}(ugVCZ03humLM7=b-ihlajP@$iqGz$V$SMl
z3R)GS#0mM{_wAW5ASujdKqy@KcczjBz1Z=&E&v_?6oV+<ZO0gR%-T3+xkCNF^#ZI&
zI<-p?vl>gAdIBXNp5NoRvlSVVzmjBQZk+5afk$F|yMq5SX*fgC!D_BSgDgx8ND0g?
ze6L<9XlQ5{9pmHSb%3%g1x5PQQnT~4RgGUVvqt&be!*+cECDp%{0*<rW$~9hD6;Lx
zpN61X;HV02stehX-@c`S#BDZ`tFS&@Q$m#ooGl^~uqSh%Uzm^Pt7NU7uO`@30=A1_
zqY-%b?&4vhvI8_Dlf<I}LZz~?v2olH3xY$-Ch>F{SEzW9_}cMsL=qhkyGWrsDB7m3
z|BzuoB>ButDn`8~ZIDO@L3lzCXgH7|0WV7cxw%2IA3*~d!K{M_na6;gAI&)Q=YfEv
z+2C3h&uLS9dg|)5J)?nW9ssl(@U=tOF#EgT_ayR<5!xH^ZE$B~Vv+@dH#<kltaa?H
z?y(CWs6&ka3d{!N5O&K6c>pz3E{?Y>ECB)-5x9vzlH}f}X$UL;3gzg8TiQ*YZqZPy
z(frZ-8@fOKtprsyGp_7%^O;+)%?i{bgieN#IY4-C(6y<>0eLy)Mu>Qs1Paw}K#5!#
zz?pw?+YJeUkpQ_4n&=RUK&rle|1ck5=Ae7xd4?fyPK?D{6-172@3Z^iS{hJyU+dI4
zDyXZg*S1$#PL_v8{zrPQ0D2t=DB=JV^26-#XLSJSPjtIY&wRTr$UwpQzTI#j+xGi6
zu#YFC_aQyhPu=u7U-5_(?V-#@qz8v01>IW00YVQ4k>{6`5~zj~-s_4@ws8#9#5Ki>
zngWMtemkfmH~~-;2BB>|>2Fr;zK@DJ{OgZ5I9#{MGpRF@2H>*oq2MR1%l5Uwp~w*S
z=WF+0>?)NFzUnIFwpzXD9B9k>n#)rGX9p_v0Kh$LHgkN8fcFE(P?nSo+&sqJu@-b#
zl|~*y5`zET#AO^mvJ-88(V9eSmNiJL&wdkD9mk4w5E`jO1gK9U_U>P&e(pVOxtk~s
z3A%-G=ruF;L7oCBKw>3!_of4SV9}@d01BetERPm!FSW9Hls5nj|D**x$&&h?2(*}*
ze**!&_kpFEFF%GrY*$kkz}VRfCl|tD!2i!Yn^-O+1?}za*{o-{GAKdJz`Z?7^oI{0
z(lIe%Th(h9N0QA0{dy|&?!VVd{tuL^Cf29iOi?ky{diL&b`O|I=PlNIEa2(7cJ|8;
z1A$GEhrzuIZlH}m&rZ`tF(Qrow%r&M_)GUq00uAqONs)&`M=l&^#6r{plm!~V;~G&
zJF(R!l9K^h)mxeVA;GeDAzYts_C9qT{60*8VQZF>J4w;5jBSC(>~fZR8ibHnS3BUO
zee-D<qR4)zb@hdR{25^{91+t*+4e_*Gwc-yIcE6M9;`7eYkV%;|Co~_sFkV8lul>&
zBDL%_WP}zGfgsAg&=g5GH^rCo&OUZ!l?y7cb&JgJ>a+~f<&(jkUwDo{s$7TBRE)sW
zEgVlfc+F>y6@(A)2tBIU;FOKWKU>``I{nP2D)Z9ZAR{-L9T;+ZmfFNRDrbTotFCyt
zc9>SAU3_p*QM!jYY;7rQz`4^^A>-|qxU@Ww%K+#!s;MAr!e`K`mPSC48sZu{0fD(S
zE~AF?bv_**_z~cB_&zDVtPlj!g5+nr9R&9Glul<!G1GImQZmc&xGQ({#T;L{cA2?I
znB&I(DMqMf?d|l>o;^E!(Wo=14uorMCbbYR?Zg=X(Bv9V{|05#b5fa|?ml!s8a)Fy
zhz0@Az>x$}btAn%REC#ahOQgc?Ik2(tlZj6W&;n*Hmzuy{_Mo=hsEHLWp#&Tw5LE<
z=(Z<b=`7m6prH{1RSzEf<uAYqK`^%lts!(G!J(nW%1;#M)LmVFSZZr3nQ7Hnt02VM
zP=&PltK*vUtBW04u76+Ha~4VWRSmO!DwvRJXVwedl93%rvqQ#}=0A%sG<?}Oihr3!
zQczZwUVIgNf%fS&DZe|LvS$>NVAnJg+;d~B2-Jo0huR%LXHGdK(=*w9tMGS;ij~5F
z&M|!NB>Au1e1eq%ZN*LWIp3{z3u-gPf?x!4!3ff=9q80oWT~x6O;<|<;7DyvNl3_;
zuM~l5;k$S5@NjW62pBb$&yGM&DNug0zp({Ers=_IuL@>?URUUg_Puw@cq`&ye4eWr
zJJf1Qh`cZ&tU+wu6PaL~(8o!CqX5W_PE0IZUS6ha=IBo-nW*uQ5c6s$zltbv%Suiu
zy*?fHYR}A{%VPVba5YJx(8qEi$Z#seh^y44{SKFmOm_tc-Jn&kPvF85bg=jxyEW?a
zU+%Uds-RI(QJL*z9@0P~Mh`)#KY;G00Qi*w*s%zJ3Ya1MhXcayn`Q0lHYh?j6;X&h
zQqC}K8>x@7ioD{_63K6TDN#F-`;zz+XIF}CUz*>0)~M~P%!FbNFW?^g1q7rx7KKL^
z@joIZ)v?$Jy$iM<q*KIgfO60s9P)%!6lhgEwhO+Xaxhw;#w;Qxwvk^@JOs$a5nwVe
zSXq_Ouu0@W^I;W$uAs~Vy55w4#GW++k-wo35P4!tx>{-Ku4mMKDqB^rCFX=C-(`-(
zQQ)wVq^HmPvS+TDZ1kw{24zRO#-!Fot|HzsAcrK{ZzD1vP}^fOS=z}#Mh3^C1lRyL
zVDOdfeQt(Wd)<0|!6A^|5fBp0Pzk)rX}cCm$IGh)r0EgXe!dK!>x+4n^I*uj;Eh4F
zSXxH>+Eo0mW$yZS_j=<fm8h1kiLqdOBf5PlZ-0t10Gm`1lm|Zk?9$T$6)cCLy=M59
z&H42FRW@h@(^S*86EVGZnfJ;8kS?aSG8+H|&3fk|U~~YxXM<W#K2XRJg<B5GiPX=Z
z?;>QE)&LyNqhHyO;MX@o+6b$4jTWFTlG4(J`xeL_d_l&g_iB&7hVS@9$B!@CMorEE
zGTg_1GQ)E5cS%Yz;quZGQ4|^1^(cCTi#uB9WPvEtf~r?MAn>}Ldu?<+Opv47Rv6z0
z^|J_Q2O>p*)5cg1p8z%WO6434sLYr8|LHdgy=)-s0Br|s<|DZSrvNS-<_N+i0AT{k
zNh*SHKNI#hT_4Fq$DqGvN!5RuYtwWlO73;25Z7%1v`{XLySXMeP5>FH1*(lX@ExP$
zmDSZQ>zY`P$siyD04`KmP+TY3&wr9Q1=`3qH!ZWWvS5Q^<fBS*{3_cQKrlwoY*v##
zhKhNBvYD8+E62$UJ}9+#{@ciaY?uSR6_MnlQRaQF8{JQ~b+$=$biL1v_)ix;MPz14
z#jf{q&PTN_4gkK9BSz+y`RijVAiF?d8MW~~NI++<IRo~ya&l1+CCI|EU%9Q0yHeGi
zm1>!vR6hnWVY#sfh!YbiSRhLO^Ve4{GRb@zKqMQ}ahXLF;*)JHU%&o<O~Um+Q)wBP
z`v|d3L)u`5m<4PoYkj`i$8X25ViKXK%(_97@X2s7QKM5*^cFQKX|yugL6H~{SB6Yn
zrgZ`_f|MF~W$zo<1v<3TJ!6pF1Fv|SnCEoLNq0i|uh?}Fe{vS-v|bvDANu&y*YYw>
z2!YEW5X}8CEgM+V#Wtc}Qc+Ry5TJM%$X)`AKL8tBOZM2lZ#56<>2(0JID&3QbD(6N
z_xoHa6_Icx$P5k+28M>J0>3D@JO@bKPckeLpArQ1t~q;gZjDKvzelD%uVbp+SPBkx
zofCA*wjWdfSn7(R+RF-h<dXQe^{-SkrClISkV+!2ERzTP;-PKt22SovPANjHr1a8U
zRMGYiAJ*sRtr1hr9tJw2k%>v{Bq31OKz-W`1S=Y^EnysfWn<7VV+PDBp2x{N1j<JW
zs*UDDnO}h@Yu{zqco5cc(m6O#&AgywZI^5x$hYt6;FQ^sk(f>iX4s(=RX|cACKejD
zf=4<SP7xOB%Fm-1RiGnw^5OxeGOa}=z2x3DSDRlO-QuYR1r`f&-fQ<IC0%RWhYuB2
z1QVtRcHOFM4%&GdJQ|PTyWvbv&h*#?uR&|+&!@MWEISQKEg_-$X0r84Sic^PJY`7h
zD#4=5&1gI8qUBM{CpoLd+*|u{0)3PDep5VLj9WL1PlU|KwYs2VhcV7HHd9U(?rmUT
zK%--^{N5t$_wV0P_0^J2lS$S({Chs<&an9^0~oI6-W#8!qv-_4y#ZQ#F;#d!-q@`U
zpQ`!sr+(4pt|l&sBP7@XgRji}WH{P>q~k9I3KWeKo`-fW^^Oj1w{>l-^1?JJvuaK@
zCo@2!tQxG*y9sV<w4+`DHEpoRp?bl<@_M_&b#fE`nzfR5cHMJJZ#4Y?;w4}a#iG`n
z?v<cw%hCSiruzqiuhp!odFg+6W<ov9sHL)Xw758gl9ctVHS6@0D9nUt&bRV?bERvu
zV+Vaz?%%x2L0Q;eS2O)`JmaFaN6d|fieA_76ND5rsL)jfwAB!D!MU}^Ih9ufqB65w
zIzcS+5mZ%SHd=-byj!wK^Q9Tv+&ijHmb(c9SNBYQP$VweXZ`MQGSiM;Wnf~80pWKf
zy)dh$qN1Y8>%tlI8`CwW(DR9IcigT0m9fHjnENzmxhD;nT=^tv`6NmC<QDj{7yDh%
zQXxmLyT<JDHj(08q^#z*&l8q`5`0oy9&_2NNmuGwo-fgETs>;1mkkC#$sFO$oZg9g
zXPA|L6(b<1r+HT9sFr`YiI-=mxb>c0Q9;Ok7cI{#>ad49qWgK9^5XPnDwSMW<)N?R
z^>bL<y!gX0?xlt8`kgUED-Z}n@zx>!Kz5P?r3keW-NxWhdqtZa?cz>7!K`MNl+0#F
z5BXQyFKRc7i4vMS^fC{<^vqAiYGCAq*DCP%>!vA(I&CJvK*SNvEjRgKe}7+A)8BF~
zF5_t0ewmRkJ^hlz4a44%VdvX^g^zM)Gh4f$mbY`84OOSsT6imI+6N!FtaE$kwe+Xg
z&GyIMIqWuknVaOzgJ(X8$}9b-hBuIV;jP96D1yRRvpRBzQ)pNu>+HUaq=p~v&<+{O
zjcAKjxEiK7%2qmBbPuOcc0CKRc-DBCaX2M6)T`1Q8K7X@uqTSBwn9@}s{KPA7d@AW
zf8t&%3F0H@2j^$Ed{At|X*mRR2b*;%<xm2s#jjYO(EzpPJY$8<s(rFo#im%MUeP@4
zd}{D{;_NC;&-9l8^{R6PEI?NTc?FM<Fh{jWGuor)nk5TH-@WazL*A^GBgmJUS>e%@
zkdl%95$r)|ht@YG#gBMo;hch6lOD~(AFN<>J-rcBZO@Y!U6kKsSyLyB6bBSHoCr6<
zPyNg%-F*KhBy07`1=ZaMvc9-QS8vv*OWapP%^E$VuDulY@um9WFM2C!3%g>-`0I^V
ziM}(;UNmN*V}mwgkW}7o5i<{T9X4A)U1TLt2}s&Mo*QJy)nKBpr|dMP8C&(SFi07B
zHrZ6UQ(ae&b%hK~{jMHC3F#YFLmSH%+qbcR!_5^HysOdqFGdQN$f+b`T{0<f<Kgx~
zFLP78PsQgO4!)OHdof9v4>iT|ocJiqOr7Kpd2r7><f(60-_mXV{IT`jE$F1^f0$=5
zD#u^)tykCnYxm&gc@058I}Lkb+Pr@|PnM@oZ;wj5Vnpl6^?qvRyNryr7-Gq$I_=eK
zJLqm13AI@A`EfZrH8+byKxoosIW{XDkGuLqyOk)RpQ*>**G)}s&{w1#+KdwB?y}@b
zK1Y=1_QyN27jTNDg!P5v>!jli1I&YKVuU7h{xj2MGWQ|-+2HyfIc}F)>~%-4drd;u
z3+4+Ka=wBRy=$L{Ok=2S8pgSt>QV($Ui8Vwvl}A>oG`Cfq&2FFQ75U`UHq#R6AKSE
zrg4!6=#GaZ`UAKDng-NsW}QLvFmX2Ma_iW+USWZx^yhFptRfoc>fp!I(aFiWt4E`x
zB`%%<u*NLvnr|rjxo_rQ&7QPTn)94!C)Mvh;%{EOUdA{L4plKhEMT8^KHsr6JfTVj
z=J4k;!BY|VO~qur5z#Jnw!h*IP$S~$7E(S}dC;4Ywg*&Jx(D}6L0_6XXiAC(x+~}o
z8<}uzTso5M=c8rx9$uqrd1BKMSa#kl<FLH9YHl(Y7JoQ8&%Qk@NKD%8`SCSrH*adc
zap*|(p#~jO4?E98^BYYkcl^@3`3kgXt^>zWdH+5WRIWg)T^^$MAR-F%PJ;F|^WR?{
z+H;YBlKRNnU<SL-)g{prG~eMPr7LeQU~5hNb0kwW=GSWHp7ZATr|4D93%;tGXUb|-
zYXzGLtW_4jIRouQ%@tO)x=!#Pe{ZnHU)~(OVx*6*FflaM8>{ZG^68d5Yd}$anLD|(
zsksZKBAvO)`XHe&eIv<?SRe|6PantwxM8XXWNk#Zi^LG<`iM$QWT?4_t=G@|uU&9}
z90Vkr7#q_8Vfl}sLlP}hOw(mbr9(jFUt}1Lngp^<?tIYKWNf*z!@o?hZW7uVCY>Q6
zqG*b?(TeGn^gnHNhK{XhHEX7O3R@&iy1WP8%FQk%YV%RLnw~4#@NX3=q<^Q)JlnPK
zR&D8f2IUE9oQB0b1eIPse#zsVWGC{+tPKOM;>jQ<oYJj0_mfof_;3d}T=cq2KTD_a
za4+k}h|JNF(D79{T{f;)<`@XlojZ<A#+M6#Jfaf@!7p@mlR@uS2576K<>S)^LNF*S
zio}h-6%mn$2n-C&=O=seM8kL!U@nEHyWKcsh$l1Ra5P1FfBkZlvRCw5vdG9d+{$}C
zdHKoFzy+2blr?`8<&PCHV=y!d@8;+0u^v8>_=;{bkUF#X-1-3CvQx?JC{<A)=ibA2
z;j*vF_Be{|iWepnKGH&skDBRHks0i1X5MbE_qUweX_$77b%xA&eb=ebuKPov)kNFc
zW23iktJrvc-QP5+U1114H`2169E;XN7p+gPOzn5nIhcDHBIhHd75e_g%){(h54lJg
z4ksBtJqw#tF2~C>!+@hsIoP?*J~e=zvE(h!xOqNzyBXwR7MZVcw6o{h6^<-z>Qz@z
zWR;MtkZ+&|sqXO34P0EW=cq7;4TpOPb}XW+m1q$r%Kw<!#FthKeQsyOO4X*af;$pk
z`jVTeVPXx>=!v?$5@;jzbuY@z6-U&-MF@2=0X77^b~jB^pYP)$Agx41Vqf|EBarr4
zrbkJctu`mh(x6(^mV-TUtn}9g<^whqlx!7V)9CFq9O45tXYH~Aoqn^_7kl){2~@G3
zNxV6~K2i>C4;)x&(Z71N$M%T0)yTnol{-?_z31J<0X`uKT99NDw$1YpbEdP8Ln?Zb
zy!1_;Pk%jh?mA)1sD{dM@O-4^Vh)UAI?p<FM9pw4=O}ZjAgkLeymIG1u#y-`Z>qq9
zhVn1|*s0*$=*s%!DEr2Paor{2x|qw#{x~q=!vFK|6{b^~%)8Dd6CgKK$nFnY3_G0X
zJDoh6+8ecq$n4(Hrf_-hS;dk#`?juDtSUR#sk-9@x6NZ?N6cY;L`_w`1k^kb1-rV9
z;>clGRQC$vu|NbI0;-5Ipp0EzboKGxAm;)FgMxx0pa04Ngbj<ez502$ELm92s-UAz
zXaDueS$|gFPvoH|j^>xN?AVdds~Nc+meR;Geclr#-6I%j9wmR%U(epO@OdwfIH%wu
z!OO`Zz(A8-KkgkJk4Hwb#pp`iz`^gaSuE*083+XwoE|u`h%Ui`Un}-`M%^)>ADh8p
zdg6cYzbqNQI`NB&)Y<pk8mHDvsQG5;8T4e|s`xNxCyOiR@V8Coef$)5<S3f)gq&%C
zMz!$1h+l`<VT$-8%-8%E(^Ed{Y0THz_~Q+E?M#VJbKpIE67yPXgU|}mZWt3c?`q^V
z@uAdSAS>$i3CI;knm}bkQ0}%M!tR|D@P&B32GMAMXdwO{N*kyBp(NbtAjYcF9mKO+
zRGRcB6TuyF@vUnd;SjUsm#_~PlE)#BjVzy6aNAx!iq2fJF!qd7+K*1DmzUu5v5!Uf
z^m9K5tR)&c<|&%Gj$XB)ukAgWyYzB5hvP{Mk;h;wUKWm}wSV{}Ec&O|A*GQ`TfpYS
z`CNsxi4NZzjcEb3KkK!k?QckLbl{^oVE@YM>bc<YHB(SC0ztnH(d)bPFCX?PDYI*y
zFWH{0FYvj(qS5Z#x6mz76ArlYIo{?IC@!B{I)!q_QR%FSCKL-vnRQci`w>j0eS9hv
z5va})o}~MuJoiOVc~tL)inDy0`h%z{IyR-eRPmrx0;|{GymEvRq)qEMJE;z41iei<
z-hXF^w94*0fxGw8vqBT0i6+KV%Zc-6{LRq2&QoSI0-)B!bct8@6AA6-3A2Bdh5p)0
zuAOPyHG|r174J9!5^Tl~#Pfa{*QjWlr6?oR`bvfgy;(LMn~m*fOZz$B1)P1bA6Ia4
zZaT1|6#G`BoRLK*yt~ZU*GsV?mSbb{IHcZThLd*dog*7M^L-wiCjUK0GZ%<~?s}X&
zJXvXMR9sKPpKG(q{PoKh@svkj-*Uk+BL!kr!j+Yke}g8BUhqB3xPgI=&ULX_QM>yK
zqZah7{{XcV%}vlOFxLgjro9>_Np0XsjA>UBZWetAAKY?ygEF*BM{i%p05U({M<(I1
z;$kP(Nv=10EY6Wwds#6q928J`9DE|^;nnk-B{WVw=2P-Y?0u!MGfN_M-*lu_E+^U5
zr~Pv@GhN_&N=YY@EE^LWR4c;CYyrYxG3vqtg_3Tdy0BO)X=rq^k88IA53dDJ5P19u
z=<RJhkX65Q9N(^t4Gkzc%Ot8RFE(d}CsI|IvUL?P_rS`Q&Uzg^N4lcv-8SZRE_w6t
zi+cU3&!muU?;#J?4w_HCoFro~{dl_6d0-T)^9M*UwSNQYA9LHk$9A7w&K*+xSX2WW
z^5U?bXaPm<ddCzuLqkF=&O9l3b>$KEi-_J|NN6N2$%ZFuWlKd?>g@ztmfjVGmK8GN
zR|4izp2I|_IO|!>CGM|KxZZ-xy(3(VtP`|4_Eq!?bW?`$4Oz~D1%BssVTz=YTLm@n
zZUvl##0$ae)`gn6wK`foJ)h{Z46#@GExyH*dSu*3iL77vD(a+;@%btOVZ%8tolWl#
zY%llP4-{t6IoL;>FUrQ+#t(XEd`wCjkJpb4ovN&=C;hExwtmM)m$+R2Y%p1ACSRow
zAe(`=M=)hU0>I-9;y8AFc1t0gz%qWUr3d8}R6OBOtl&XBHW;|SfPF>gvH3GiDTig(
zKc(gTC~0~UB$<vdpWicD*0A`#l{j)8Z>&~2js=2}_0!E4WHS!zm>#Z-oVpjoOj{|-
zH9!P%*pJKE4S`-IL=zX(ZE=xXoqk@ra(@o>*}3D7HGR3>!Ua8@wBp=7sSh<6-6yQy
z;@X>XV(?hzbABwS{PRz*JabB-4%ol_599i)#!WzFbm(TRvHf+QW*4;Kcpo3o>z$!v
z5HkI)z2p+HFg0CRC%^ulGV4YY;^%i4WpXs9#O;C+(o{OGJL5Dg13D7~W(Rv4c7gcr
z@aL?;)w}cHEh>IwubTQnway5c?eEwh7h~Mg@oaes?zCrJ9{q`WGo5lihf3NR8M&#Y
z-N*2;E|)Kye<r;z@t7L-6=k`%Mwq5unD~xHlsqpy1XIh)1Jk{AI6qSjPILs~=;gdW
z97}Xx@6`4f#dmvX8MC?Et|td99X5)$U|O7=?GN<Z3ZM$CZ)`kad4F$Xyac)ut7qI!
z>eAzMBGB&9e}1$fK~+v%JTx$XutOMl3YhbV;hK$&trP&T!l~a&cHo&}Nvibw*eF0u
z6%rO6mpxc-G+Apal><-h-2Q(72GVru`n3y+ir@?w#RY5}94wX-r49F4q>c}J5s{%r
z$#>~s`^OS?A2&3yUsuoeC(4ZHK`-}a_?Isot1>r!19L{V_0%;qqy&9qf8i(dCM~{5
zo`-{C2`fKbf#-UrQDlHzOTjHKFW*AyGV>a=BAs^1d8aS<gB#G$ei8<USeL|taE8qK
z+y4YF-*R}M+X6#GTph-8H+KTt^Kf#Lw&fXOJa8rO-?|WZ_xp;R8@olk-SQZ30kobX
zuJGREJ0!PnrYN6O@<rSWas9W0i`dU1fzQEk|JH8%A_n_^<`S2)ZxLuASs>NN-Mzni
z&ldgVOBbDP{vEFmCrzn2({4LDF<T!=sxaJhle!h^$*#9~r<mridmVpqr8Y&2VeJww
zXdUj&tKF=`Yf@hS)q<8A+Z{1z{-;`kpf+>EvIz`Q(-T<+f`WCSdXsl$-%sWUjVHH%
zsxz#r6{^2_*6lpr{_d5YhDNPk%vKX_P@AWU%BbCBWkG&E4+lZenL5MqtmTx7s_Hgz
zb^WshXLfe>rB-Jgqv1Xe^83?p_5^20nR+1$K~UrL2j}j0W$IpNQzET9O;p{^&G(vH
zue-8!z<@@JCJxHZ{9C0(rf?VKbwkEdDV!szOu244a|ST=p<dKG%*bfj3|jXu$<!mt
z8Sce;RY2>WQw%-(<eSx!tL^eO6BwK@mCxl}SBiEoSE5s=u8nzQoSOYQYW9{%orx0Z
zWYq2LRl_gh{q^|utnrR)FA2C%Cx{=T@<o<kkwslGQq*!nh&7?57KJ}sIDOy#XnJre
zVqN!q$mK2KV9%tHj@y}I(ao+ro_)+xtv;-)g1_h2<&=VX;fMFmKVCSD8aF`-%d)B<
z`u+h$@u&BA6p}{PIkX<MeQYI)Q$4RLmsn^dSJcsgk6IpYG?L%LGnvMIyxnL5y1G_I
zM@K_6NNnhR#i{Joot&KN1OB+~&>-RA;gQJe=*(_E!~XjA9U+vYua>`{pkSk%Yc3hm
zSDRFEX8GvrZ|uck35{y0rW|l8T2<yM`!*Z9*q$eCecbaOoP`@sUZYKgpuS*W;7hNo
zbGP!CrhfF*s&!{q3fvSIj1&q+Y6PRQCd+X{Ik3qEE<2C6S*5;4LCMt7_#QC%ywvt)
zdwC~!J6qt(p<5l-gK*^wXdN=eP)^p^$QR_3>e5;B&AzqFsPRTgitz7Xt=MK}UJvdf
zy?hZa?#Z11=YgIh`%PA3eAIs-#78Fh+atx);v#A(<j(Hg4zy7(w(VIvFI1fBk;f2U
zMo!U^40c#e&`FaXL3;uonGO24UK3WH%UYS36xdTJNlUKf6gM<ta$J~qca{*T7<mho
z8(MBnc=zYM9saLx|2{1gxAh1?1|zRD&NBshTN74aFz>)i398TV)`}rwQmZ+Qb7)_V
zZOMTv3z*y?_@IZ(fM&gXU`Si@XDF-S1AdE@dZxfETVeU}^98{Poi6Uha!27-L$5o(
zeR<6;-tUZ+rsPHcZuV0-N=_~>giJsZ@tS1G(-n~}i=)RGHE^`C5^)*`$d>NwoBLbw
z?q5NYQMS>V9tq6*dd0Lm{(anxX8mhO5<2MEp2Z5s3qSVwoSp6oBSeXKBO<D0nM0Am
zC?7t^6s!Bb@<+6lp!noA+Uz`gc#{-6*V`!*wFUmoWe+KR56dsVH-xTj6Ovl3$lA8o
zg4>z4V*~mlkCCtL&tO_2ZDNpwqHqsI>-*r%BUiWmNn^&2-I|j0$(Rs1UyCwYF-TF`
zX_0RE6a2ir!eym2th+35(Idbs4&yRt6)7gPsW}t(gM)900ZKz3inr9k-_MgIly`zL
z$9+g&tfzYiieGrXT5%gA+@YXwTVIKS;@Xei4{s7e!Va(Iw!}PTE?p3zEPZ$?;H-At
zy~uI!L93J@0F^=OFSO+j!^?=7ap4w~Mdy%-5d)e#RZl{_aAG9gh~)#aECqdym@au*
zk0ts81YK}^2tLC-qgJ0_U--4_%}t>VytYBED3y*7-DGvWx8cK5k(sf>Jh~@@&+*6U
z;VvpcqIlQwW12fYX#N9-8crgDBAHfsW-BSHg@hF4+;MkiSndA)4yp#E?N1oFAR+xs
zvdXGBrxa;=<c7NiMPodO_TY&mPrh%}4FvV_Vg4I55w#t&trI$(KU>~2WZCIXaX6_z
zr5b&@@M0nM$uT!hH_79HH!0)#lhmD-A$#hY%PTF|v|Gmdgs?4IlK3JImdz9I!Y}&w
zlRSJR`uEOS26$US7Y{;73`uRt2u@*Z0&g=76)o6Ijqx6ub#s1D^R+3|Yxb76vSI|?
zRwRB{;a<cAF5x(TvibDlDs;abBGIwX4OM?#>_5?(e6F#!orJj|aL*^S<qnCy6Uk8H
zn}albtpGvtc9y4#)CIXVAB-a3+mr9kx44u+y2;C)v4*tBwy;>Pe%K}|ICu6p>g+e+
zlF=X`_3{}woXj?tHk%Lcw^f*`PJ)xY&4kB1BzT-jD^E+|)qYBB{llOOKm+v#*FEo+
zgv!ekC&5i)$!v}3z%qkuIekuNmL!)WJhaWY{qrJAmzSXHB7JIVs-RRO`_2L<Xe6k2
z1PrmN^9V$?ard46J9HoD(+Afmi#8QA#}!cn#HS<JIsO(|Mi!0^d^6A~5A;YJ`xFTC
zNsXM|Cfyc1Qt*ZUv@DXJdBfmf_`vpVj9Em>9a6MA)B0u+{WX=JWV;S8sL#kYV3!x4
z`q4;+oR%NM<{i`ym4L(8bsc1w!TDA=3R`w7S~_PZX&-jZfk$j93sj3hIIQCzC#Ec2
zY}uIyPZ6T^CG#ivVGzy*>lCI4AEd)1IET-7?ghY$t_)&ycFd;-J=MD+EM4Tav@YoN
z^#K^E{?oVh2b+j(ttslrq4DJBckoaQ2OFE`!ylQIm4_X&$6D_zGBJTyc#&L@_`%a;
z=vubvNW_$!+tuU!IjFV>t_tr8V5Zn$jbvKYcuZ|9mpQitMi_5IM=kl_pPHT4-AQ0K
zLg5HDl0iG3C5vQbn0|Iydk3R|D%NQkUL|HC*jhXl@TIapqvdGv;I+ElayyG9CjmWO
zqT_4M6RGzO97qUmCQI}eq#Nx!m#C+?A<l-rk)iy@6IS2>Ta40k104OTV$DYo69}PP
zGM^hS1UkcOP)}dhd=0Dqebp3-@}+$#+?5a8IjI~)b>5Q1Rzor&d?MJ<iWW7r%{ZbP
zB>=@Ra_rq?<(8_dDhGrst+?1a$dpYJXN25sIh+tQb2j))7zAK}I@2~zB6tp<?rPAK
z4f}U@cQ><hADlVHi*6|AMa#3H>ew=&KMz}^XuD5-ZDU?yVW?U{B+F7?=PQ-LL^sOJ
zq&fA4g%wXm$rB`2;-ZYACnSDLcs6}o8Ev`$9-D%yt;{lnzU-X5GpfU+j8>N4)`otI
zcjN7XCaCIjF>Mb2*&U2K87E7EW7eCKQw$00yn7v*(K_DTmi^J2&M0BAg<Mg$+0R(}
zVdrloLZoq|J<#-+{<waj`Oy#5Q!)a+j%Dcu6QRGboD^B(blqT-C*B{8M5u<Q@ZKSH
zG7m>hN<LQRMxloYB&|XPKYKk+n0nso&D>@l4X4zuW_R`ur>&rzy|<aItfi5nQ*`&D
z5L-^e871X<MsuZ+9Ga+g0Ul(u7T2hLnEK<#gvXAHb6;Z;N%5V9L=XOOap@!hFZcHd
zJ+wSXP)SFpG4{m!Er-{mP88@&Hc&4vvx>F=Y&Fv&!BD(JO^{eZVMtiAW}r(UN#4aO
zTw9(m)!;%et?VNGK&f<mgtc&@{&Gpf@?Ll2dosgwc7_&vd)G@y!sL>Dyj;#i1X`=t
zH3t2AbeRvK$Ym7Tvqor}RO>vOnSUQ|rW@5qS@B$4O=ZxB4&uV`9y;(iKV-wa;6FDh
zoa@soOLIcYw$c-gA;hHc`h8L=B#P7<@bft`!MGumT2Riy<&iB~3OSD!$<?Lx>A_It
z(Ku4TH*R@_&pIxb4-2$n_@GbxXCIhZJPtx@3h7Nmjq#dFUO-iUwV!;?+A<a9qL3Y@
zsN42YR70aw|EJbOk97*|ix*t@r$%f<v+A0fb&N6^^$^z`HDu7egP9p3t+NBKV#CiO
z?9`gOI5Eu+nV<lef69c6pyLN=_lzq>P}em$`m+|K2^~1Bg>G|u$u*V7s-78%(}2~s
z;j<AB!{PyopkCV*<u$YOouMyDB+}0v9`-v1{YuHBowM6~v*>GkJ-cGbuyv@jx!w7O
z;YYaU<I0Mi68^fskNF4X9=<C0uvJ!H)d~4~x*KlJX#1ilr#t9SWyzO(@VwUh?hhp1
z_OE{3s0Kbz7W{2#x?e29-Xg=-&5R#*D(xB|7^aD$!+uxFn>OZJNTGK&8>nGAn=|=E
zJ)iD6_YLC?Axq@x{xo;1ikBIT0o?tBn9X#+5;_wD9`7_kzWaJXynvh_tX+rdkpXDu
z*ovdxQJcGHv$A;}q4oMfp>;o?USBHb-xa=dv53)!jaa$)_ZW)Vbd+u)Z*Gf<d5W*{
zlE7SqB9G-fbmsnS#8`&!U;0c>RD+b;SRxU3J~3xld3?ol!2d<u`Q5}}sD~k;@NgOI
z$N0M^_Bi@T%XI`(L5EK5L(X;wnHe9r7Y`^+-EzhA-;1u}FK)opQ`3u<jxfSvF>McV
zrj5E>J4QHYKg^PxOAy147M*KL{}*HL0Zw%v{*RwyXYaj|QC5`f(IARQ#IdsX-g{Qa
z3}qw~3fW|@BiSK@W0NAX$~YwaZ`Jet{r=bW`(4-Xxw@X7Jg0L$=kxx&@B4MX-mm+0
z)2Fw8l|Z4ZwbHCWo7E|~<V2uenxDU?<>t$ku*XRh<~R&$L~eRv!`e7;SnsLXw4g^{
zSpF1s+#Eb*r-OrgJ(U&3uQhxWBQdm(3OS$Kcyv{j<@LGOzOLJ%Z-W`TFi)182W4Fb
zpKUMSb;c#{urstEb`vwX=8@T5Rw59fco5O;0Yb)ohmFIOkPOhmJ;N6Y<J^nK24SIn
ziLO8GA|Jc(K^XSvyjbad-XT7lXQ)4Q<IJmsU85<#XN0`W6XWF1D5CRo569>Pv@%;l
z>8zc<r+T90Y<sCuu|)zQp8ZnB+S-b>rSawZT)H)t8oq^%jR8g2qp<Ms@V24*EC{fq
zdfH#7P;6NDGXgrv$>StHM<#L-Il1>6Wkw%GeR=k$_ysB`CD@c<&(Tu(eW>NR(v-pG
z%SuAGE<Otx`E55KlB8cmua`)8#Ii>sG*%jEKF+d?Brlu$aEz1yj(=$sbVtuC1NMJ9
z8~MyX^!}(?!2nf7bAR2-1QiP-IE$(D?5~@LEBOve<%1EhW|}VZSRSfMQ2BCZY04Y!
z%+cUGVI{vy|4M4V6y+|%F##i?K9YpvQvx$m;#-SZr)=as{VwT59y&a;GPsd~7)81d
zCo9~)btbC_Kfz(6e6B+P<KVGVyq#wX1`b)0O9=T!pbq_&TyUX>1Ha`}>`Re(+VWXq
z4z9pw;rV}R>uupnXr{OK7;>m2@ljb7MeG9k(>e`XA#}FA=R52cE%=WN;&0{nSxo#$
z3?fuR1jzyb>;dVHNWk7^yXo<6?tU3PRRaA#R$%BrHpcRL5y1+`*#bfD))_{Yg>x|^
zigb-0H^KxN;{$yb71c^?^v@+@U+d}K89_!MQ*g4l8Kjq1%JkyBI>P(kTJVVYcPH}1
z%Y!psBT$R8NcQZ*c2y<Cf3(8#s~|!L{D7_J?)~GE(93cOeF{B_=FqpTVeHM9{Pp4D
zfT1t2mAoNZ<u3(VoL(8>y3mx3M4xZo`$BTQ5ly2ZD@v7qXSc1tnTaZ)I-lT#Z;Mpf
zs6wWC?|)61C3n~x9|+n#@Tx8(JaQSit6PkBeadGcmcr6iO~GlZ)-MFuv(Cv0lwxin
zB;7xpKw^gR@V<QkO?~zmt|7&1+J8t6#R_*;agx3ca{0K8ec|Zspq_uUNcW6whK;{p
z%Fb0q`J6xeDljFhhC|)Evm1%`B2_N_$Gli9b$B^kq-Pv^SO2}C>^v9soqr5U?)Pon
zBb-KxMiN;<(k{yg#b3n#Q8D%XPti1}7dZ->lrE0y{@g>#`CWfkT4hX%<t&nV{(BbE
zK6GQTNSQ0}!M`V8vBH}y`DMa7nTbcwcPezc+|Zc>dG&v!Pr~XUb0m&2p<&NYlF+44
ztq^CNW5vqXk=sQ7eP#ug*X8W(csjc2bB0H(E|M;ko0fC-Xsx<$FI0?knEx?AYzv{p
zb2sJ)F>lLe@$ccN516(n{^I`cJ*4yKYjH?irsDDC7I9+hR-eG%Jk9@?J7T*?Tv+OQ
zihgR3@oal|sUX2_x6zp5FZTcbZbc35CcOwABchRDmA=FPPmQpd=r=pbgGd#LVM;Ku
zfPWHJr*T%v8kw7q=v+9;g7@lAPc4q2*U*pTrW6w7m6`rs<M<?@kKETCe-!%6G(YVW
z<ucWuMN7C>TvX&hAaZ`Vo|EY4#Q3JvN`EXFEJWym05JOAq(G97R}y(WCq9=ItGF8H
z^rjH7xhI=cX5^O)!qA0K!gpgA-&t}cp82!1Dy8(X*OzdW3VLs<s>YLs&DDbz=Wu6#
zYsuTDrK+mRJ+uc5O=3<OPmWc%EArpI^*Dr1U3Nk!$rqo?NHWLDPcBQG=Gy;eHm5Yl
z?@*5svU$^bnoH`%9};f+9C$6=Z^Xv?Y<NR|kyrMagR=qxEXe-@blmi*9cKsU776dW
zElS)QMFhM!4d&O5Q(oGdD!JYK%P(p*Vinx`4~><S+9SFxLH4+6Y+&F|=0aRHDCKSM
z>iVM(4xf8{%0D;Fj!R*8vz4Oll6s%Jm-10RDu5=&ctQQQ`K?kly4S9?-Gw})jLXu9
z9Rmv{M<?a`{9K8;R*0=oyt22rYc0gGqp#$sss1c$?uO(_ehx2AXA$UeeR|Avb#<}2
zbIm9BrMRHr(WjA-={s;40-MG3UR6SyH&<{r4(&pqyEqg+vlN#A2%3nzTj^fAi9D*|
z%3KNi;0!TMUYlDSq@%0Lk@1IMX>u2ai+=Ow&B|`|?o%tg+o_H+OH*Dm{%;BkoeSQ)
z>EXS6k?zHd7jy8EH=b*bUz%nIq$1qc8}8n{%V*)*_+>L>7$`jQuI%aD&01sl&u6Jh
z4TJLo_bd0+7JlrVa4qb>@|BdRt5cR45jG!O>8}f7v#FuG{V_E7qKy9rfUiAQpzdDp
zR&UL0a0H2`Xv2G3yEfX+4_Ctb#o3sqYpCA5S>muEKPeLb_NB9)AFYegCkCoh%L-2+
zZf!HcEyYSM9cgK5x?XeBv$b7zf?Rh!XRsV2`Rfbi6BLwKTxk`OJ-xjegmL-#`Kf7!
zUhk=tN#hi`?3tOF!vzg%T^#+=L97CMj{6?`t~Lmr{-29?F49LEQ5a%=c=K~HHR?hK
z$}L3Ur5*HFj{ub(jlrr^l@zT%!E^hR=~}E7oK;=VEq+E6BP<|LsH#VgfF&Y3D<qV$
z^4$MI9T#&fJuC@D7$-*DQQ>~*d+1xaXvP0&Y%G}ekz#&90TZIgwz%{P<rHvV`p;uq
zrLYnkhX)o4J6{5prVty4FZHS$NHEh<H8t4gmrb$oOuF^6je;-Jz1z1pZ_sxhh`S_W
zk4TrMS}JY4xi$f}Qru~&^xEF~SQgZC3w@~fc5+0@+Z(g^TJV<?L8SCj=Jw3-i>BD;
z?)WnYJSfxNi<Oe>^W}D-KOH7aHymXyXar9@OAo<?EdKHe<)^nb3D~1MOH-m4?|m!!
zP8nRuPXA-aCBu$)&9SkuJ9?|7LZwJP@9e+`E9Vuva@1-H#-oPIx6l805Sl7mhDJ}I
zBBYtD$12>-pPwI$<;2VXvR(NZ78cf{$n`@<j5#CG(Cd??@){#vYc{AgJkeZ3cLwkO
zFRa8{DGJEx1S*A`7ZQ>ilJa&HW0pHR-XSGHl@ASZGfNdRLu9}_o|(%@9tFCY!a?xd
zxo*4FV4$V9W*9tQO#dn~hK*Tlm6cm%fIrt1T$h{s<xp_qZur{F$@yBt0dsUaIcZjN
zdY^yMn>Q1}0lzfG*Z2hlqTA+apTcttt2F!ZV+gjHH{EOIK*iJZ68aZ;Km^%kBf<$s
zya2+9VkA(I$LyEvv<0KRL}AtJ!8anHylD;gPaILirKSu=W0`#256MFqjL#KNoL(t8
zI6aiC$TdG%nepeOV2Dq0DbBP1kT-B2RNhB*X6;c(kY6hcqsmO?sgzko&k}{nG-8)(
z-j>2m>H)3c|KPf2KR^v__c?20H{+ZF0}nx#_j&edN(zU*rsj{OoH%Ja8%jI{?|vh7
zNz&_z++*Md+H`s>kaGFvmC1Bi8#uUN6>0N7uGyvo^5xOk*x1I(a{(a%0RxUliudq?
z>HtGU7);K?iDb)4(|NsmkL3v!?w`5<-yqm4AWqz1$$%azs;H=#;3UfQ?mw)^&i-7o
z?&A(=Ec;KQUuFpqV-|9<{R$|goOmy5b5QsMkYzti(}DN!M}JWJ4%i|0`~Sl^Ou=8^
z5~;GpO2mq*hNJ(?YiBd_dz{K_?!TiLl}<JaS)xXgM)FTf)iPXxDgT|tuzB-r(#>_B
z5dNiFWG9EO-QI_xRCVhQR7wJuRgEV;gSYMioOi9qJ83T_Iff`K)zffg|9RUlr5Pc?
zNS07k&(E<?IX3_b{jQhy!R*yp;oFP*80al%AvMHx7`fiy7&?0%%)q?nLJ8eATM0Gr
zb#N7s1YXW5uh@?8Nqao`6czs@m~M@Qkd%;`kepD(o_gibSLpYw+p4)a;YbbK76ULT
zNI0=51Ea7}m$?JfT6mIyO%360@XIHQdMvo1TcN=?*BXk)hS<lDScFpDdZdbR-Y4;s
zucVxU%)`Y#EM?IeH6m_|G17kivF$K%#u^P@uHicv+~}2?wa1|;O>VI`a-qvX`?)m^
zy$*cr@f?IjIO}+CE;?`^G)*+*gO7|@j~ZEewe?dgN)$E;O%Q3$YV_uz&2<=3_C%Hg
z#MjGY;@XyUefKMFSZ`b8DgHVsk`)!XIRe=lWTq39LyWg?8|QixheU@6IJF$^zWe;y
z3=S4G@YV612!d1=$3;3!uSBkym!nJt=~aarNJ*Oea(y4;2<~kvy(oKnM2qk?-C<2+
z=;Sj^<0!T?V!$=V<uzy07}_FltLQ;okD#FHM*jQAOC~$S`iYm9M2B3-YZ)Zz&)(#w
zc9?j3>e<n)j0iHaed1kb(E^1{h9chUer4IldzCm1Ime|>JoBtD@=1fJ<j%y+op~`!
zpSM;tE+WJ(UwVR0PU09xPOxi2z&53ej=*;f1pUbu!ls>KRqnk+6IC8PN-~t3g&UK%
zW1@s!s*Lusvu;@lId`sfGX2zC$o$%rj$-cEPzJ@HWzoguKpdFw;1ie0tjHkvaFk8G
z2bH8h6TZ#R)Fd?|og&(E8MlQ$P$7oP(lq6XQ%0&QjgiT*3ET2z5{|P^w$XfRvf&=3
zSJ26qSqlrO2OPZ>f6-h139s^RTJ4BQagQEkR0_h8K{cUQyYP`>1zqdW;alZXeR(d9
z8z1CcpIMs;W(P9VABHSx-BY-}b^3zPIf6_Z>)Bj?2Qpd0qV^z(z?c!2ItIps2B$~c
zLVqz}z}-h9oQz>cvb0EncM&SeJ2^YQgqI0&WlL54rxUl;VkSfN9N#{|)j)E*7CeyK
z<fZ>)e{La{JiSTj_`I97<gT3DDc1PFh<3CS#^Lsc`L%TUKFwprU!;F8y?KR@_0g^I
zeGU^Q5xNoP$NhQ~MNUiAv0`Kwkpf4{1{m?XID6hb-*wBFbo6^VA|nECW?g?v#S*t8
zO5dfT8-Gjn*P++;jrsvPjJF7`596Lpa+Ny?{g+zVsq2dMX@B;9V*JgrXE?zs21ouc
zm<yvH_Zj&lXj43vGI{(4xp$`QwB`7Y*sb}AMCmyv6Z46mStbwN$0V0NgAZLEAO}%H
ztgeaog|WdsN|rlY!4F!KIf-O9sgvpMs4p%%yjo(7zV%}=akA`PHJmm6Db^AKJWhG>
zd6-Q6M5044Gr72Fc;EDkEi6YU+P6j#%=z2)X1=phu`S5FE?34{oBwd91?Tz(7`Ry9
z1AFO!4gXn#D7t<*yxt+3Zt1z38;8|d0Y6U(=#~VK_mbAu#9}$+a#Ba_Ux{UZxR}8H
z@=u|W7~dRst2ix)p%K|I8Hn&I_jRVgs`^FToKr>hJ!r+a<mXBY-qs{<%zU}FkM>7b
z_1CEYL?UJZEcE$@50bLILTt3!ZFK_rfC5~vUBgBDEBqRapXqwN@bIDIts^ce%Jvof
zSJuyY*6?MA-KVLR8=e64=HlA7Ia{nzTt_X?cy>thn4lUr2zBSq&qW;r$953GD&Pb?
zxOJs5|DhK<3)dw(D9=-jT-!6oSiJw!0?--4>3%TzWB^+_Qsq~&KnU?%C&pv+T8L0t
znujntj0F(>5bnRv`WBCJZI7ZK-MG26ET$xFCg-vsP}=x2!F}Cy?lpj5mdSCiY4FY1
zcc8>Wl!EKw4u^-qd7jz8kdfO32w`=D8Y)RD0a<>oX|=yGSUCk}15txRVfK*zd*~qA
z6zz#26@V`_$0HbReIc1iJ7&b(mx(TvZdvp%kdgOeP~8b4E%bf-hF+4*r#?@HhiizK
zqLE2nk(=<};_u*yAX6VqQ?13{O4@h!P}Wv-`|h-=zw+hDlH=e9jsN*@Yjz)jfXKHN
z_H*yR2r>f?Fp|@|mA6x*@7=~lm#eDtmi75}d~3k(W$`hbiAdxOe#`&#6@V|FnHFVU
zJwJ2oy|HQFawYy2&RvT|q^X3ppm?Tz*xP|%z*#b@($DPGoXy2fls;UVZUtjUveLrB
z2ISt9Z0pRy^mBJY<(xt5hB^)|-@j6aZe2p0PLpfe#QK=Ulcf6P<#fnw!hAM`z)w_6
zU)L6X(d}XtZ%LXxO9qD|O(3{NW)6BAv5$S3Kg$92;k@3}tmO9}lg1X`m-$n8Q~FKa
z>9lDCo}+^n-iHq2;V6%%pI&@t!%$MqzN@|;+VIUJBnR+<+uTn85OSe>Lt`>ayLL7<
zwQq`wR&7k~V~^@V#n_i|QqdI$>fXJ(#*WvT%(k4OqG1QkRcwH|(#L)Ntxi4l9`U9{
zi42#oJ;|Kv+w<rjo^z!7%?P^gQQl|P<n5D~W`UrTORvwT0xzy$a+@o`#|PS-tidM2
z!iSqWtp&)bwGh>biI?u{d@~2<y|2tF*59SX+d$GxPW)vCx|Jsp!Tp|m+@+36ldZA2
zS5TKNjIf@{zfy;7eqp^iT=+lU=}_b;tCf*12cr#rG*S>%hTs&33hpA!qxS-!%&rgU
ze=b!_aIX<dfp?`WSLPS*(9=e2TDWy!+`o8r^w8)@Nw=pmw{vly*(6?Bd4n`p;Lzc@
zfq2~YzPrjpBmb*_xXjXW4a_8bH7Gn)bCusptO&ub97-y|U2hF%h_tA{*{jW%6y~|+
z*P_Bu&}v7pQwtssXL@^jo>@jnF7e{rJt<>(iE8WkUw`HLG(`9pgZ&(ozbxf7ziDkk
zmp0ro@U2_@#=RdX&Q&XYiU*evW7{h6xuwiRtYjauBO<QA)tna=7Cth9C*mcLeZlfj
z!-f-F*>qo;`ntXT>|#=gdFWb*{+u&9z$8#uhNmv?4=n6fjxhC}KE{n9@WS!x-+bJ>
zratGYpJ~Y#*XKgDR?dFI*n@fV0=>)k!S989Spjaiy9cg1Kq@*?;tj|w1?fTUr>n7>
z6LIT3alSf%eEZRH3UVo%WiK%E2zgR<$WNRfV|(0p6({}l*8{X=f87xf+*gc@jNbXT
z5Y*D9CJnRSG~T>6hd(DfM=SVUQ0wE`gML<S=2&4|d5Kr{V$6XQspDyZ&Q@_9CFZjl
zde!V0It&>}oGQoHzk;0v@$Jq#=tB4DgiPw_c45M!hF3fNb>)}uW1IO;#0%GbH*kxB
zl*DhAnQ~@5R+?zQ2B|I?2AlAVkDwWG9Oj$irIz%2`fXm9l=L?NWqJRk6{$Uv>dfeR
zc_;h#uou%3!w2RJarRwY?bH6oU!Dt`nzYcOtU>guO*|lLu>)Dh%z-PW-CZzx0U6WU
z-0)Pdn&pv^T2P{8!~A81)meSj%4mxVw2a@HO5OUaF74R8xq>#?ILzy>n^0r%*-&7W
za{JP(w_4dm|FM~+Gs3&A{NNd4&lfVo*|TRcwIJ#GBz-0+RJrWsUrx%`lFvb$PGc*2
z<qrzZk)`_o>419q&r(<W^U_EAZUmYpud&Cpn+3H;(a}^mcn1p7Z$SXkv;6tqtbF4=
z$b#<P^)wX!QzFH-o5|Bh2TG)S!7C2J*POau8{N9qw^ufOZT#)ljPg6L;J(*vm!Y78
zH{9@gF<pTY94+D0(r>WYn@+J9UuT&2-ZTEAE2<a#?GV>fl2?)Diq`NRw@g`q1Irxt
z)Qq1#2uW^hxunJy|MF!A@;uVnbh)d*JA|YC&SSpC7sgy%#Y}iHgdI3~_;|03WXmE4
ztUtt{;h67Lo(%WC(tjxfnP+Z3%8|3b8Lk@YY=VCXYQWY}v^!eANc?>#=61G!d9?Gg
zbTDz432t(qs8j$gZVUIpY1B~fUrO>zPATqEO~nz_9D^J%8VRGMgfn=VcyWaC?}cl<
z+i<Ej%=Or^7W0FfZUu5v@6#5hyHn`&F5aprRXx-6_+NlvwrOJY!^!B3R0*oZN$+0T
zefu>!#2{qj!KC5G%ITFIYD~-}uB^d7G+?(f1=6~z;t0=S5r-<lP=>y?@aay?V9je+
z$9M!|h}F&$;ny-85H%9xO9<o<yM&8o!*=V)Rm5{5R(yx_%9%%zV&^(4`t^q*$;_$J
zoqoB@^`fqfK1FZXJ`+{V)%us5^9cl4B}n&Zy!tC>YdJ8mS6Ie`xf1O)SF@XV{E-8I
zoqwja5i17cl2~b*d5Q_6<<BIj|K(D?6y99THYF0jc7>{pD<CrROSC|UcW9Ow{^Qx!
z3h@pc<p;eKJ$f5IJH^qh80)(X!Uo8g_feJd8e8XDl`3BUEzue-&0jmIJ`k(q+6acc
z;>k#|!UKN|Z~h2Y5!&7%)V1M{Nuu-h7A`qwr%E^tC7!RRo#G+FD&Xb6v#K0)Nxk=%
z`Geavyz6^YZc{q>R}<s(K25@u@2Nr(F(wRoVQ)FB57U9sh^`u*I(aZoBmPtbe(311
zwzh{LJuQDn&=}u$LAk;9!*Oa!{8VQ41|-Ywp23Ui=y)FA%b>zJDUR;q7~?iaWxObP
zzuxGWUco9j?Tg$L^pd!($gL}z%(h9s=`4tt)D2@H7r+T5qn|>t%kv@`OvjlvaIz3G
zNPn{|T{8OTM?{jRFy$l%+KA_IBUm2}lIkKvPsmSU*|;x%i9~LWL$bXyB8+>~n?iG}
ztK_pPCa1gz*I&4#ze4XzEf+)*skI5eC=8V-i{9TsZF=4$KNz72$6Yej`nn#CFH+r6
z!86kq_KDD}&2#<z>j*Z2ZHlqt7X?Qhp&*Dmg-Y%+Ff}F)4Mh_;@Rl4I`3QE?FlWOf
z1mx`1IgwRGF(cAcq%8h#a)CwfbK-q-J8~mes1iO3qZL94%P)U-TfP2M5ln*T%H=k#
zN~AHd#G#et123q0O>EIPBxL9a)C)c-i?5CjDAO}KS>|%|v^PzVk423zHA3Zr3D-?L
zFnLF)>KmdiM9IjrO_6{2d~S|4weJs=%rQpqa(7VtZzEMS8mL(AxpI|J?osF@Er<#-
z%8;<s53;Ng1t3XgirR75tLDukT+)v)Wo@|<A-7P;s&uz@k_b$hIE-E~lUO-CBGWn*
zau+Sl8P_*eVZ^|$@iXMV{7k+&+w<!j=7!r4-rU3W)yI)4B)jdzXKE~jZ+YJLBZ;sy
zSUt7f7Dmev-b-jcmUcfx6Gh*i_@sKkG>v_sT$%}4Mn_NU`-bFT51BkJK}OQ(Kom^g
z5!K>X@1+v+1xcm}H6zrWj)EqxqLl6HN0-dJKkpJcb*@K+y$UhL{av=yaE|MpJt|=#
zoPLzS#FN)Ax)s&Y(~Q~p!d8ww7T)#u!`aPjz3@Hg!APshn~rdr%qTLWQ)P2qfvEDq
zl0?LM>yg!y2Z4@w_s9;kaALl{Te!h~`_DPc&p}l*5_Nho%L=6+V@T*CLBF}@8NaQj
zm|K?~k{)^tauISu!#KLJRWeo(381$oF{|9y%9&ZU$^>I-%HQQ|n70v<+?l|(y^eH5
z<8X94#M=q(eH_s-AHPISlIeh0dFN?h0+3qiKfEj<=;$ls;D<32Oj@|Rnk!Ynaf`30
z!Z#hqP|~vJ$pZ58kG;~e?V~Kyg+-;GgUJ|=+)h)5;ZhU)i+m=Y?7U{76C1CpCRwAo
z;_-Id!We(Jj4PClm++b+?{k?jC68YqO;6rAHy4ds>`XRopeRRp+IcB^`w;A15cgLl
zqhDo=J)HKbPmdRp`>!f!>FC!|N2>=J$RKrGa=azU(ycCNYH4XP3a3A?{Qf-N=tbkR
z^pZz!S2{5+k?{CLd$>AajOzPD(W0y*l<&2Yi`$6{;}A!;v|i((MEpp_?|0XfUH)n?
zFJDd2jB6Nb>Ee;{6m=S-Wy^oP31^Q^r_ArT?EEFv?i}AVhGE(2ADRV#6+1HiG2>un
zgK{nf^~UeYM7`N+$qy||5ovK>$(ckFT=(fv_9xGWs!|tQzxwQ<;w($K$vUir#~B1x
zm`GCm1@D^^3RW)sa9X?67)M6O{19JP$4u#Va3sDQbD8}@EXE~!$zWVJWM|gfJowdp
zv1pi-5&0!&xQrh=_HyfK^PHFdaVANbHdf8MRq1Cpr-Rqya!d?c(=Ae#QBf)ucB16E
z3yeNgEIO%;kK`G9_ae=hsO>MVI%{EG^Sv)$P5e5)WnO;zu`|xbN3p;)CNZ*{8<*G*
z*gT*}*NVgU<BmW3M~MA%kfK~qyB}8_c|`ij8ngZ&bM=Ww^dpSXR5h*n0=9HzW<AoQ
zYZQGA_4?|YgAkJCnbVAN-R2t?Lr10~fuGz*n+e)%8l#fm^UmEXBIxKxYs2HDefVu6
zO>3f%G5z!I86ZEj&8rlv+b(D%@*)(7CIh1^*8*@c1bb4uzAFFo=&NIVduZmrhi(^(
zZ8lpN2>RUp6zAx_qF$;Z9&WCO&@fX{Qg|=?z`HFfzBzvSP@|=G*3j`i4#QSC_FinC
z4vx<EY`pZBxrwjCJ_dJ$XOreFyJPqrumkMNp6*I0E0pTRnr&78L(f&2!1I>l%qAQr
zR66$+|0k~8be~2pi`v^BOD%U)OHeJl>LO-4!-aG8@use7o$e#?t2UVKF;|kZA4=vc
zLtnR?_&FA0Y@&xsYO~}pugrh{&G&mbdRSa6i@%zKep|(ezSYX-=7#n2*JcGkYMuH^
zmd);TR!I1%M%9R6sdb9Vf>|anWuIf?QDBQEGqve<#NfL-6T5i42knnehnbR`!fEeH
z6W=-H9G64t6u)4tC~YCr&nrFE@bIK3+r2rv`AqrWMy{d}4+^x~n|P>TOsI~R{54x^
z=<AA2eebk>t32Uv<M?eAZkcurzP~DS`Aa5WKKkfGR!m{T?naaK&dLZrE&-_uG?aUf
z7oD+w?%I(H<oRcmVD;OJ2fknj84X&2@Mh@PZwK=hM@b-EM{i}x2LjEv@3i6t?$pO+
zC)C9^=*(IP<J?3P>Xya<0eG$!%x1jUC7cXpx2SSS0u7-%c4qYYe;X`fdOVBJkmR?p
z<#Z1EP7%>mgWT+;WrE7kW%RlQxtkbHN<w=1FD}PNI1#wPHl~O7P%rx$wNR+NHYmuQ
z($>C$_-bZ;e<|x&UzVL?;dwL#v`viHEoB<GPT1yIKD)&u0FXACnT3T9+8o#uZZ$yn
zX??>+v;Q|3@gnO1wP#^JnBrbyFx=dK|E!~9@s(6H#w_{>j_p4$!$7sw5{w8d1KyRj
zC+D=-S>^n;6Svo1eTO3@FFLoo|H8E;%C-66-B^Kje9Dm6!y+rqWww<vCZ-Z(Kb$+=
z(blH@G&<_%Ab5<AK2x{)LPeki9agN+ew5LC|FEB#gq2EI;jDF&eLMJW4%^#!JY6ea
zZe}SP$Z-Mfk8vY-JrWv%uaUROC82P$;}+w^tx5~2A|jiP52?^@Esh8R7tFhQ+qNV{
zM602fTfr=OA;maoKMB1NI^p%U1wdiLC7rf~D(F89lm%xi!FtZ=zbzcy%EpBK&X$K~
zP_5F-a(3lcV=KY7X0aiHl*w%sZcji544(-;|13=N=OiI&R}<gGb)r5N^SHhR>yhWi
z2?yH&$US0lt?oNTGfHye!yoR=`@Q>0E^ygThW^`gyj6<T$dN`LnmwUDdozy*ZLdA$
z7)h@4d_M}fWO#g8W_OUXIZK@^I$`~9@6neWf@gM47AG8X(?~dtx}>QT8v3J-7U3dF
z=fWMx@4b61n;AovB|wmzBqj4wA_}o&ij2F3BIAL)7i`v2Nm4ejNHaI?5*UJiQMfJY
z#>S%&fk&oU%fcw}t}3ytac+a0a&)(hSRhAwS(-L^TR21U^7|tm4ZOciA1Va@V7kc_
ziXA1zTshff*D^{U(uUz#hA<Wcp-F431gIJ>iN~1Rn-sdFzeJ6`|M|r0kxK&;$|Og`
zfdgK^FNpH+;5(4unzH_dEvM=gV;bx}RL%VO9|vuIuKNJjy<8^PA0cD1JOdmw2u$48
z#xS0DjZ03w6)1~)?RR6tw9Sh-R1>w^Zo?U8X`MSE{hO&ZOpzTGjW63_>M32_i9^fQ
z_f4N}y+&je7hUGkVqX4|KNxMboY<H84~mVL6wI`RD3~q@s)inv(+nAvHw18T)rnII
z+><&x#@783M=<2^Hv6Wy2!g~#?HvxMbFQOuZIIEP3fnVGBut5!ukTm8*noqg-p!#T
zsSY?r@~pG0<(vZE?%I1%Zi;_t<1UZ*&S%P1a$_p5#Cooj3~H(y?#y2E0c@L&zpON;
zQp$0%vrIxG@pP`8!THTpSocF*yzql=`HOeg9WXb6O=#v1n4u7mW<`o4c+g2eHzuI1
zzTQL}m@<V@3C>lJ3P?+{N=mI_f2p&ox~crD|7JU_Ztf!le!u?g%#~OgMAR5UH!h3c
z@M1vQKSzSX)R$n7D;XTUd*7!<k`bnb;0yAH`l9ntrUOEu@#W8WjN9xS8$<9puA8;S
zahXtUDl%-ll5M&R=9e4P9*bf=@E^C@-B;_d3%AMdvHGtf8N*RP&JYL7M`h4PM}A4q
zDEt2Ai_#u?b1>x-a?HF<-iwR_A9MRDqW4)U@o$dg*dKfj`6?GEw@<lH(|_?{NT2^=
z>wxMO@G#O*JDtoVNx4VFLNLsc-t_A9)N4_9t#Q`BlAe7V5uJo%+SRy8BeXl_)OHfe
z&RyIP4%a0z+*8bLyRdOg7Z<aEk9<FAhe`b3bs!BFeU*SO;jU{2)=qIw*S4=D>Uz$Y
z=XNSrVhUunP+3jF-&8(y7$y>R2ja0}oDOq6!7-DzTX!lAlM%-q#6caQPsQkP*GQFz
zA2Y=FeW;+LpJI%(^mKS1CORC_J}*74_E&7RpHD6`Xh{0u-TLngJlV*MqJG>U`Vwbo
z%p6}vGm(*zX-POvXe-!!d~u*#V(5FyD|`C%|D(8M&>Wh+tbO((%`fqYf2o$4<`2U^
zH9FKN-yx1dn`lk!3IK~G)MdDkP2W4y)q%J`*mwpd-c5zdi03zv7Z>COID#QeXBUq<
zG;&2dU)YwZYY|0-cN6<zvGsa?`@nu0;d*&HpVC2Z1cw-eriqo-W}R5Lo<XmHMY{~}
zu~i6=u_$cL9N#A=knHK{xsKoiBl?nHG3Hp@A8_y+o#B2=BFpYnt!!#K-^VuI66xUe
z+7yYoVcVO9t4=tx`h(pw>*z9*PD8c=z3iUQrY!xp$fP<>8>IN}oYQUf4i0qjovK_3
zDpW2q$)Z0I;vaaH?a`tr(P91~)Q&F(i%i!1DF(!am|@GQ!dA_AvgjMl5vMTTwMKh>
zY@3wyCL3z=TFV4~Q)sh0=2Mw5et*Z%^TJS;$zh*4bm%0a)jVEqcRZsN&Z;HK&e<SS
zrQEHKA!FK?ofc)uaIc|fV3?J4o@tCRFK@!i$Slp=duyY)W}V@&zN5$`>IrrZ2{72_
zlH~u52u?Te{?lMa6jomBK#jl^5D=)$7%JnqjS2O-k2EXC$V1j0H+&w0*qqTg(J-wY
z=Bhe1r~-H1aGvwIJ53{KM_ya(v*$fHeE!(7!RTF<d8O?b1|umQAkCHT1zl%osR(S4
zOZDta_mYqPv!%GRhnLZyB@hV#>4F@5p04$v@c#HJ(2-Igr)b1ZN}lu~$$*r7)MA2{
zD6AfiBVOrkziRSA9zDXB?C*@aOMpt~8!pREo*b45V<nfQOO2x=GI|&|wq)8G9&A$i
z%HjP5J^5R<HFWeQ`L7njaZ`LYqRZ#5(&>nJ2!Pxv>PG@P0hO>~;iW7@?=vp9I=ts{
zD1!=oTwJ!1M|zf!-sDQm6_iOJYOgt9@3)|XPcKP^&-o#;f8b0zl1x>KToPBN+c1Jm
zb^p>P^ID_(;`2!Yl>WVYLC>*A%&U0&>tJHvmkFv0KQt)34(V#0VZ4;Ny`3Ew^k38f
zj+h}1D112W6g@qE5Q7c6bos`OZZs%}$B_3yG13NO31m)0jBRHUx_*Pnk`tlfkWLo-
zp$`4f(Lr2_C~Rp0!<!RV?cOisRnnK6HgHZ*@VBFK95i8_uNa>>7&y9N$cEQCGR{6A
z<(=ui-smui>?FM665ubPzPxn$3MFgeTI+N72I-}WU*t)hA6i?|6g<An#<;KRvf09$
zXc*`nveSEfvy(8p_HA9m;D~yY(MvRdZ1=#yy@rjR(kJ~jGt<-g0rPUfIF0gO*Q|%c
z4EgkAx98Q_k4;Q!-GzNu^#xVRaa|~PedAk8+=nAdKABt@^DnPIr&Sg*QH{Ri!21-B
zo=CX;bu$~g)0W>v5P?~nG3{K}U-hDjynb|sMMpVKj<-2qqMq&$MVQG8Y!JuB;cvLJ
zfhTh65)f9ch~$?mLs%^V_IA)_Tm0Sj3LJiLAj%=6W@ep>K0}0~tdDH;kzMCXIBCMg
znp#ERf)oUK*#VBwta+&)>00$Y^YoR8N|(PsjqLA|@@~ck(TU%aWyv!i7M@A4Y%mXX
z?XNRv#Eu6LDw}ziESb&P`m;si*RsF%o$xi-MUelTQCNTOe<s>XD<r28J0Q6<5g0M+
zuP`d*{oLKD0DF`s7$?Cf*(QYz#~%F<uOXLO0;w5(e_ebNHva8&cG|I5e|r<w+j@MR
z1*0@Cq*t9Dicho7>3-{eOd}Rf7QEVz96Kv)YEs;tj8}26P4v}zBQuAVw#KVwH9YR+
ztKl;yV!I7O<k0@E-3b0b!8Du+D%%3kYxqfO?A!A~LM~7k;WlCuIEhy7bEbZKE>rt7
zu)*cv>XM;3C0_b-cd|!{*%*Hg*Zxb;ef)%*`q2?+j?3IjOj3-zq-Z!DDm64U2YKK{
z#-NJPxk8z19aXEyr3q6CW${&DYk^YDW&@fHTN%o2olecFx<V<z1r5t>A32G3HA)Vk
zpL?+sRCsJ6!oLN@)Oh0m<;ZVW=Gf;pJlq<+%*_`-@xq*^p7Cq=%jRPU&h|}M2DZTM
zL4%HkACJ)AU{ULzSRpS%A7%)gx)k)+@g+g~l8xqnLUi&=j4CHeG^5u+|C=T(<dt|$
zKA833<lEQ+Z+^frw*b*(qq;habj=EJU;3jHvxpN6^_$hGbDzkBwwyIM#w}rk$(RRv
zZyVTCpTR*Q3!IBJTqJ~8=v~%Oi+%d!$=d_<6U|+De_hUV_bpJ@O<dYO1r6eE#!FMK
zpznt{1CEwc0k2=bwhDl|z%+BfP`NJ*3$U`XLehxc@SKzR9gYpUMmTYcFdRamg^3xw
zj1JDVY`tG_K{1De8SxKp1KpRggL4B^Z@p!g+s?!196i2e+<><Kt>Ev?&M|g(_W0Oy
zy{XcEi_14U38Sv|S@PrKkF;lb97Ac%(k?lyNq~OqNwV{~wfGA0+e1Hv-d^6XzWnRq
z9h+WieY6`5b+YvB2|-5cM6oz5&x}~LrRuXmLF~-d!NQ046<m7E?L#re*sUL_5z5#Z
z_YLJU2AKzG?(!TdlFz?+KaCeq^A%<^(_TAc9+RTRn-V+GVzS+O<s)Y8^~Qn5#G)^s
zdsDZ?&nJQAC1ftmOTM|o%fm)ir$a=;wfQeCyPCH|a1v41W?_Y&OxZ}?{G+`P@Y_n0
zPN?!?L4%9J&)og>hgV-d>9!#6bZsmyEp_$`JOHOl?}8=Ybm)f{!E<=mlolAa<V#^-
zUs?-;$f<Y688erRtcnkjCS%10SZebR4F|t|Wg2ss$LO8fo&_8G*bf$z38UCkb4j&0
zUFJ!w1bc|hx^Z&5aEH#3nf$Rbgb9?-g9fv|Mng+VO3V>YZ}QM}mV0;eq&Td7`<D8)
z7`wYRnX<ICpJ^y8;ZS_()tz$}B_+W&C;LPmA90^EpPZCbP%x!8O)@TUU~A>#cQx^J
zwwtf+oJlcs>mG&oGz|gg#}DASqyz3Yd>7d244uHX`20+UdR~@|Mg#bVTOJ-BMuWF0
zk~4}c*u|x+&`{X=V%mqYjf45FE*0Snb%vH7vCg@{+c_MbdiF9d?RPc;R?g==w`B_B
z%l}Uc@b`um?%}LAb(7^}YAl9#lh;CMpxD{now>*V^yu3&`;z@aaIH#*!8Xkld!NDb
z^X|!ATvq^Jd#~ww&-3Fj{FUu|(2wXzxq*YCs$Djk40kZ`j{u{B<7uD{nl`}5|7a7`
z77ggFMvbwB0&RZpygLQKMsVmwBH@79Yy!KHlHRSxmWv$$h``T$ZY62)sy)pFHAO2_
zhKCU(p<|zuR{}`)azq#8<^4}*isDaI<kTPwi>n?@cukIv?<=~FSzOQ)rjM=*kXHz<
zz*DAGnn9z{zdfYos9{|`zkjqpVt0ut+?09_1Y{(43O=dODus%6Fq;d8h<51P-zrBr
z7{9d{4E1xveNkNO_ywjw20NITylA>pz<sAc+Y<cEt~GCDRlIkCX+ej4>d(Ik!iPa@
zV`FDuKMU)q#(Por*8zO9Wyv@i#OMU~rmbP2o~`J=8|=f(H=7<4boAua@UrgHo3ALU
zYigD#*ku`UvJ5ILx^E#)?1^^3k~PrpLfUT?XXg#aS{%2HlUr+>B1uY4-dxnwbXY15
zmh~XpT<A#kcEXg+vxhXS*B5@AoNOT2>z$dIiACQt|6mNx4)YxSpRn9qLJK*yzb-Qw
zMkJKKsa0n+&xk!m?xG}jOpD66`KllH-dEpyRhd5dBw?JaOibNsUKMAfWbKQ8$ew41
zZ$;E5GJg=n7;cIKt5&Pc|Fx*<^{yu<cYCVvA{aW$b!R0Oe`$JqV{K*ja_<zZfJspG
z+H6b965L@UQ$~WdwKd!{wZ%udY6b?Ax64s6N!=DqT`-8@w^nC7M~grI`6nObdH%RT
z+?DkXThbVJteMt9&KUd4cAw0lB9&<R?Vab>uYY>5=HivSr_IgHM(7<fLIE!8bi+zq
zn=qWs$}0P@b-3puZ-kYN7{jmdCZ)YaJ6uR3)hf^~xKohZ^VMwvdsvn!SimEneKzX1
zWZdDeT~4B_=y20i72c$;c2&<L*Ob)HJiU1r3YWT;+h<u1mqCtuuRwdK-K@NCt5Gtk
zn|7s}+*}Rsv3ee#k=vQRi>97Va8ht`Sa7Z^=ev<M|I;T+zCS>W9;a@(3I4X;oN?x`
z(b4+$pCxx=L`Sz8F^uMcv+)75@!4mI17<0ZEJ~UPO1vvo@Syo5s;b%f_(&QEr(85;
z<PGB_iWRNCeb$3jAt)fI&ahv*aD)~8@#DwFJU^kdzu^z);;y&&S8i_yDx5rn-~q8E
z3m)yu%NrXTSuNmQyY=hfU@5jHCf5mK&3$0yW-`_DqT$_RM5}Xh7Eu2F2bXD;g5cLl
z%*h*8vM(;sDLiK$b1QH7%(91{1-SNVt$(`6ZeKj!5`3KKATIMjHFK|nkp&k0EUmPC
zuSq}%?CWW~n6+#kp{uGVT`0MC9ddtQd0RwjLgD&8F_~vu%Pug*(-S`wL&H=y26mX@
z=<1l&x3ndv9M3M-3?qD}%}D}B!JO6kTiQUtA}+|3zpaRn-iJ*UyMVzZhKMIg+NKyd
z8oqI03Rmjn?93qx1Ewe-Lqt`L!X0WXMhU*Y&lxE#BNL(QttOS>R`)z#o4){J?sGL>
z0mur)K~0=QZ{VG}7>Do4$l^C5_bB<9r`?|X5gF=~lWp1$Oom7x>E_4{b#W+aaePiY
z9b#eH;B_i0rihILS7l}8ACO)%;F^t6)$^2P2cld19!m!O-fM(AKnCXG4)4b4>kLcY
z0y^vjn>>u1tn5^>3ilg8LBC+`nDOmmouuXs<*HYL%WxwNHBuj#ego_4!`};?@e1Vu
z+(cn60N)AQL`b9B&3Kc#pFo$K<j2v`-Q0T4E0l_vOYNijNou^k9gGHo^zNFpJTaoK
zhX;sF+4J$=V=i)}4pFnF$^92)mhzV*KF`QD)1Qr%5+)2_O4+7_mMJ%gf4hE%F~7uN
zjk&2u;FG=8pUg#OcL_Yb+099m4Xzh*_h!9j1F=R_;ig<qZ$8cMONIR}DkAdn@FUm~
zJA(D&K}2}?1OWlTqIKo?)h0^Ya?}U#IsIBOQIWz)+V%zJXfFWaxYh(NLc8P?6xLt|
zU6v2NBw?KS1XpgeL4fISU4_u+z}gw)Sf?g?;X)!C&H960$K1Mq%dxFKCnYOYhjiNQ
za7KFL239gXiR_X*V%*~`30!$@Tygt6rSxF)k7^ECAD32`H=V3gIPyTW!^_Kyy#=Hk
z*a*{aDZO7~mVBvl5?LVLLrM;YNz3h!eYJ#Bh8-Op@8^`B>@-c<efSwkNfTVl5jYy_
z#>K=~@ISptY?v@4LP`VwkFB#Dys(FKHon|(zF}})o*0%fi*PFbphZPLCMpKbY*YmF
z+m~l$;}8R>uTYAL>Pz%f)D58F3yn@|%i64|s-nw2`?J%DqD|!6w{MKh4NWaA!80)B
zS{MPIbVBe9C%`fcz%;wbhvxBBqkcO$Ik3a<u3XAck3UKI)bu;{4^0_S_DvADIEnZ(
zmz1g$76Rm3a`lUOhnBZhDDh0W67pf#$^6-<Fb`=X8+&`1!SQk5BCs4$Ieg~II>txn
zbD3~c=37wx*KdR;C&k6z<mcM+#*ZE5=JleK6u)>#hZ?uj-8|AV5fC2#!ERU_F7+xe
z&-z7EQ=p5DO@1Y;9^92T@Y{tnHCK2{eM3Y@4{yGrMzBh{n79gc-^Q<b&Xe8;yZov_
zf%eTsx4M`5u9{lsFX(^oGSe#f6uIQ<3Ktm@6Z56xff>X@J4Z_ms?Hxt)FL~~tnh57
zsl86U`ReAlLp4b8dOA{T<z-+;9OJMfN(eH87Vk+((Y#dzvp+@sK@MCj&IaKTnRg|^
zh5aK*0G`hk_p3BiQtd=#1-m{1E=-%mb8NL{UH<mLL3EO$)6s??M$CA1$JXB1cs5&f
zGU=<^{+ZE<8bm}W#qQ2dU|{{2?;GgctN`!^$lx~r(+hAxa&mHnCno{Z@Y8azJ0}5$
z^=c7_L+~{eHNSX!c(8J87=Cq=s)Go&Mj^X-xdd?6g5)lIxmeLr?4gT=1-Aq|+l5sg
zJH21sq!U!MA#r9FF&dHNii?YPyhohGB}h)iRlEr*Pwab9>S=hCo#h;m5R%WHiV;0?
zInv;QJ5qT3w9wU4Cz5$vB`k6bVcg1d@w&k0sCcJm(~lJudFhY}rE>CkjnsQwo8uD;
zB()ZuaRmLge*XRp@c6XgP@8)v-+VQI8<8TYHw5E)n$@6$gK-2BjWFbVGHasdLfzRY
z(vT{6(P!9|U%;bY2l*QE1LNg7Lv<MBxPB5QU}bmxtf^<QZGf5B!%ZvniwR)n@Ibs=
z&62MvZxT}-jEt6mQCjx<FtC{(Jc=O3RE<tRB!0su2V4KWtApyRJ3U*CHa|YM`)}?4
z&c0Ldq?k_*Q8B(M^7tGrv17<ipu&X?48UElfvs$SzB%{r-(Q39X%<rsn-~J!g~c0%
zaij}lj}m$Bgq=7$>5oj;GO-le$8(plcrV<xy@%rBqWKx38k&NXKrW3{2UG0%8+zww
zA%cm1w`r@_0ucb<f7ZDepOLYtw5G=U^rpeJYuC~MQt;nCL7@$?qQS@y7GZ6O7BJ1H
zXbysrXR<I|LT<pf-@U}2c<K8axeomawA6)KCOVq6!tbeHWRuoPIkfPr)#g!YO8@GT
zXjf;>9o{>N`gaL*<L5omWS5z-4c}R6uRc-u{3*Np7)Q2)mWGu@+yTmoP{3=}Nbk5A
z>tZAOnDJ&%T=H2_Gn!YwcMbRoes?r*l(u|hhwr+K*D7Yhu4+Q+spR#8IYI>=UwEDc
z0F-|EhN(2NwqCjjx3YXmZo}6*=x0)vEJaFW3e%$Ry}E{m1HL4-JiI?`r;%OT8E7H4
zg1O|J-!vuR9~Ce55nr|o3(UKC?}91*X4f~e(TN4b^z^h8Z^1s5(jcGgfPH?ZCan@}
zv=Yn=RtJD=@l9%;ZNsPYy!`S0fHh=<mcdojVycnaOPf;ae{GG!SkWnm*sO~|)MiCQ
zHQHa#ZOIPICEt{NsbBnzlZcU>y;@7MT_S{#rEal}LSgD?dLsz?G^FpVo9Tr#L)qY?
z0DtI=?QGuYL^N+03~YUtjB9eMWBp`jl`0u`tBIbVHtyW0{=z{J<o9bK4&nwk@`h>p
zWzRjm`KYzERl7Ly8_ldaYi9@c5P+Mud)<_6il`1hT`(Z5ufCbZohHSXG-=bStE+EH
zXOzUTRr+o%et->r2WHOiEq(1x;jsI1DJidC-`|11^V}b?H5b=~z$M&c4yQ*wMNtbr
z@k_n}9S-Y=&ob+^<~9{S;D%+TcUYF<-y>pG<EE}IEVOs^_V(W4Iq!R;o6-Dy{kQC5
zzz!M4kbg$?wFK!W6T2<yVD}^Jrf&zJM-56LX12m*Kz?NEzUS7>W6)4vAI?|r0~HK(
zZq=`V<=YaAc@ZSF%hDO_r@~&GOgxbv9UK&0{*(xFSOlefHs=y=zDndz>1DDnPJu8A
z@=}=!=x1um$jHb^q5Cm|ZT4P+uc<TT)+>}%e+<@lSsmYbD=+^EF92khdy(fm>F6@V
z3S4-7GqGh=yHO{D<z2x5QpbK$xN`IeG1lS1>#;B~=;gLT&4`Nurg1Sv$Y9j?EbY*!
zK27+`78plHqI~_j8(^JoiwasLO^9j6_Dyz^6B7jR^9snm*PtZGyQRL@%yB~U7(g0i
zhfJ<VtJDk#PdY>Rq&E0bz%5Z!wFvY76f9!Op(#f{t<Nd~@cg+nwPc7SRxm<M;H0x1
z#sAcANR8aq@gMf-opa)h6O#eF<Jx$tmZ%ng4^$t6dgUOfYy&>;c>U$Ih#Bv#K8fHx
z@RK&0^PS0aZCAJM(VTNN3|o*z*cTIGPJ$0$E<_Fu8C-6>{Lij4gC9TMwiZrdQRk+9
z3L$D2*qwX!?av^mOF~mmRN#)KaFSk~+t(v-Bi(Q*N2%djRF2;PZ>GzP;X%Q{$BkEC
zKEB19Gy-qks?C2Fj9@nsU%q@v9s*%d9q=|JKq9~vG!W^e#Kb=Bbe)OEoS2ZY%OqXX
zPye&6hTe=KjB!ee0*#Don(%$(Hixb7c(^Wf9E4;cTPPZlW7J-a%#;@^9$31#Y+AxS
z4z7X+s+E(7U81A$q!Pz|cwc45j_#Z4Vlm%ErubztSY1ea2i}w9aT1(R`<x7FcLd`f
z0YIR_q-|*IVLo7pZ!ijv;sl<Hr1U;rQPp|PpQaWLOwEm6J$RiRAeWQZ_hS#w#+h%x
zQ>tq3Df1bs@DjE$c7QPeA^|&qsyv?t3L<Ct-5#b87U~x3SGmu>dzdFQ%3pl#Uw-R#
z!9Q@qmB;!0JwI?3PaL-}B%_k!S!CC_(Yf`)6(K|Ss!k?QJY%&w&Z+C$QyuJ1n2WRG
z$>UBDy&aPf1`Hmop)r6~yg#(;@wK8?coHIJc^ur}H+^pyrIUIu13dAA(L1UK2L~xo
zuW3Uy(PMEz3E~SZMgP^8ADdm~-(9;x`2bO%ZI=rQ2w-k_pkT7=j;A&O=+lZlw1JV`
zy-RRA3RUARe*FhngXa2bnOI%yti!Mfiuu>~2mCY@s1ig+FUf@PVn|nE^j!<WaO2>p
z3!19q;^oegUYo@p*2-UOgMUMKzR{U+s@+WV<e)yfaKtWK{2I65)0=ewK|wLy3HUKq
zR2-m(Tq_mqB>^Ae+6h*`rYSdHMZp$Z^}S~Z6~V)+r{!owd(-~L`M24+$PLN9qtXTW
zD_~SrJxg#j%OttzpjDz%$beAPeBuiOan3`$ZG+~PDT892`+#wTPQW=l6$9xK8<?M8
zdnsZ|5?o(HEg!5|rGLxfLK^r0I&1|h;Jgzd1r#N<2A}LqNOHSC`;I0pK4fZExV+Lc
zW-*VT%#|JR&!+GC=Jj@H;@@4Ob*otH@G;w#zpnVKlVj(1C7L0v*8T?JCic0^sW>Hc
zVW_>p(;iDiN2=ssC^>l&_@tw_TJC-4iZd?-;H(MY(inoHOrE~tTg|q%HpjpC0R9b$
z@Xi%VZ6tgb8n`R6{6DS$6VSw~44JVUb}Ts%8i5K?*<JAq7y*DARi#a7k$U~IO`<1#
zW;)w<=Y-I+>Sf2C=#Fs`WyLSFXD*w48cP~UsI@0BPrGG=&O*r5Dt1y{X6<|g*$8q@
z*V0z>FaU6<qfa6J<fuZJhNLepDTx^$9Tf)}Df{>Gil-X%g69LeBE1aZgSbruDy!^f
zQl6-6A1KQYcX(i>Pe{YDnVBG2qA)w&q;s^<?bSd@;K2u|p5UX+6HEpb&8M^&U90fF
zu^>vGbQKxIv?sehJY<cNGAEsHNLXZ}@4<Cptm#yK@S9_vxmdL+_FWMdtl)bwmJ(z#
zV$!w!ZuFd;E}(D=NQo}2oz0?)kH)eMh4Tmr2^)Zv0PK1lLi9-m{tO}%16XYlEDmG^
z|5o!jm!=SWgoe6B`cEB4>E-FkcEWi<oxQxfyGx^*EGRWKJbY0^QZf~O-VY_&dl-tG
z^0fD>8ay?HfgJY&b(A5vpVS6^49Oy-OL8NOmWzM0Yq1TYmT^ycuA#~?-9FffFq3>W
z^RP@&5wW+VIbeMMz~~Y~5dnc~FH!X6>gF?~ZFs=ejYdc5+xKm5x+oWCDt$Q!5<3Cg
zI)&Z8NTY!K5IRdsOnAO_F3i>7gSryLWBe<WJW1U>FmY<pH<RI@nd9}Hf*~*l_J)ab
zg2oEfFh}L%__$u#=(8N<cwzY4Y&(_$I~bs7y*xZ*HAzQK_#2v7Z%139f0;&_^f$rB
zQ-XwvwYv|7gJ=Y~F5q^2&o&Do2_-HI*2ziJ>kRpIC$_j&&nrX~2i6qkaIq9IY!!`F
z<ns`Hh6N^V6M=)vudvPG`Cak7=K=E?+E0=;Od-;A5PNa5C1I8rdI%!Z2^BS}Kd{w!
z_UX;o=fOw&2FiRA*I8IuKj`p3{Q_Zyd@Q-#2SOJF1|uM6PHK&OWXf+u<N|6(Qa?3F
z66(I86?Fs!t^)25x?(Z>4A)QYmcH&Ug{>n~m{YRkQ{COU=<)R+!?lF{+E@HHAQLJ8
zSKR`S2W0r_U<lUx2?+@f1=>A|nzShUTkzaJf@|Lyr@jRBPM5y@&;8%OLqkJ%0rWF2
z)i3lcZ5bV4H;i@CsS-;(Rmxpi{ozrsf@!aLv16aBZlBGqfpWwQGQ#&0GWO)jgZq%E
z!p#r+7bg{fTnQ}2E~|`Jkw1*iYnoB_^6;j!{g32$CLlq|<>uwjw5d6lJTaz=IFE3`
z5q>AM?4#>`Rc0Y5&VD%XlFPX&CtV0AL}A1xCq%^C;^JbL6X4`E|BzulQz1A=9+*9d
z9R)#&P(E`Ky@OvsKAq$O@C^S7hv}GhnDjMg38Pt!fWZgKw`&ckf@LV7&SP4FgS{EU
za~j4z&d#=cgnX&ZZ~2|=Hc)&(_yOYx0I$Yj-0cS#xX3^jMBElmp@B42ck()T5qutm
z`=2m7dn@Pc&(^*nMkXfP%F$;KUm#-jOU5NtZZr34z2G@%0U3%~ksy^AYi18@+q^%~
z?2QbbiJ4PWyuME}bthJ+s<f1F_p^Q*&QxT2sF(2X<_feN8fn6?DHvi`#moUi0%H~=
z)4DBsAVk&A!D!{F^@E^seSglZfh=!O4wSgntwxxpl7_uV?EM-J6i*@)MqBSh)ZpnS
z1(q>3yK5Gf3m*<zx_$%rAr}rcE8c-(FcBc&x+*vIX*4iDlSp^qn0^(-|03KYQ0a7C
z#bUANVK|#I5b;+oqzQ%=G7OE5dS%}}%MD+=2YV<Im8~eBmBpZ>X%cg8znKHV25rca
z??^%d)h%*KlCyuXcd?D|v`AzZOfNE&i>ikbn0;r5)1|fOx>)5R($a8@5w{Fb=lf2D
zqaXbfN7WNqsv7Qbv>)ZICNW<l<Y;0Xx2v{Yds&du*Bv=Z1#XtKx~oQO{w*K}4I4};
z6Dr4?3?&jwO+-r`?s#8cb6%R_kDCfVV9SWfwmE{eQ;rd}G*=tfPbUhSgZOU*=pcRL
z;69v*V`a}5QN%4sY6_?Pe|)`pAk^&_KKvQ`5<-b=JtIpCnq=QvM6y+qVk|{eBMjM>
z8I|lxDnt}TO}1h(vQ3I1+fzaq3?s@e*|+zop6~nq-amfxPqNJ1pL;p?IoG+)b&?EE
zei5VLmExKFee)K(*tXN~@YEo0dV$luoeNR{Y@3TqN|g847Dt?4@CZNLzb%)1_2d>s
zHZ=7t?uJcA;#W~i)!KVubvCJ_niP&q5NXg*I>|hG@=FS6huGi}@e;#;3OG3(mrD!+
zMn`<(zoWZdiGQyHUzHMfQG>y;*>j+SlShrfKeC+28&Bm7mRK;{I1XCGKDKy-m#HDu
zRT1PTq;8VAX@#w0c0@E^_yIdVioiQ2J{)X~B5=7mokp`#Ui(&h+}70Nyk-Y=u~lrj
z!B9`{z?#eMj@WNXi!kxp>A6lAtC4k%Z?9FRuJ*l>C`tRKkLxqho>+f)d%z{0eiLVG
zFX_`GOVQd>mi_avvG@OPDJO}pEPd5Y9N)hU#!a8co*&$kD%>Sj!f}>ug^R*weW{o;
z=>(T2_bs-3pi<F=9^9we67`vbqb!AmYAzJtUK(-pchY!Ygssh~?W!(<Q<m6-*0&M(
zhVARy{S6iKw=snzZNnT2Y7*Gq;)XYQ*2lXJ2#H*z?|m16uYcG^ozY0S_KJEn4Xd_G
z@%Ufw)BDG*Lk7+6f>UygtCT+L`zM$4t9uS)?8!S?=Bd8?&%|;U&$WTO>@oPhW!8hQ
ze9LQms@`&`<1WZ6bxyG`$P^oi*}1-PN{aLmYR;A6GJXYy&}KU-k{{EMO8c;t=@Knb
zEo;5Tx*$VujB%puG*zE)_1Ua0{O#;G-EZTk`H%Ln0a<sO`<&ymxsZu>vbYfyu2Gr^
z@BjAdTiC1KM=qyb)z6q!aA|+il_4chP#?w@sV(87;s=s3v*$L?4rC><dVG-O;=pzD
zchrV*e8p-<M3_YCh^u=~O3_c%SdRF_p1;!3$q_8%FX<leF@PLmgd5(?aB<&@gtDs3
zJR_XW#Y{Plqm)oH_KeHq5J9DcpM2)4>&6co4`?GB84pb>h-hYOled`__leHPF(~A@
zRKD~jRf>25;fg0+%kdRvh_}7VP{^c;C--%GXp^YQfwn{KT`%J>Nvt6Urgpu+xa5)N
zl05Tn`l)RmKF;XyY<MhnWk>VT+GNOF9ck|hl9%rHPwx<+P2+HwHAfON7L!h%YpP9#
zpT${9TF<1`CjV=*Og7{=hjVa`P2miFl7NQ+ox`=zR_(+mYYFHUb|rFqXsf9Awa-}=
zwE^Lnq2o@CkJ`JYujo4F>3(i)ok{1okP^Siy`5TW6S*I?f>APlwaw4F3Q?CCSRWAD
zc9Cux^LDTM<$CG<yTTqcS<-y0=KK|g1MYU~TR^)A{JCA_0p1m(Q7>wK9mk?ZxJvGQ
z8zm1bSNrsBM+`k@4>Z#hZMi+`6T3aRV?L1&Qk(V#=zIQiLOJl^yQYbS{S))|in*Hp
zWFV&5=|PF&hPoMzd@9>dCi<t(pOB!v7aoVm?VU|fwav0eTv~@{QL|`b6qE7<pFX^L
zm+bqBJZIb9<@r<40y}xZaqE;M?y5{{=`d>&4If;B86sqBK}cpBGpoml;KNP^)FwNy
zhV~x&Tqo+i>tAu&w6EjGf*0gBoa(yk(;S3bdhS=&J{c0`(uXT}6wH{DjRh=^;^JA;
z`O~D2eyR#e_sa~!o8SsY=kv&x=BL`;I$_0*ej<oFO^8n>zp}$L7phW|G)k}KuO0d+
z!kDF`j(Z>4T5@xWOjRBbVhdojQbQ=twYHv7cMf4+=Wr(JJRcZw!p55HkF~PUIe<o$
zZ}#7sPxj~TIJ}~ubneZT1umr&f3}h@&Rg?+kLi{QGVh|-c^9M^378K5?MmEQZ#R!{
zc_Qk|Okb?fQJcSJaCdrb{k&XOMvEpi0-#r4j_w2iQ&e>9$B)vigqfk4#Y}R0MlEIj
z%fM1RCR?j>I;*w^*?ja<wyep6F?UCLK}ULTE9?v=N3GqdprwN{eNF@dA!z)^1H)QL
zz2JqL!vUgBzCM0_76Ej~E3=;$Wsz_j3~cXWKyv6Tjx&2&B^Qxt6;<c@dO3F44#ADj
z2L%S2mppYGqs)?X^e2A&2!PEDiMax{DV<6AxRt%lcP>(%DU|fPXB)K~T#$3Gui3pl
z)#}T~cqY#wf>;t&<72(}iWx+g%I}-HG*H2UNec^XN-qk9)b|VoQJ$fFG`ADt>~FmM
z^yNmX*(p>$)yKQ@)^gS0wy<S<PwANfm%u@d6lcv7F{5b4IUUV+7=fg(qS{4SMBAm$
z@j+%!*&936`O#QeCGP*Y0N`qPrE9<QhNx&Rl-~GIxvf%a6%^Dx8`9n1wcXH2+M=O)
z2<5K!;WnOKtB7K=U75u`ZZvv-Y^w1q{*Io(=C0%!&xTKPjL>x>%C{YfZQcHYN_c#z
z;)8@PpN{tYp77L8yuLekQyhhV4~I1~%bs&*_@2XWebYt<jU@F0qwM9~>l+vvPEU>h
zi?6k?pGsDB7A;j^Oh0!sBq4X<gPm?H{V*WJFQt@%7fqrgNm~~9r-HGE&m$$*2KP<5
z(@*V<v337>5u21ao!s?GCe764gm?*0(!qL$OU$(at*jMQ_jASq`NGGECh^YrBCSw%
z_2s9x6Qf8pPnBz-DE0SGPnET^VE1@$EqDLNzKC_B#ef)Xn`g$HB>NhNHS&73SjNEM
zq0Hc~Ch-eH6Pyi_WTT2YPZ8^kZ(dll_;r@G%?nBs^DF<*Ki?gv+^mbET$B~YkZJDp
zy|HXc0SPs-S{$CU7J<aRce3Vd;z_|N5ict~*=y7GDWy!?V1GW#?^T;KqNK>Z#78A6
z@5BE*`p`Ll?FcmKOlWVvwahk=6qdx|*RS^?h)bPs#hh`HE^Z;1;Igzkzi{YxBzrK+
z&ZBwj4YHB8`_TIt&-yU0tKBnr<2WT4z$ZnQ+e{&0>wTeD<3&rJemji4o+X?VeD;=!
z;|I#1q3*3V3QG=%pmfwK);1wgxQ?&kd%Vi|7X-|kcSumTTCXPA&{q$%OCEEH`s;z9
z>G!_nQ7r0p=&7HE2i$k+-3mG~qP|N{FyXBL)4iCm5v$IBR>4aiu@}Hb;C5+ktD$_8
zQfJ}CHyq{gG`?9uc_v*IpQ?+ouYP<JZi)oku)Z2urQK8}8Ao>r(tB$gQ`023D45Uq
zQHJNfhoXDcMc}6SJKnfzsP<pL(c<01_9z<uxyyaT=yBCEp17<G>+VoOD_TfU^D^Rh
zWA!%WnFvEGhiH3c`Wk)HjnWZL%904demPJof34;Qh3#-iK%G!J_Vl&a&AZEpsrai7
zWuEM&eMA$lSabI;lR*OBUMU}qi1YngXYLzGcAwO4F@3Z3fmr5mY`r+s;awgrl-)hg
zH?|DFz>%NacZQzh1L-!#zI^|S^6raAg#3gFg(-)~k((FPCT5NM?{c!6EA9wUc9)}t
zQ)KsWfVor7H>dsB;(7J_=zeN9Zdm`8iSJ&Tk6)+S2qnT(<(N{f?8(E}yUtSXyP|op
z8R7q(5MAsmipM~<zy8M;dE1T}*cfk5x>q5oakBdjzq%#kQqlD*|IE(4QXcv;Xed76
zw_HIv7f?WxmAJsr!-gdGzmiY=pi9<wh}(m<sa8x}l*{qe*~J{mrV~f+4D?4o2~Dq0
z-2FN8zYl|Bt)NWHbEsmqemYs@nL)Km-L?VKy~X#a58<z<-zU76y`qmQj}b`|^n}c-
zMGbx>x0BL7&7`eXtu$0&--5rs?dhpq0M^6}2h80s2>zCkR55nZAx~OeM`tV`-AFC9
zVoS%TL1Sp6S!=Zztk=`mr+*$@ZPSn-V|2Z_xmh8_)c)4XHp~Pb0-EES1O$RWZZ8kO
zg$BHj)zfhQbU#N3SPLGGY5bv<Y+W@Q7i47bb>J&!TiEGtHb-D)>#S$U)yShy&C63e
zJdI;Ny`~o7!|zIeX#0L?*Ll(2yQXrjDn4s#YoY5QpbRsz5sr^Q4!3e?<mEhQV*9i~
z9H@Nk*{LJec0O%c1%YiZ&0oCw)Yhi@BR~LT(N-0rmCl0ZgbYNrg@6%f1dmu|EENE%
z`{?sI9Ar@~qrD+}5*KXynGG*taakni&q?}t7@JTzPA&0ljnZc;NBo$NVS63ueo0hK
z>%z0rCkv$ed1MF~BBq{u?aL_y4vRjH@~s^wIf#=mOnG9<fY9NMoK^62jDy>+auemM
zm7f7o2Gt-q;LLUm8a5WrpKA#`6Esd&RJSyAx!o~_Kq}3k_B`<@l}vSD(e@lIBw)J<
zGKH@-pF~BbWyVa~rnO75q^p^{D!!Y<V%;A~=;#g=XUNfUYIX%w@<84;LETuB>-qwe
zv0Z7_&8UZeqPFh}?z-9(^?Zj9TPK6Ke84uf|CGe$J?0Z0BWUG24q=5DLj0T><b%Kd
z!3utAGo4hOCaTsZdp6Dm+GX<+BXaQennr)ahVL&aso_CHrdj+(y1Ms9>a!_zs<mgH
z*&gZz#!;#wW&2@o<L!WeJ6>Ncy`0{`vK5z?;#eib@J{<{o@(=amC>KC`wonl2J`yU
zl0#$}O;kdo?+LRr$Q;TfjAnGz;>()OzNQ%)Y{C!q_|D$2)SBt92U_+FbN$^yA!74-
z-+Y?*q5BL^=&jL(Yl4M5bqx10X=nH5-G!9HnS@PyX^PsW%qx4YR<|*e^W$ohh4LeR
z?8fTGZ(fk6kBOzc|9gb)vD(M)F;xFD2$X^BYL@OR0%1m;kMixq!SKb}R(o?=<h2_A
z16sXd=Gd#*paAEa?t8D1`)__RP<SHoUx<g0fejeP%chE@TxdS1A%l_Jmt5g#Q=3K2
zKpE48?y9s^w+*a0EPNEx%1D#>@AT@??@jh`_9DTzr*i^u+%d4XaCr4D?J&j0CfQwy
zmuqnYtVxJ2l^Yupe^_|n&bg<5&6_G^ewSZ+Wiw`#w2Vfp{ICeag_S*&jMRA|T=HR>
zhwdhD<7L!OF-L1*=3TMuzkdYtb<n^$x|!CrKQU1W3FVvb8siMDv)r?{>$x?n^}xsz
z1M<~9p2oKY`rmffX&(DuE^3+k-*{V&e>^qweNPElBiFX(8`iseD%!i+@cXM3MkLnr
z5ql`xckKyVPYKs}%Ki80{trhwA+m1J9kqtCj&aKG=sSjZd>~)#QNOax!XTgo`tM4q
zyx=>A`iAMgI{r!TeMi!NS>>8mjm2K)U%wq7H%HxlL|c+q?6kIik4EYt)%t-hr{&~v
zqp&ur7uNND%C+LTWB~~=vTnGEr$k!+sEP28xEsLDDvlmDpTZhEy<frg6&qw<^)BP(
zNR;IJMB8c`mz-~YT)~q5;`xamkvr@UkiPyeDY9Pif8xC6Gaj64Z#oY})}`g#95apx
ztDx>TH`}?lx+yz3Z?O6{M0(aNCiwC5RV?&(iNeZW{D0%awxTdH#e#)4g6KuVruXtq
z-oxFw+4}aN0N#X;m(nRL%Q*Ac|5ibVU0Lnd85^Kd+k=~e&JmHg?|tv@DaogqVd6b4
zV`<Tz9*^M~2@`rPf0cc0i$8z9-B>7g{vNI6-$gSC+Bqjvp#`@5!qE__br{nro=p8$
z>}`G;`+s}P1<>?i{Af1eR_cHoaVy2=%X(Mwcc90dIy_g;&)S#v#M?peR{wPaG^&ud
zBYti>_T;DAsJ7qlQSQRI1Msf>@h`)anR+4G)cX0oB44@7S^NpVi5^FF|4F5`rGKZ%
zI_WxYo$z}g^xCz#0bg@mQ7PcS^;Ff@B9z=HyfN>IKM>DJ>Whr(Re?KVGqZ2c#T_$n
z4nAJjoW7;7`0QSHK&nIizsv<zJ1O}U3jP^<`!75C!@0N1VXpB@TI|0XuNa`wOv}T*
zbEk>#B(;54oFPLdUOT3uQ_2z1_?OqYXGxr)uc?{0JW4GgcCfT?sv!<I^Gj%S&I$O#
zv-Hs()~7~S9ry?F^*jG(1For0y`;<$81AA=!~!JMbF{0k?Oq+s(X*RlV@pb&+WEcp
zN2xV{Aeb~c!jTertn?cPB&oGYG4#kmDpRzB{9moRp`Vi5?9H)MoWpJNy6dCArkCVC
z_~esV99TlZaoNPY@@Pw{*u7$oKUS<w@wqOEGQnmfu6qdRrK4fW@trGAfB|8fE&c@x
zE;E3QEd8pzlD$J{`bJ1dh(ks_F8$)`lD^R8YE!xQ^HuAsGhsbV{KuR*t{&#@>~zK#
zj8?jEqVtaaalNdp`}I&2mm24!neI|=%x4YSyIe(`LV=Jyso!$i^Vz8%Gt_<vw|<Rv
zmnaLfz039)ZRhi#Q>N!q-VUA1<Xhln`1StERtLNu5=#@cd2s~ssrT82^8170_s8AL
zjWo@;@`wHXRzkgLIg}wwF33QYOZdH?5-ofiHlwSOaBNJ&@*NPDmDoa|58U_F-<;Zb
z?FrEXwdLJ2<pM2VR9l3`0Uxy`8c5I}5AbE)E;{wp+K60@92z~W@i!*_)j{<@t?Rff
zx8Q4*d!dSP>fwvDMp17c(Oc3Z<KrQkCkoqpy+E%!WDSVT@u>W&wI$ORr<YV1d<Q>?
z&ZxgQJ(38WnQagBF6NSPahUFn>bUJbaf)~ehkz#44mMCE=8-QO(j2j*h^fE8&xuhv
zD>tpEKeSQuGQ6ajR%e{uxgn+$8JmS%W9esG)^)r&?eWwC4}bf#ySv*cm#n-|W4*@Q
zNMu7bMM;=65SmuG9oq{W2vy`c6G+M|Hd3hv0e05En3o4XtOzOh&bKX@Fsn^Y+W2p7
z<y9z*8E@};cTKD3e#>w}B6g-2!qdLdfL>d3?>ctM!=0RR+8vkLyJliBw>T03)~Pwz
zC0Z20eTDni{QNRhd~|x@EF7*t3CUOW@_xP>(yQN%IL*B)77ucI{wr$fzT-=L*Vl4P
z_P!sV<O`Bm-^O8Rmmc%pAgi(OMRDMqa&{NCr*m$J#Vm5vis8oD57vu%KM%t*x-Qu;
zra5Se<aW6$`pdhD+UJmSju{vix<Nl1Q-tBdBDt^jw48S67mCtjUJbWcu@pvf2+%O9
z_P2SbD*`${Pl&~YczeP;CSLljTw0uI6t+`XUL%;nIBkA9_2PL%&7ID5hXDGcOWk-0
zeXmwC^$>u<S#i1KhNYz?y|^+|F66c<x79fSg=|vPLnLB2fFsG#|1jA`_0Z;hIy!E!
ze9#=zMlz`RQ!IuXcXj9hG+zv)-lUoCYi79L6_*@t6WaD8N=Ou><t=u@6dix&)g|yR
zsg1Pw>2~P7qV@;o-e*^JuaM?K%&ju#TtC}%*O{j5B$Cw*ITc)94Lo$C;Dy6_Q38H~
z*0DfIe9zY^I(!bt$G2veHK%KXYo*>Jh_rW6MQ>o_F6Pb3OHbeYL~xbUlIx)^Hp{gD
z6D7mVcQZQg55wc~(F{b7dqv5@+)I?fN*&G)`g#1>Pyfwqn>KB74!chs04mZa6?PpW
zByj9X+<WfVCYOz9!%W=8OH5Adc+@1f`XbZN(hqYW9G9+P7Yx%a&7ybU(^{~e1B=x7
z#6;!5wP7(=I{)K}va%rj4Q42pjt68p8{#m+vaark-i_riYsJt?@)axm>9(|?Kxxk2
zt+eSSC^=lEglY6~uaL=eeu063m1=ExsB0)ts}1I(Y%rJ#a<nF(^=Qkq&LoZMT2@xp
zr`m>RWo6xV^E^d-!58S~xE!r!8*%V*pYqVVdwN#OU&OOQ1~np&96m8Iai1@K8BoXZ
z18dE{P@|a&Z{0Q=@bqVQTZZ*n%T?o<+kXA}rL*t%r%tNZ!eO_CA+Glx%XK}gy-zv`
zX;P6B)Dlkr=tTkU<@Rkw$4p`}pTG8>9*B_BHvfwt&)^@OvS$tE^Ji0F*~U@;bv=3<
zE1#0_($YbhVCE`5h|Wxar5)}l5pz1jvI@qbn*p2e^LxC6$LQ#2Q0Ao{|KEz0o4D0v
zi4gOa8L2nz_ITKxvLi(potCmjn`Kt2xo^{tN(DuC0Vn&M)~~9GVUY*kc`;s6y=&dp
zg{G7kb*9)~P_tpoI*ykhjFVpG2jW6B=|fGa)~H0K2-Y<b|97Sz<h<k-E30_t3G0{t
zCSt#5Q7(^*G2)nG1uTSp&2#dy8oTd*>;YTmuu;cR%_puFJPGZVM}O)}PR3%SkuhMr
z|KDmr*h~Iz%$p)nxe6DiGVNS{RoHo~<jvMy%&zt5v^^q>t}}h%Ce(?VHqJl}rITgi
z*6PEtC9xr-hW=c#xToC<1Dl5!m*eZ9e?5+S+LrP#tTLX8iV7UDYHWNQI`^p7uGHM(
z6JJ9y+HF1Z?1RS178Y`e(m%$>EBHi_-XdT?wkqL8B_x8+K{C_fDpYsu<GN0Z!%Whl
zah9%S>5wMr`ts8yZ_^bil@D7rRtNUh?A$M|$F+i)BGBCWZjSE#)!c<y4^q2+*;x|{
z<dYm%8M%bcKX^de3>+O_REI_JI&gw?ZoQnJF0<5i9Zzg6>5$@cn30%er^~lZ(wa_-
z+3kECJG%6<y30GkyXGanVOZ|xWqNMBN;)GstD)aO?#nT&D;vOjLjB6FDJ-KK$UWcO
z)J*oL(4+TkDicw+a~Q<ly)Y1Ofe5;bHHcZm#mNzh2g|O^)(AUFy=8vYI|MA@&EK;n
zn4mH)#kjBSH?cAV8U3%&qJ&JzEUb~E$?g|kBK84q<Ykz2^M~+_g=Q(_pu8w4DH#mt
z4XI6D-&oD6jHuD5VlBm!8$%KlOsN!L(Z(B>7qDu37O6d8QO>_gSi6~(k*WDs!rU4A
zJ7$4f%(PF6!Q;p}HGx@<rjFw6ZW%b=`m^cK=$~I!B+jC+n{un$zJ2A?xjDLTC;I=T
zrR6lU*y02PFSBeF`_o%NFrg^jWy1-8*MT8kO3;h{p!0;tKLI0Pn2&StKkcu87k?SO
zdSuJ*kzQ`}miJden?3z4L<c@*>wLHCcTvHB=s)uU9NSp^k?C{vMt{O}nhs+!RXDqj
zp7}fd3cnpEh7!uE%s2O1u~QL^iTv`vk4jd@d#SA?5!G3y*=SN66Y1f}217c({Vj;B
zx6ia>rB`t*&|KKjX!-wCwCjH5rB-7+GNu&_3{*C%U*MIV>B@>jsWeyeGqR|C&ER=@
z-}&vBZw%E>cgkl+wU3$D9+oxFGcrOcS$`PO5RESNZYpQ*u(~Mr@9ot3eiYEAs8wo;
zS_zte-+AcPqW((1Kx=I>Tj6Gn{o$B%jEhzCK~*LDQ=5J&um4o`poN?Yag<}-s@Z_$
zE8kQ|A&AckBq%DCPV+EUsof8S9@R<w+0=EV<2r!lFIiVDwbhC`d1mY*a9S^TcCaz+
z*;aV}g_^k%**x-D+&asG=+q&s8A_Y^DW<p*c^<#CL?CY3w|`{R2j*=*^UvT1={}e#
zZ-y2%CbRD5c`o;*72YsTe7GQO$%^3%x?|DIuV-<n&y>{CRM-JwlLOJQ?L7iI-d{R)
zJ$C&)34c~jQ$2vH9_EX^N_@pa7rMunru9-C6`V31Bik;Ex13Pclu3wA>sjMiurP1$
zT+jkxHZpYvzg}C}p7yW7kjS5<YBsCl&mH<@ayJBf@<Z>*?tGMpS^vsEQ?-|C<-C+i
z#nne&oLF88eJrrRQ^;+7>6e{Zl&SOIsBMsbjnDFBLVrhmUJyr4yKwR`A#3bO11y?(
ztIIoPH23vp3X;UAQ<vYHP<r&J;Rcnx+3Znl)q#t9rLA#C46=tCD9fadxX#exxy=dZ
zGR1xqX@|2b4qoP=-N)zc9Sv>%qQ~|UnKzSBO7J%0xAw@Y=nRW&^!anX(o)qJZ_`cX
zP{Bc8I#P);mZIB`^{@VC(Zgx#C;?Q-<K;aG*`I|r&QF4Os|c25z=xuCpSg>5o!?E0
zoOk`D?3AS3#M4tKJ^qyzv8dC7V{>{8SL!6I2T~catiHeftL$F1*WUx{H4!LYDB5xI
zqML3!kEg)IEw|iB)Myxi<8%7(Go}Vd$)Ic)n9UX!1($@PrtS2QwnyXwZ*oLr<Nm)h
zZ70jnVzf~wzj>9qvSy!9pmVNBDuoG}?VIy;)SUWgT|TgtkbeBxgHWb;--2T)c5Qop
zWkjUL=tD{BHAv~)Mp>(A^C#@m_zD+&YNF&Hgrcrgk)x6)*k^vCN1HeYm%hWpfv@Ko
z()u$gEp)EEt2u<F9E<Hv*`E)eaaMZS(yMSlHF_7MVJz_yuBnm{b<UXz4?nN};g^7A
zNz-JdJi}&phMBs64|7{xn<GKR*eY~m>EJb(`yZnkeslKEZhl%Lp%?6be>k8)zRM$a
z8uI&#!D|mWEFNKMcV=Vqg&CmB+|n{UbDotMh(K?JXTf>|`1%2>s8c?;Do$lprgb(!
z2^bEi#)lkLG*$9w-WBIG%+P+EVtd!OxY!t`|H<*o7Bb9uDczBd{BF&-jPsvGh0UC>
z;*Q%TU1QHV=GDxU4?~`d$+yYQMx0=!mofQpWCw0Mx{J~Pq<ackS*K->;@Lt5<UYZQ
z#{RJ`inTN4FT;{di~DRwS$T<l?o(gDYmkq{#KqfU%O+CsM@@3BSdF)I{CoCYO#_ZF
z$6~1%8I2hdf&^CqxH4wG!H>vuEKZy766<#{+1@1PfX0McEpTgXaO3wJEOA%!MJ7KQ
z)}DO%!!gg!l&QabV4{MM#S|HE2~j@)G;2NL<Z{*JJVo_=zdm<mDAU?rCqp%ESC>x#
zdwaZ}sRS(miyKVw6?8qc%J<unCuBDa&%pGp7hR6+PG}ik&cRGXEC$2p<>;hls!i;H
zuYpuNRn0*Q(aC7(2AO^Rl6UyY#^ISaRaNV%aj~P5lLQ~Vj1pKO0dX-usOcX5QoCv;
z!-Uc%rRQ<WS7eyE<U?2Kx3`ON-f0;Y04nICR?+!ZaCK8^DV1F<m1Xbn8822h5g_vn
zuIgk#g<6NI$cxA{lhePI)iyRsbD&Ko*lx7f+&=0>s1hzI_bR2cW4XrMH&6xi-@=ZB
zqW5i7eI#jRZF0Rbin$Q%xXn|2%H?tYw~cC@VWB@4?d1?<WKEdDw(r@gD5wx7NY7*8
zfVMu|nzm7d^at#CN+z@RQJcyo3sjCE&N$llibuzA+EoxnrZxP^cnq8R7;f2YhiW|r
zm0vNW+!CyT-hb33jb={QP423TZbbeZPn&MlgMmCWD|~~M!Bf>utWeS|U_c)J-}(RC
zrIqcU##!9tyaWCDjtOzSwv9HenL?lX6p~yQ`>+1q4J*?=r{L~e>x_1!ZK9#kmL&oU
zstg$_IZZ3^y&48g*n>ZNFK>L5Vi7DajI8Uue-(Mm={ObVS){4=@ynWGOx*$Y4fGIR
z=A-tEKBB#xlA6uCkmJGrZ>|faFhfyaubViU)`B``eyxv=P=eTBvza%^M-?F6vCMhX
z5wkUvpYn_wAJu53&*m;MJIiM&>a9E#pTccq@FyxtC^{ra;rExFngm{Q(MA>labQi4
zD7$>hnj@ieW)g^>qhU4FZ4brcBZ&{5!yHDYDef&qeVwAR3fv<aK`cTjs>i?wc)VAX
zT%LP=l^_r-4Xj_6D>Y$92RHpGRe{P#3=J=NN{iS=`6i#lQl&eR+(aFxeA_GVWV<F>
z#r${mBs^)h>*fXSLcWEIbk4eMEXtgrn^r`o>=-8!2ZU2B5B%L(gD*;-&bhLUz0j>>
zU-7^W--~&;C8;j$@8bhDo5QtI4iI7^`+we7WZgN7KOfPo$4Viu$7zM79`1Yoq)S!Y
z3gG|(+z~~$I1JQ9tyfh93EY0Ud!BPGWZv|yare`EB9|yKb&qbSMS7i#5V4W##L53S
zU+!hOPikS8LMGAQv7AVhis$_GT=($~e;uvQYC&c~HYQPCMF)=`Wjpzl?!<0$^|*?$
z%iQv^)7+PT8V5gr77x~MY0Vk$9qqx5>mK=DUesRJeiWTRXJc|$`8;B391)K4mT*00
z)e&{f7mZIiwsTu8pXp8|uImU@OEBOn8vn#cp<GIM)^Cf8km;8|w_aQoimGMR#$9`r
ztq^{VyTamqhdaCxz4YeBg8d(_H+tNJPh7#^cWu11UjSEKS3_fg&uVoGQ5DHAY+t(&
zbUSAELk`BlEf3<xc^tPP$`OQgp2q^zR)mBy<=7_~{w$mz(4wAMh`WvWb5L8VXO^>0
zI2DOdHxmybxc7Yys5?iW7}Q81x86IUh=EAilfo@mk%LDxn!^h1?EEV#3w4n%H>)gI
zJT-`#{i<p#{#&V*Ncvu}_xAm|`cl2!Hy=owv*A29QqF>c<4=JE+=R4)*v8lzR77+@
zvn8nxA9Y}4=a1{E1wOJttXVD@{qtfo>V!>}+`+{Zv6`pX@2%v^zUl}kJh<!qjKF{5
zdDSo1Eke8jTkc9&o;;TMq1p+>!+lqQNb(F;D6Mrs(PIw_F*OjQtUDul&v*XE1^Duu
zyW3!@#mw~hWW%n-<;}!0Bu&_U(u^i1@pt<EacbOzC<D*0WA4{4XB5<kTOz@&^Q%s8
zCcj_0DUJkh4<8JR<dSfWQCC%z<2PK!I=ESyUNhzCw(g&ex^qgVz}ocTAfgjdog067
z!Bxol3x@3wv9Q+@gK1@d%x8OEj??wOdANCc_LyJ*DVdHSf{P$U5;yIy(*>mC5zacO
z6InM9OtGJ!bIYeql|bRc{hqwbkOC~rw;^BY@EY!&j`z#eUmh!3KQlqUA)u$RE3tm;
z+youe{cctMdZw&vJwdOH6MuxuFCCKnHpxN(`Mr$v1D8*A_B_mo^7;@EC^%s^L4T2x
z-sQJ4O-t%&`u@Pyy>k`1%jPTlO%%T8x19$}=U1E8@PobQP7$DrcHBO23=|1d+o2pY
zpLFI0386kUPW1f#B;w55XVcqTZiZh-<>F~@R2$Z+nK_AUb%;2ArocW9s@dJ&#z&^@
zcWq9yko2niM|o>E2mlQ!m5sf2{a7O)AkbiCZATuCQ}3l*%dJdiY{2DcbeE<i!Vq}I
zd3tsdDeQJHGc)WLUyhdYXN1KYNd-I|i|B>N77dq->-%0WpYdV7a=zmAu-@AAZ6<Uu
z+#8f%RaJF?{*F^R7CIMF#!9hD));<dqsB`{B<8;E@UmXI|3K91VRN1757G0KD{Z^!
z;O$>*vAg!^=vJnE$ZXq5nhi7)Q*}u#i3Wb+=c;H@JCLi#sy67uKNz5>c*8`g^r>4p
z=>n{a<`$PnvC-W8zlWar=nhlk`r&`-%x*<Q$Zn6j0M#loNs{I7!Z@Y;CyOt)el(X!
z-hc69#t8+RB@XDf*5@%d*17yZy98t@*SCfT3V57=PreG@G<u+|4#FsLNZ_7Ls{4cH
z#+ZB3Sc#k`&X>ObWF)Qo^2w9)UZvh3u|zodglkBA4CE*{o6>d5p3qAHjv3cPTHp-=
zeh%uow4{f4{&-cjHe29BSYb-Ll6FU{XzZ3f)*J0r+EH}?ok;^Et0+<4LTa~E_r)Z;
z*N7C!lUsA?dzJuUx@H?~EJju1kKyq>oIlzO^v$l%f#SdioxF5lDVMAkt=@j-@i<cf
zF*%_yu?qp!O&-vZ`6?2URLS{s!Il4UF;eG0!%}lSYjeUuD;ZJR12Uu@j;NJg-g-at
z4x+uqwtuz$&pFgi7#bQf6kJT&!S%q;$LIbk$E!_I&;$BaJ5%^ma}7R?c6zo4Km#jY
zk-eh)w<VvD8r}`o#t#~AdE>3i8P4)(zm85UAX68nn+GNp7QKJG`gCv40j-~8Yh)Bv
ze(SJg`~bgF0?&BEtoY+XA2(^9GU`|~VPa+A-iQ0~RljU%stVW{4}Q#;z56>JPQ2Z`
zoFE3V=kVM|yholRy<*XA(KfL%1Kk><80qV^Q1E03RxJW)WG@t=HUX<IHHek@y};Me
zkpVv&5;0{+SM_V?gUWqqnt9HQEiU*Owqv!BsgXpwqD!=OdiAvgs-LOtBQ2scdPv{F
ze%qF2aHlU?TX&r4bP+^sY;6Ug(cz=Y$v({l3Y7Z&D8)`}N18}WoHYC?_-@JlTgC{W
zX~I+w*QSTF<gI(onBF$bnn!80+AMefwyU~GT1@rCKS7y=sm3DV7Y`B+{d3;a*<V6o
zZ)sHK9UX-=7M|{6_k;begFDtrD{xlE^gRkYWk%lX>+>@(2tk0^C&2JTRI1gkA09Zr
znOfODO}eYMeDbUAGjZyLEpH-LHxmhm&ZmBLMd@6xK0~pr#z_$mANp*p&?s?Ey4GGQ
zp77G~`bIMGlDuOq9))FPoW+e3F+(zlJLkS`o8{3@7vAq0ZDFN2u^97XYU#VtMlY@i
z5O)&eK-%g9fU5U<{&~`vtk1<@5lqP?=@+*9I^O#L7wvO+(3a;T0{c@Mzh#jX+j@KL
zH|na`h(SQ@E=^4wgsAA|4z7n*vmH=7Xks)9X_H0LQU==V70H}^$_Ifm+9v;HZ@c+N
zXfG{d4*XRMl){{N_9lH;GLtkPUAGl!e?GDm8G&`Gf+BZ3;RKXA1Zl_5*g!qb+gDwI
z>GLS7)vZJVx!t9<GaNERyo{5`Wkj0QQ;VqSRAChJFf~=cTVmJ}abY_&)%walCKL3L
zRwZbQ#Y~N5mhZ{XZAB7b$*U&^cSzl)fl|^MEA<>=zd1p9p|Y>Ef_C)cQRKkmJIGnq
z$JgW^DDW8H*`=^kZwWEw@Z=$`qH<ZM_G}U2LlMvV=LLU#vz<Dju;qbWgkr`&C)2_)
zG7=quh;sPq{Wo|+iSGPACbo;4k{?8l9J`RGTjk$4Q(jEU2Zm{2@mZ%vL51L;B+f#A
z9pm_^BeS~5TuHmT+tsT*vKcLXoQ^5vxoR(rqN1V^`yo#|rKDxp#*^11a!d@?WYWHQ
z{)2S)^$~_HnOuy;m>`W-Z64-3>(9q73KA_4%Ee4$ey2uP6qdzNlW2*;dgN8O+?$I*
z!A@wRW-r@6BhfDIPvNAED4R;rQ4DiI%`;ZgCGBtGKKCV|#;k)My*Kz*82!_%Z@x<V
z4B`C#Z)!*!vIDKoCm`VPxu?g@m-hq$?czT{tq-3<u)V1D-S?v23dBAT&28Q*w=EW<
zk18x7`_5_92Bbxv6<kapaUjjC|MvX*uV?M0$qrbP8=x|$OIqTk;q9&32&_}dk;RLZ
zefnpr41Tru$M*boz<dz`hv#t}slMWE7dEq5I7x8B4gNNAub0HI10n&P@8*jj({i$;
z+p<jtRel<KcLQ0kdO1>pO2h5~O~0z=S3reP32dkY(l-^du7w-7K9T%!r57O9mwReT
zii;)q89Nbukpy15jtU;;!eMvlnaZbNy1b@|sHZmd-ZPbxak`082)@JV$xcknBfD}O
z@;nAKW<O%}=Omu<Wm!)j*clnoCXWsZ((wG^78#_#lt`>wp03kSV~P>W!WYri@mqJt
z1>b-8aKTD9SZ6h=ZXLBy1;PQw@yy0kZBEcU^MQO~{jU2LvQ1Nda&PIiEkRLa(heOs
zqYruzCdw7XKlEVtaue~o+XNTylI(|;*OGF{Nrvczb4c&}XA|WRTruU+fj|^d8a%s}
zrOFJUge+9!i8krmIS%N+(Z8mE6YBob;C{_u5zQ?qP|hpGEAWh_qY#jSdZe?TACaJy
zUV<S@PkjIG520@c8)M<sT?zD0gm@$t^Q^Dj!{Gdq_c<`Q=+BuXaU?B=w6368Vq?&C
zdRE2E%#1TB#=1>R;mg*=rzEx8>h&0?0g|Mefg+r=aDpb%=$|M7S6=>+^O_&FBj3ly
zupO<fiL*MBzo1@cbob339!50mi4<9<jO}di{9`JmB0ZOz*m*@Z+2S1FQuyws)@_yl
zDsqL5`Xhf-%G%3q_ii`Kb0jj^hk$T`FhVc1iND^zPfXal?V+1m&^LuAK8Q^&c^GcI
z?JD&cXq3o_tW>~*Tj_|fy`$8MX^B!>YpQ?>wV=EhTxn&{%$w+N@x_StQD@7v$z>ek
zkC^+|`88`S%Vm}za=&k6U&nf9uj3vb1tCyn&WhCH)5PC11qBxo(Dx5t;!#yqeTnVR
z6;Wt3Jqsnh%22wTtqkv<GO%SG>UZ^{>WmTd=l2G|qRm2g+Te2r`D~u5YblP&&q%2c
z97UbxKwYKme(i&unh&DtWTBerjX%`K8ljw_<>KfSt70;Fy!Z2GFA9JjuiEYT5GejN
zGQbKhVxR4V45JU&&{hz$N+!?6fT$7d_LEkRSgGO`ce9F$ydOp8MjN1Il6qE7TIt_E
zw)09fCsepH19~?E=gg3aIKwPX5t@KwAC$C4RJ%&g=uEzb2*>V8cbV1f_0K`1yu4Lj
zA`_nPZSfP#(2dGT{p*LrKA8=lp4C8V1|AIr^oBp<PRfMk>_ANIp}0A{#;J00{^f`!
zqI3yN+C>AxWDJe@%2dv~g6z8sIg2Y`_mnj>_7ZjC4}q$xZ0&gOTaglHO!HA1<_2n8
z#eD44fuV0Zu^sKDPVCT4f+@5@4{#o0+>haA2A-Z>zhBwj-ZhM@2oT$k<H2aWXXAvZ
z=TrRT;D*e#4bFOF^2Ly^Ac7T&pxe9T-=j1ASSc^gvZAGKbCLps^)l8Obuo*T>pOW&
z4w-)+9bG<0F?CA3bfBvg<ap*6D%bPi>sU{**vJn@pK>Z}@nQSd5(N1U{9KYXj!)v?
z<lGjJZf)%qxSz*rRlWY3m-_P&qlVZScO*+z^Ass?t&U<UH&X_hViP-!HnVMeIbsd-
zmp3?-*nd`CRs*y)EG*k5b6S6tljq*o^c;dhxG3a|lHJ>1<U$7XFjAClmU_l_@lIs$
zR%%pjvRQpr;o0xJ==~L6n?=15;zihvB=Q^%+#XJZ@pi%<nS+XIfy~0PW-C%)JgM@@
z#4$*4HjtJYp#5jX-KVV8^Tt)AiF2Q^$Tbizlve+#b>8f5L8jMTCIO7`E=ftrJ;8k$
z4yPD1tfZw>PdZ-EIIFPXy*Q}Ectjqf&y*`akg{$|j{J2$1_km7L9E1-PPe(hd*wz3
z&R;aZ)|~(2c5sS>PE{CLHEg{c3m(*jwW<5fzG7Ac;DQ}!6Xc%a#HGAd86M+!8nE0A
zL-$ER$zKubJ7ffpzg&T~Hn-s<jc<!GnBeftnp`e9Ls>Hvyv%)lM<(fZDcq!(p<#qS
z_r;5I*se1MQaslmf7VEPNfJ!pO-)U8jne?{NK-+jOlD<h-5F8rbW=My(RCE&NJ(%p
zEpjCTGZd6Q{D?$SM=b<g{-+tr+&Mh`Q&NP25dX?#`sRcVm*%v^SI5IKA$<g60~;i`
zRy1*1r=U~R`+lSWP#g%BZXi3mu$SlqX=G`IIP10)q#az9^Bq0~LW1f^4NX5LCO~_O
z00?0W+RJo9i>*dXy#I3?ABa_yt)C6Nj0j}?3>vxucv0cwYcA&BpYY%}K7D8D^m*L)
zPp=Mvl8V{WvwLcjYig54;BAkdA8ic(b^&n&Wvzm!NM6Q!$gu0Sq<X}bVy$eg$YaZi
zk-ZNoF+3aFpFbOHViFXMiX@Pcw<w7xq)8}N&)L!qyxAORN{mNTP{FM>L01=0%Q>X&
zTN9K7Q%EFNTO4h#;pNzrFaQ3#uN+J%JWiDYIx0ciLDb3qM4NKVF&=6hz&sKyK33!-
z2EDTABw)|3zlMbUDk#W{!-W<vwtc&ey`^Q5KG6#O8u!c&K;v#GRab?5dPOZ&S5qj8
zE{jgUD$RRsX5;kn_0`k0p`T@L>XynPIiRNa5feC3me3Pmg$m*xI{Fzc>u(^1&)zbV
ze!2~U0asak6Dg3DC{H`?9_z2R3|%G>T1?%0aL}W{IJmx-l~A_9NelS>6HsHbym$d?
zY1!HcJKv4cUB{gzs~)}K+XIqYf@qvFSj!}FBFP?ktK{JT=*a|PxwFx*qbo_FI}_{E
z_I>p3+8BBNgM2>dk=zUAsV;(x@nm1Zy)Bihwe%Z@!>~6_-W7ia8oIq!R#r(bM}kl~
zT#OOV`ugW^$z}uWYf>$HoZu-j{4hQ~o<@t7$t>ajdyVXaV6KEj+ASP<LR&1{PWu)@
z-^9oV!&?SoJFnlAmIm1G6w(Cgxlv?CimADzsHo@r!>RJDU**dOZ=?1KCzO2*`3f0n
zONfLDP!c=QCaByqZM<N}i$SsQgqv$@DGxOXg*}iyXlP8|X;ty+?c1e2P}(xax}!E(
z`f*|2OVWr(_1him?chK*h<4Y7@=<FL($T?CtB@p;eNxMCRa;Mwu0}5=E@?Qx*lHC;
zv_q{(QMVv_paNVzrbruPNOq<+t(hC7i+Xrv-GwkxMv_*Kd=;R2BlhekY$2-F4)#S5
zr$$A@B~HE+$mbm)(nK-P_4WgU$Jaks)r0hMI@<Z)rh(zV1}>EF^Fk8~h%urJvv^TE
zk#Y`4cA_SFfA{=q`h_%1HaoQip)ifzHw#|n{yUFu7h(g{i4g{adRDqT)YF3JUXXGk
zOw2ykg8isa5sJz^_$2)UN@3@5DF_H!!Me>Ssqwp>ZWKvKJUZxX)gw;zKL6_9i|E%n
zsZH+B^Yhm_z+5MKNrDvYcK{Ev<IlFA`b01FmZx2KH6Ia`rgGwJbD_Oekb$!p(cyks
z6-doN33fbh(&JzW>KdXJU8mdcKlw9+u+^~huzJF=FBU2LFH#PKSsH8D4%zj4Yw}Tp
zkU^A=0^0Az$ZH5alCmu_)=Zxb8t%UDbtd;LCM2z%3z#a@snP>k`<7A28p2Q#EHyEZ
zf|K>pd`Br}cgxJ&iMCEs;OodV?j5_)%3CtzVJIf?O(3+<@$lh8e~osPlgE%%GpnR+
z4o9r=>Yeu!lsYEaG8#rV+=ogK7pO8j5Fgq5_P^Lf)JclX(-u71c~Hd-bkob|Q%S5}
z@shyDAUd63=)Ofr1%B^|XG*x|RY+z9F92TXrB6(<npG7{;dQjOdX!P`EOMWBBNo_Y
znN8<l+=of*Y@$@{L13asQ9DojC=m;6aUNna6B1iou3;mq-LBt^q>n7Gsqd))5#S~8
zUFC`0dE|B71pF<Py=OGzy|3tH=pQl=Lej5Zx}*_OJZK13lm$778}AhUp)7UVXmn(x
zorS^`K6r?Qs)RhuBNv>GR;f+fgMJy@NAO3lhQHE9q)xTbjt-vQ;@hIB2v6i*tkE*#
z&UweWO&h*zwTckEcQ?_wG8*Plrmx%6F$tK!^rTa50<l=gvU+%YdX>6?qJH9cER8%c
z|LYf>i?N2WP5OYkWby1&63cf!Lflalot<F^H%k`G@_r2s|2yQldapRpsdNT3LQSyq
z8KjJah2ePN?XyV@x#TrAraNd#-UfvqW%jAXPuqjit#e3zEdC#CmRF<rUl$e42-GH9
zTalr-Ja}vPqoam0=e?MZK+nqKE8b-<Xp3n&HF_Wt&UqDEsD8E4uvK)dDh#YF9$q3m
z`GyR@P+}DJdTjSgBiUcc)8!4nxGO7RUn$;yuZTFAYwTG5s6CvJqgf~A)FZ$9{Jk%l
z&zF+Yx7?;V3CFQ|I-?W52B{1Ti8{%F(r7S+>~`+QHLcAF8)C6n;qyDW=U5T05}l?9
zvPNx_P1$zph*x*p18B6Q!o@IZ?`m=Ub#Cwd=OiV8>yHD<$5C``P_-zA7M&exS^AIf
zxNWy4G&MC%!hM2h><U&B7Ev+|AnZ>`(@LOVmHf65{nS)dC3WJ5cBALv%o$KD|D|Zg
z&-g>r`KE8X3=Rkh|AvpXBZ91EUq{Jp&L!VwT{sRtPK<C=47fhXogG~KtUTez05p1j
z*qhwmTV7hvhaAgkJpv5!9k0du8?1a<KaXy}I`@UNK&%q-row8<rLSJ-3wC~S(2=^m
zF@Q91k3XuuY43f_UeEV@#0!BMbcvbHYj9&dYWNoal|IOUeEIT4hmBDKnv74_i~u|;
zLZn#taU)<CuO^Qvl<n=EU$t$JRp0Qj-%e=9VG?(Lc$Pxnk?HmK+=YocnFFu%iBuTb
z*GVq1H1r-tCKj{pp5`;Ccim^$6qmpWI2Dnk=I84>7nnYLv8Xf0YIYtH<*>hIDel?V
z;URh&BwN0Dxh_{a-~&D?;FHeb2wX_|aEM$|LteaL76Ae$8@EGEuxnzyjKd-AIR%Xi
zS_HZq^IpCT`bQ{AM=Z`T1AeKoX&g=!?PJqJjiF?xQ0X{B*A1_9IDk{kNh=N$0&iIY
zYNJbabHZQE3_t3&!ZiabalI;C?YUXyB1ND%b*405TRyc(Z)=pn88N4~yJhzvRG*%T
z=;1g_5jeUKlj@`V>wS7+PQBQoU$prXWh#RG_xY+8^+prJ<C(wWIu~Z~M##ceBKM?0
z|66x(H?>d53Eru9o&eIAsFfW<!+aF9OzlZs`}gfD>@?=@@G(<Lvi%WY-&5zEHU9Jy
z#74n@l7y5(B15CYM?aUeExK+mA|G9sq|>tOTN)%z)0k)RB&el6(jK^Hko8I20pRCU
zm{3KcEh<ecrbZ5REUIq4`uYoDAvh7Uv$HWponzrLEzeH9aQu{cv;YJ)kDaB?*wgUM
zNFi-%YAOuW-2A0$lRcn>=(VbvVUQv$BDl-i4YE>i%gP+oY9)4~_2K&2E7-|O#UpMI
zsznC#5~b6ttPCo<`uZGrAKMuyapRIbz1h&)h5v4F;=c1!M`-vBovv4IEKf2?yh!jn
z8|^daX&rbwbtQDwiMI8_;bKyxKtZAhhzimj?Ce5>h+yv?H*G{zSiw^8T|@dNGpHK3
zd9OQSh$<VSdH$stpJsA&&>Nz+XQY|aJ}IPhy-yp#V^DNm4<d3FS7wyf4np8hBP9}>
zToj3I?Oo5{lu&J!-+PvQX@Ka(y?hmV89fC2q=xi%BPTv`d>q`6AjBZJZ`}p1@%?@C
ziC~BhIW?kclf{_h0!S>Tk8PjdrAwC%mnDfh>BUsXtL8y~`uyK97KB$zOACVCprg)w
z#K6<XykP1GN6w#veuho+s~Nzpbj~AJZB7tIir?lhKZ%M3M>qZN2}0d$U~w_qSBgR&
zwhX)8KIR@f`o27}7<J4BN3fe7AhpEQbv=0SAOs2CMx_CXK|)Yd4_ubdK0Ow(Z&n;{
zS{pnK$h1w;iZ5!%r`<k>ONwtdx5_x#Mi6!Cl}C?*S0`ORJTr58)*JY1`)kv5)4%b9
zV^I)Q*O5!^Q}PRiY(&VXO9zzD%ZP^O1idsdW*C6c8)z?bWN=9NygGEk!)|8t1G;lY
zCx9I<zd(lZJkpw2KX^|t2SWcv|7G4)cVL<i0<S&y(Xl%uy&H#7t<eqE27>`spz4?M
zgrv`Zoe{f(Ybd(zox)DdW+l7m)}EdUy`+XmJdZ`F?&-s-TUx^FNio&K_2(oE4(wjL
zvys&4;t+~52!Tko#Hvj=VHqC&8{n}Pu`>DDB+|$zDG+2G!0S2+%N70U0^s8yMR9wn
zay?5?(r<rCE(nW3gt3kQP2=c!Y4w=8Wz;4l*(ooC9kH>onFHDj6Jj;{N5}L5ayV^i
zInxm}@%*Gq?veE29x=9elCb;)gep3N5HF*DHa4cHG=b2}t>Ko8zQ?3R#a(AK32Q62
zuu2_SBtKhR%MlIU?9$Rjko}l=-9B;BO&m_1WwvRjD<pGX8kt;(mAijj6X=<H5FO_Y
z1eNa7qlxxPB+n5)&($Uyw|AjT3m;#BVe1Y=%5WH^Ji(9v6C?q4G#HCn=XWIF#^v8n
zV+E*49ngj>4V!C?0e}Id+Z-vlrTug8c>H=6<gk2Acsf7Z+QR!ZyA%~Xo|^goWL!60
z(hk(8e}SsXy~wm%)tg8TF?ILh@Vu^|y^+<p@f%|ok2K8EKLJs3umachZIU$2&@_DC
zi8dL7K^X+1Q^)4Bn0Tg6E=e7v@#<lMDP>X7brXnMThr!*Rjb)tm<YGL5LqZcaEUFI
zhlU%#_1NArQ)!#uTzf2ixJ}Gy;zXMf*dgBh(b^Kv+!JkPPFDehVktD*E20frg_b}P
zZz&&rYYiDbeGpDcSFcUe?@~3g+@HwY>C_0)HF;&oW*UAdrY?)C%Ji4cWG<k#s_36N
z@DH$cqMCjv-#^Zd_L3r608)JwZb23zQ-cplSA2uJd^w;1Gjs^q;Oj($ThLZ6fHcB$
zj7xx8YR^}wCV*JO(VHh&33^?K_6Hdn9PeXYMlBSR=L)?{;*Ps`nL_{aB9Q#_K!}$P
zXh~CLOx;gtLjLtiZ082P8hbb6<mt}xfEiy96(fiw%z~rA2?=X}FC9oDAL0DsSF0`?
z)%?j1*`E6oAFKoGX&Dk$t7QM+jf027;pRsHw=6%Eix~>v!DR!I&OtG>uI}!a_Wu3h
zGVq*d?#IAI$3%N+^GNS+2~lzJDlRTASLeDOgsg~$E7!+!cxn?{Kyxb*2v)C2{0PO=
z>b6M2B&#T3Dn&-A_xl5S4_mDofukG%_k#6{kpNN`bPWhp^UDQ1A)loFVuTP?cOF@D
z&(jqKo0r6L2_hLq5F3*R$TzCaHKcESVrb5Kf1;B#!is)vW97<1Q3ae|G|E7V-IEl#
z0?k%<j}G=7ecA)QgL+O?q3=fvWYd#E#nGsErvFCx^4njmxhQF3@PQYX8{A?Obom)>
z<4#gb4H2U9K_Ki3`<fM56Q|y-t*tFmcU3deNc{Gsu(?4WN&5`J-j}l~XJJ(&SREya
zK7dfj-0z+IHP=R`uR!~QwfzHAuhQOO-VV6!#*M#INa^3YrsH1RK5<*I`*=`U|E7(s
zYi1q@+wTxKC`s!q`Q;BypMM<MOxH17D2yyu%7^v|5W1ej?)D2;K@jc=V}nDd-FaK#
zJnj};gi6!rBb}tTjHcvzkg9rh*ub2n5R^d@+F`7dEN8)tPo8PhjAsUG)0^kQ&hGU^
zBlRvtTX83$EH{!=3oSFcV5P0Bn21Q!vEc`egMe}R3N9u`O0Y95AWRQ_T=;>N`s~o=
zU@y3=3Bm~hh{($kGI%d8-phMZ>Zf&xqal6z98?`-7~u16)XGPjvd{CYYY?&UA<}2s
zCQAUVs?uDA%t$(Xd?d+!7up1lV;S616P!e_!!Jg%z`EtMcUi-Ack=AQuWHz$iF@q4
z!Q(%RfXayY#ytzJ;}eoRizleUYd$&r^e7nL+u;q#a#3}s5i{nw^o^5w9*M~#M_@{)
z=OAVC7~@h0=vu5sXgJ0t<Oe~0!(7|qYHapE;PtABvy3&R&YN`#Z&e{Fb5|559Oc*t
zeV3kO)@Tbg|IXugabP<tP8#&K_Vr0Co0A2qUTlQ%^Sd8)p4^>=pQ=B$(`yVA8G*J*
zs+j`(#oNTy_uhx{7Mn5_Vqm)815&R<v*_1>6bYGdr$s%YuhrFN=lymy-r%lBXvP=q
z&S;smo>+rAfD#-)CAx4-jXK)e>}1eoi<N{2n4#TB#<qkRKrkJ&wY8JV&oNbNbCD@Q
zWD?~Wpp>NjhD5rOPVz-4o2cG^BK4%<B2+ZxmMhm!f9()N9<YALxw)lTY4KKz&d3PX
zFIM(d)~H)m`}C+mx`#Zca4*;&Q;UtSR(~`3a<gA<q%$mv9u!*Kkj1;QE<7ar>bo&L
zu0b#k{kJuDY(#sub0*z@^=(7$!bKjWgryV~04}x;MHn^)l_Sfcq{EF*kaTLdKL^b}
z-S+HD=gm}_;Y2yKG$<pjoiLdD8M6Lt+>_TfR^>PJ|0WE)qAf&S2--z%k+*M-GMGaO
zLw^p|uRq_?(bu;wDHxy#XZc*N@kTp^J;Y;hFtQ264fOLu>>>)#ZbHn|`%9Ef19Wwq
z9PJqRKbPFxbOfj@f~8Ncm3=OZ9C47KX%=J0?#i09*Isz#2eUiQ!dQWI6GR4Q7TJ;i
zr@bp}YO9RGZ`gE1paCI6K%Kc@gkh5+iy~Vm*cOmVn?S*`HDjQLgtdhd7${6im`3Q3
z%9;R!wn!u_MYd_!ZJ-kbWCp_038i4EMIf@4CEyA4KlsJ_>CXLh@16H-&-0vfTtu%X
z;4~c23tUa^bj^i5f)8mB5n!r_!CrSfPGW^sy&dBoSIp5RY#Z~<P?T7unLUk&kbU=D
z2BZ{BA*dqL!k9Z`h#qw`?Ef{s0PGtw?wM~8i&wOj7a?B=@hddY*METfn}5!-?Itr9
z#IRcdEQtpM&SYGiIgz+`9-xZ#-3%!FwwC2tG<G5LtV`umabRs;f2e~+yD@><g?goc
zL^#hTJ&h-wAT@2bfK=`x=7Yh1;!!o~#pLbY=RUQ;A|cbz{NXt??mPrPnT$K2rKP<d
zm6p1b7xBmy*6cJjPO)Hy8?Lw3DE~K>MPszM%a`ias#j#5nm?2Yv=g8L^6P~<O{tz$
zk)hTu!a#$%SDzwh7}<XsL4vQMY<V(PBET%?hMS8`FgugpI2KEx-FtU!KM&ZlWX15I
zUy(V1!||cPfV!ZGU}PXE%jBq-TFN8xqWe6<4n1=l`GOiAVGS!$mLR%EQXO2RS1w*T
zq=biigpkug$0zTZrhO!`%;*tXV;}E81EFRDOv_fhSoPk#(cC#|2bu*iPFyb@TZd{|
zI7iY{yz?fB{nErC;vfLl5+%+THv<1{-MyBTRo~G>>aRmUqkG~%`#p^9d#wUS7kFu;
zH2`QBu}TCFG>DBih_>RTT|7!a($C*rur$00T|!0N13&VJ>GD7#iNEWVZI5IShhPF~
z!Gi4+alzIQ;FyD{4avO)Hy7KPPU?Gnb-ORDIx2?_u2$?}%qcD7Wqw?A*86CLGO^5a
z>;jybx+TMsp2Q}q<IgW58>|Xka-k`$m^%bVH;Tanx@2;OgVK?}5|cx*>e(jD`}I=k
z%VY+N<!K8QA%;Ox-I6ialUea@icPfMaS#zwE)P7XQ&oI!3eL@+tOIU2p8b8@PozOM
zkw{dSu}nXI(3!jxSn&92NzK=03)@U(VI1Hcmqh91qC+~QhVr<T_`HE~bAZwakak~a
zr$?f@0i;^kj020r?Mr0eO5Jrq5;93|nGQzI=rs83w$*d_3ab<RsParSd)}VG+=yKX
z&HXV-GAf1~Qv~g%v&bHe@nwIGShZ(vjNCq!8r29X9lB2jsn-j*5`wo@uMAsX0o8~c
zWLd$?!a6CO9rYC7M72Idys!9!!wHF`k7viGNK0)Hus5y99*>vrs78WjoceD4<I$e5
zyw1j%0~6WPUB(f#yPnvILgP)|eod6_lzkRwt{o-h+P<YUo!d#BsR52_8)!Bk(OwLK
zQoqqzx`|?WkFPp`gyMFRV?YA$sMC~oXynZX^sH%IKf>K+@;El6>%LrFjqqYj)_;TQ
z^>Uv1gLxQGVYP*ZY*7clV%v{N?%*U-V`It>5ryid6P4srp_;$(fkK}l8MOl(6TKF2
z1xLaDEF|dj0Ri7;F*vpRqHTziqpY>&Q~0m$IlgE?EGX3cOW+;sb(D7a-D=<|t&Mit
zD*GH2OLShDb~V8pRgW7=Jvw;>YK2pqJ8gJZS~@YyR>L~8VJU2bZ1r;mMHV!r%2D6`
gfBh$aAa-!`2uIGX=@b-Ai4k1qeFMo2XQFcc1seB6-~a#s

diff --git a/docs/articles/vignette-09-paper_files/figure-html/figure_ex3-1.png b/docs/articles/vignette-09-paper_files/figure-html/figure_ex3-1.png
index 5409a821ecee916c353483dad37e34c7f885de7d..96856a642feb6d735b5b362007a351f926cc1ea7 100644
GIT binary patch
literal 73670
zcmdqJby(D0v^G3|q99=a3L+SEOGt+T!hm$gfOLa&mjMFOAks+pNDLh+BGNDnT`JNg
zG4#N<e>~4Q?>X0Z-tT??{<to$ae&#sz1QArt$W?~+Jvbp%aIb(5W`?FQn<W~It)f2
z4};<BUM2+J(S7=T7rYRfD$2=#SMYTx-53a7h+fF+xx!%A-JpMXGnVvTFxXufT;{Q+
zSK8XNx1pB%X~))nbCUlx)oTwaT;T?}B>DE0)0OSC4lZAs2bmmf*$T9}EmL}yb+q+s
z>av81!&5A5v^Df?O0sQ|8Lc-=9ZQL)eET<}BXq9T1-S@wAQ$JH>Gsn$_orMX3$A*(
z_U-#};KTT7LQHn!6Kp;1z+f0H?W!x#9D^EO23>&u(|#E&1qM@n_kZ=akLfNhGxzKD
zZgBziYp~C5a#6IgnVFd#jBIQQ1cW35#aboiw?f7leW$nQyNAmWqP-~s-D|*LMt@vG
zHml%AsSLB%9mo}ZR!iCOVHhG=Sy{Scayqd>3`R3AKi~F8%)Jy*&vFhbMNTc9<L&M3
zYO@Y93k!>61hFbt!B7$A%@u+9_M2EG{S+S~_r!6=MAMy$BE?ZffrZUL9eexI@p%}f
z$WTziUDl=Q8nU<v`3ZMzN<7#~y7$h4%|O=UhWYmfJzN*96V^N$H=3H0N{{!Jiu&%L
zN@VM}?1s^6l?tpbLRjnqKVov-rB_)m%MRH*vKANQl5QhKqxINV5+2|G_X8|QgiTV?
zI5;?X!y&ZoBkp*=#B+0I;*-fM2AjS#;R_6(8IVfJ*?(vGunWDTC0xcG%?xbnqbOmy
zKssI-$@~KLyLUf>QF;5!N#Zhd8x`!6noM_NGz-oU_^`g9Vf~|yH@%h`&z`-@2t4)n
zG|<=AcSyRJrMta0?jt5qZgg^eq+U-0F~h4~qFu(@NRee|bv>p$!Hy)3L6{L&FEcCb
zGHrVaW_^M9%9ZKWGK1Pj71y(zCj<XxO@xFdBc3LrNXgKrxZ_?f0+!L_3%L&&N8%aF
zlH9Un41VirU4Rwd+}oHgVb1XS+AQw2q<W(SPnEu$F5MhesMZq@c2v}07K1-eW96nE
z#~_Wq9e1iVwU?C}c(axOMicb&=TG(8@pOqcL^Pc^I9x%59~<zNLxyTv*#Z!y8hETv
z&~M@b#3CACF!#4@zhc^c6@EH9J>E=;Rv8Y+^*cRW^qEi8Q9v;=F%<`%oyHkfIrpXt
zs?huI#(r&U6Y%=`+<kLqx*1t}92~+{`$=A7#9`noXX{FBsXfZ(dhzL^C+1ZWqL?Wl
zz^Vz!aHpqR=UC8EF>a=%xnN`I+uJJ`<6%<y{UlvuhN3Jl|B_sLxKem+sDjVhngRg~
zb2IR0D3jm$w@l=1!E&^}=i1oEup7)PFRyU1)CR6#iipX{N1MENG*wlzoaer?-n}bD
zL_~x_lk>~SDW?sz^DQp#pUhS2bJz|sXo^NzEY~AjMsM4sR(oSq1;~j@_ga`a4;c}%
z$^{I|^PA|)@`a<te1t_|xvAg?_$Ab7px<4e3>+#7qlLk+h5-RBByBIal?ycV^z>*%
zJxYBux8^!@mhTl)Q&WdWMRC<~i;5bQ=paX8?n#?hjZ;ui<X(I5ltoy$Ui&CGWCh87
zS}?m4y~=&lxUh{pa%99eWcA0`!-joyg!`^ca7gpnAoV(X{ddi#T|`O>A>mSA-yJOO
z^cgT3a-Q#kt+G2lx~#wf8KbFddMFpx@*IxnFXx9zA)Jr+XQ?7&l~VWzbK&s{s7ybm
zIOf*VqmsDzcmeYtS0iZz6rc_B{j}GYE=qal4rlB2+P~kvePa<9N05cKt>PG|Lq5(1
z9QQWUO8Duvw6tg)ATMG#y=7&O?)rz8QbibTI7i|~IE7+s$9DrqJ(G~b#r+-|+8!n*
zs6BC3+>eLreSL#V<p{TZ#+&fC_opdp$IQf-n3XBh{Q;XvZErGNPq@C$dN?p+$h6~r
zoqxmEuTSns1{h8^dX6^l&y=jK*(;?9sa+zY9r*jJEipN{=#j%jf{)6J7oWhv$ojmB
zd;IwEhDRHZ+hV%s`w2ldwywEDJomxQ5@8*bgxqNjTJYdNw`|4P@wWemt<=5wKMltv
zl}_=NN{sNGfD^qq%;w3Co8t$&;;qiv@=<X_qZhUb=2WyonJW)Z(%Koz(ZoGeh6VZD
zXJ(c=JLr&9K)EmfKD>EfQc{xBpyt_4W4{Uxs)(;&zq*~__U$_37&q933RL!99BF=@
z+2CH@n{@KzMl`so1xVDAemoUEq;y0jNM5Telks)mOJoQ&SwdkYS*B)<(^Cbsf})k0
zMbb(T<HB7+`iGpgQV7jq*iOMvO&eq4zSfx>+%94i$KW1cz0d<W0cE1n4kui2)^CRz
zVPZ2<n#?dtnfBv{yoIIZ`<4Fw{-kzOwdc>XMvoOM^~-#YmJ6mQi;9X;_^gys{p36b
znx9s?mKGcL3V$poIE4jvdlYEUZ;mm-!^wzZqleY!{TW%;rSiAKTj}||ol_X<ne9HD
z4fUrCGy^v${Tp6v^qB3C`Kv6OBrP0W8@H`sKkLsV3t~7G2({VS_xD{IqV9;=4ra?H
z(pgP6xaZMvvBSYWs_N=~z$YM}ynVam-Me?{bHN_#6M1!YLV6rj$bAA*wNFT-<LZm@
zw2+S-5(YUhe<RQC7iipQInaMY7Fr(RP{tG3YP-%39}-OSDOQXfS|*c=+ns2j|MSP>
zQ(<WF^%zqIcdiPYf~=^5i@Xo&+ul|9sg>$YSQc58ei@j}8=t+kB9LX43o--L<+qFz
z#Os%`BdQ%HA303bsfC!Z2ny;rwH)S$kWk-`XwUxp_l5iVgch(HFNrj8?-U9D2Ixm{
zqs>b+#niRCsxm>XMk>dTdyNA$h^pd;*X9MWXipp4--`wHwfj&cStIK`vnjEsil{tw
zMk<T>wEEW9$>vsGxq`a4<5{IlEo7!u-|iTlyK<tp9`^epOjSj|VFEKZmc~bU^=g*Q
zP(GBad=`?eD5$B60&qvmA-9#qL;a@eoD~!l&_Q?vm{L7H$d%JY+>22GS@&#dZe<xi
zjUM>pG-CIwzty(;<AZP_Qnl6{(Un+7BI5mh<X%gO=lc609{#58ixy07xBP8xM(Q0V
zTjKua{5CK$E?Fc_<J@3|1wH-7^DZDxy|+9^LP1Gs7(9#Ld;5=ie<1$K%Mng}7F>C-
zWs;gje?X&=L+GjO1Gy-EkmcU3h+QHAV8wkPV>a++j(<?fV<~(X%o?`coh+4C-sLI;
z7WUkOTB`mW)L^G#L^m;^lb$Ic;|Y;S({E1YI8J}Sr**%92ZqVY@ZBD_!eTsc@8V7g
zER%i@IyWK$jZxauPD>y6)+bXDWUA%G5-e2E6EzEz(-Ruj{HOW+KBOJa7Rd&+a=f`{
znKWM`DQKKZc3M6X+bdxGLlt*?xY)}@RCx9L33&4I@&iYiHP-zZ;v&0;R+WJb=(RXy
zA~H4YOuVBx`m@5%pQWlw@Wk%l#e>24?>>C^FnnX+@t#s|FIRgnn&DXhV}>Uo#^T*$
zuuNEx=>-Px0%Lg!t@{7(Z$B(FZng}KRhTKF#@9W1llkPr!oq0yEZ>ck>S>hVU_sw|
zJJx_5A~isoH1hg&G2OLKIBx>%Q!(%@2r^u}c#%fXF%2IdKOrVYl^yy>Bs)O6(UyH_
zdJm4%HimVKhud!&8y0i^@6#wA>@L`wt`|bI8y34ugO!u(c@qj;a?y0d)1I@4vbnRr
zZRC}@94oV}fkk`iqMmFbBKiRV0UI?RKHSPHDCqw6!mF7c26L<+fF)eJb*m5xU^Pf2
zvIFPTu$<3EfBY@%MhdS*c)n6<zPRsR!b*2fk7nv42L;r{OP3Vmh>3}tX6Pv>{y2Jy
zr02by{U{xZ$(D^^%VdUS1wf~!>w~y&O&Y)xHe;VmKnRX19ocpgc^e*%78o_um^W+1
zju_C0`&5I-*uB2L#fr=6<awvR&b=H;AHae_Y<Jy&Wtp50bW6<Zldu1;uWx81{JbJ*
zJy%D{y1Kg17D)(g$3S!QTwXptMb!M_Vz+4p3lGl#$l)Ne`X8^?Zmi2jP(k2&Si%Y>
zg?j={r`PV6OZ1qribsx9h1k{65~tagIK$cskn~&pk8J^t3XhD`dipekmMql%Q7VJY
zXgMN0A|k%sbjv3YP8HGK9L=CtrX8)53MwuF;7Y1R+${%B6yI}f2_=8&N@Qmj@B17~
zAn@X+)cziDRcseK(i)7qQB#g+WD^k3BDh4R>ETh6uaa3FO6MaVM?puoF&qGcL2)6P
zibFT++cz1M-Ka9yu<}7?A?*@W&saVL=vZ0dppH?kvh3xz#jv=8F-MN@V3DbL+Xyg~
zHzv1m40+m%;)|JqE&0!q)ytAsXtF`=LVAlwDxO(o1RU31u1kM^Y_025wpM*!<DKGA
z`ua!kbk;HnPm`!0xUpBY-6$7!X|QdL7X~Y&1RfnE2OwluXXnF45*W<(B3SLu!Go<i
z_fdpr(Wg%XQ1$dzgvka5jEWsKwi~as2!H!l6@Y6@wat(VJ&mZ{2*-_L0$7&iwHO$T
z&i`bqBRo1fiQ$gW3n`EmdFt=OU`?T*y!&gtK7q1Kg2AN9Oh~Rnk%Zy_I6=j0;AtXm
zOEKdXFxXE4Fe9nF8vF6=*RL-C4`SuzMNiMvyISSt<t6J<!(ajt=R<AIG=m+}=!m+5
za$82L^NnHGCv~>AwkBHq8^b8^VXCela`C`Qs`@2La&s9-Z$Ek#2O#Q{e>{_tdJ-3c
zKj6eGFE{st7MPtt$hlimaT}Eaj7WTZyuXR0WY@-+%7R6cLk5YBj*cd2EBrKTi|TT$
z9sOTF7=(_2+4pzE#~SMD7yyO!OE1G<o6!0`=MddsVtMpP+is+E@JH-}`4>V7@89Qx
zYKTo(Sg$*YdlbNk9<Cd(AQ~u*cu`!t_R(Ro#^&JfuZFGfw;ADJHF3FQcSNj*@|A2S
zYs&sv7F5zGfV$OYzAF*(kB#X@EnVHhr6p@wIk{8>4OmV%uux&;r%ya@-n{ALA;AHx
zHqq#b>=Rz3t?I^yWqs#?gY8o1v+8p*<^<_qaM~??0%cbKLI!qiTmx90N#`<5suzkp
zVm|xRp4fxAXuFdGFdbmWrNFb}Oz33q?z#!spk=_hdsMFgjt*h%82J68{=m~50lSex
z*TufLT;Q2SN~w<;jyD<^;owxO15XohoD$;ez5vU5`r;yV*cN(I-DkWP>uYav85V*7
z(@>krZ<7sNBSj5bO%$}>g}^dh0QsGtDp8NMfnLGcwT=GHPLoHXFsWDP2ITkIwT2vF
zvw1%QfFV|ZWDx;}^KrR3a?vYLJV^JPv2jTRECk7SXk`uiGroynM<o1@e%PW`aQm}K
zATTnF?!=m!npOwkeBw*MBrVS;8QW_)v^HMlesbUpq4?SvpOQu=0R9wIRCHuyUjMJH
zVI#VH8Pt>mU>#Eq&d$!2`iE-xFl(WMB0^Xp<~^qY$OTzhS$AchK5dv3_1q|}uI9hN
z{-55nKHaDTKD9Ddk@KDlp}!KYHrwJ)Nl%YlpR7&Qy$8$@3I|4^6F*+7f>6>YlQxF>
z?+oeuHfRR-Wwh*EN3@QDLbx6y@Bp*AmypwK{ccxy_N??%E2w7mD@<iMwJ*Se_OC*|
zdU$Z~g9n`5GVPKz58wat15#QcRX)p}!cUTnjIzx>*y46j$^4}|H(Lsi$xBl+Gn8d7
zO(d<bI$5Z+jg8GjgF94wq5KYA*2gM9aJl^&vifoX{b;of-yb)c@NipG4|p7dv9U2C
z5Qj|>+G1?~_U!|L4CX}%d6*ioVvWOO{;#&s)$g~R5*XS$IuzqTSjYm$rwHsmllB-7
zt*FN{9Yg0Zu1Lu&!5Ll(3I_PNWfCKtjU%Du)F-UG$pzJN@Y}cj`}Y*Yb(k0FdE_oE
zWGOSOR|gn(qS?0=FcP31ECcRRrD(^KkuLN&$xS@Vz=@r|F@g8wB-gIoxvPKw)Ay)5
zWsHgv{EZB-zcl!eGlLBUuFnqaVPaweM&WDBCA1m$-CX)o;m%E)=;ke0Ku6c89b)ps
zwBmeO7vuS1FtM)g?xB|#uO!6Bn|ZL3>U(&UKauV&|H@aQRXNTO+GgUx`cKRYlB=@n
z_mCtRS$0c7D8JIC=jy_Q(%}8He-Uv9JROh##df3RKG<O$EO6%%006dkcG#<o!DbT;
z<H1y`&QlABoA(>1%QM7%6<nrWl9G}<zjuIyjxlg*&I6EyotIa=cFHBf2fH#)H)L8n
zDd03C=D#~A2hj>Yzq}f5_VwI9Zk{&QQSwrlKvkWz4DKsTH^9lKxGgN;NM((pN$f_w
z0UJw<QJ@Afaol(Dd+U=1N4HK-(HCzjw1p^MyEY?A4l7(epNz!8&sQ{}+G=XC8?9#n
zk@S*AD&p7#Fl{U{5T^)%6xbv870oxHp>{J(-fS!^19i@G6rePC{ra{3GO{F%tSI}V
z<kPM6XfmtAV<BZ<C-07~{-hCqZU>y6&$|1l?gN<hJ18EqiTWLUBBPh!0H+IOJ^19y
zmoLLUXQ1>SF4kg4_-(bv<^KHn9J?`X0BooQz;I%Edf9M^jverg%5i%Da?o$DJ>d7;
zFl1z4@M><A@WLq#kC~G>H8H+<T{m)tIM+E^@e;Yh@xCIZI62s&C8-6r!-JTt72dP8
zSqgKHYf|E}Ackw&LLuSVGff{KpQ1f>WRp4^4j0*M+9`HR1A%-gQOC&HDO1S#$!6<W
z3)lq4PRgw8oE*hS=kTyFH6Nb_fJa7;9T>S%h0>MFbSu6#<L06zwMDP`NGmDFJ?DI=
z72nHtc+@b`mZEt=7y7_BGr;49&WDvck5F#)rgFYU5{$IBx3|wm``HjXcXRH^nxQPp
z&ddYu*sCvNYyMkxT5`eYFyka-{^EJ!PZe|;p+Eh532<EyC<3!b0Vr4FG+RDiWmTe4
zpnNo+WL!)er@S(X%mh}#0CVDgkq5^7p~=MLD=pK1XO@OPejGe{{OYB`SuZa0?8K^Q
z{jCU((4v!BS4gg2RsY)ix`+F9QS)vJA;0wLC7uvtLFC~#6c!e4s6M8gaCOdAZt_n0
zvzgQLJg}<U<za454~LR&T4^5p%;?HIed1A$wh{yn#UTC(WQu|>uP!CoMuXI!DdEp=
z_MJo=i_y@pdtri#N=->A2Jz2s53rQmoy^$Z5!?!KetrsP*1RM+n3~#SMIV4ordL--
zk)wcC)&A9jntx`gk6fZqk$i8$^O#`ZgVKkCcOD&bMbF0)EtAw+#umD6Piovgb9!UR
z)QcAPJR4@Gmp+G=0J<-?kxQ0-3jFgUwqdj7xY*wiRMjA`lhTRlwuO)m0gr|1+Mb>s
z_2d7JTFbPjy%*Cr8~~p@H8o|MX{EHoPuPh(h>HLERl_0V9wlY{NLhh6?(d&l3~e#!
zT6q77qR$Sg&G*u9$yv)Q7?e%{4`ET%l>_JKYO9EGi>wU2I`ax*rdhqiD9cLu`oIR6
zs+)OGfL{<gJJ|u~KzE_58x8{X;NYN^$Ox!wN?u(eUqJ?nW2JkS-MUOGevFRlR8&-S
z{rJHmKxh45(Xju|qMjWQc<ULLKQ0lo>Y+1P<4_KxxcXCHL4^7P2UU@@+~`r*jdJP@
zwg%UBTkN3l;*|EUQGyl1jtr#vO+2FE_@sDt0iserxhjDxgTJIhn?1bj<{M!!nkony
zlS>?Af!vcU(Z;j88F*#@v;&pIsNpgLL4UA^6OE(Ts~I<#m8}<gl2@i1YXMpZEbK7K
zP6yEST|6X!o3}aIU8dnRf6L$v9QR=Imd&?M=<Zafx+BN>c5K6oV7u1P4u!MRq>?h@
z!2XUh=EJK&g2PFh1V*PJ?U_I*p}QmOA_M#s9HNheB)77za5t#}?(Jy3cDDNY7kr1=
z7DTkz_FG?)K{&bO$%|w&l>NB6io{_q0N~jG1A$Cltb;UaSg)C!I(~OsP#KIx!}B~S
zMZhj!E$3+wa3j=+@o1tT;qo88NYhD$_>83YT(OUwJj5ppHJF!cjMPUJ*(E)87SJE1
z)fe!JhJ^+6AHguUv5d_4Fx#O2!2BSnv55%3&7&b!BztWGQ`gWCc>Wotlp&^P))7^<
zHCJm-U4GNJt9WZ}q*{)i45%M+b?xO*u71afYnvQAjU<AK@jQo-XSLwOTsIkv?NX-J
z*doiZQlL(r-;Z2<fbjf?*=J<kuiHJ%QI`rUs=L(8gv(%TKK|t9V%db>SiS4JoI{`M
zFPEo%4ZXglzUkH3(H7|#>Vo-IuY%%q(7>ux5+7_3!v3;wbN8=Zr+EWhd0%tVJvkJ9
z2IF@lldja^NzAW0NPX3&b}B(;wY;}BhyE*=Yvx_<<j2{TiwMot(NX9Yly{0uc$PvE
z0&t(yo!C@^EfRP4U=w+d&|fI4=&WCCylE3D244~{VW|6xPcQzN4wN@}oVdBz+4{cj
zs~S=<@7OW!#2<lYi*7_)N!FKhU{^gnGMz)9oP@yx;8k3|iV0pjrH-4Vw(D-#<ZQ^E
zoX1*zm&Dv@y}7Xb+kL2qiy+Zib8EzURWE>C=skDddd!Jm?j@&K`}W$QL)|<%$E52M
z@Lyl7_NlFc&u7VF@>RxtbuZY@pVu8L;Bk1Ej2L7k3*E(eQgQksi?qvpy4PG_AIkdw
zRNk|hzftfqieY-sZ`ogA#;A6)a8=;rbkRO-SM2qIj0t82Rek-PIq?!mAd|W?05~vz
z>@cDD*0)zazbk;y#m3Xf07no0o;BNO+M)74J+u=6PSeYIIp{F$d(lHVtIFscA&$hf
zZ(oEQ{DwN0g_Wd0$jXX(`|cg4CwT`uy>D{lK6%?RX>O!=n6u$@V%{<-*oR|EP^wCF
zU!ajh^S99|{~csklos`?>XV18h2Qf2o@(UIRv}i^Ov*>Gq72gI-lzwu@+BE@#Or9z
zv!Z}4fwy3?Fc^}!^2k=EyBrLtAod)k&3RYLf&{&MoIu++mK}!4ueTd%s5h2>)!6u*
zemot$nyDG!-qrn21`jx?19}^#$=kI`@$m!fxK-wI;&K)C^s9IHy5Gl?8fzN>A9I;L
zeq8Z1P?;9r=H4z*B=`YyVsRotyWeccc%!LxM$HoGrIE5z&(45~bi2r;ms|G;9ra2;
z-^Cq@hKUoE77-z&xAOrbpcwb=chPeQfB*Soj__H|vwR==^f&!n2D)@#Qf(xYE|#uA
zS#%PSo$hY&v^gzDfB$ji%m;CuS--uCCr_UAlwbavvHks~Fq@Zzx@`7?Tv>7jr@(t<
zGxyhp@4IhuL~h)})fuF0oAeGFD4nNT-~A0OL&LNq>}mmXrk{G?=}zXfzyI-`mY-h}
zFE1|!Xq7igWTf{&!nGaorDt6=rdMn^wM^0|Z`whs`mD%^A7Lc1N8Q{hfEL`boI)H*
zbkni6`v02kt`Q(g*w|dwR|qGLcdRQs!VX{kH^5qD`fGz|qyUGjfo!QD&ftd5g{Mdc
zwtSDKcQSorO)`^`T3R|uS6KCHDfi_Jx%HRqcEA**#jX*A_Hcc)Ek2T(<dcCCUbWYj
zDL@I@V}NH01jr<tiv5XMwd<xtk!<O>g(lZs8Ch8^zcD7&o0OR8P_4%K3|0BUeMd7!
zD0OddZ8=8^vqcZvCQdsd+rtN}hZDD+CKhwM<L6;-Ka?5iulay@nuo7_mOUdlvxb_x
z0Tn0zV+}IHbC|q{ar@<?*K#nvT3eBRtx-dw8;$vW-0bt`ONG(#t#hV=-{k8BPzBfm
z`Iko6B^SUdG#Xv)zN*d0%xrhKJwH)l7CLT$&(%Su@IbU-<m{o`(VPuA^Br2w^dCud
zDU9x|JDSHOIeEjZ=u!uDuZUuLa%(C5TE{Xc$>S$EbO)&|>3a>^I1cy4mrEOM+oZdr
z>z*?o1tpDFhm-FUK_t)p`}ZO4os?d}P;ka8hk;c>!YB^&QTmd@L^T|gGXfP)vR<P$
z<)eN-q);^Gx@}GCp>tN<3QwA5I=uMa5xKDyM3|3g^f+3>kz=ug=|^*dP9G3`tjf}j
zzG6izXR63RivL4-L!kVSWO+I%Zy;ZN${jtySk5lRh`gNcCZ_iKM-J@YUH~}-e<Pah
z{uE)O&RV#Lw)w_7{88vVI-Q6r6V!O_=LW2nIclVD>F)og24cwaJ08lL^;E=V#xxkN
z6Z0FlvQKZ1Rvz#9KGWZADnaU%R7n~Ud6Ti4wujLOedH759V)lU*M?z?pP@lX2~vg1
z6)Fx@W#y>xJXnVRqgkxaaJqUyH)3_ON%OP?M;DJToi}=>&DTLN@cg`J?adH30KyZ*
zQO&!F0J@7MbLgHwj}kFiys3y{0ABgSr>&cDdsYe4v)ob*o1`{4IW+fdI&KTbe{0J0
z5~zrOe+yfgK$RHzW3`$AIkGaC6MAyeHItMP7iap!U8-W2e_1BO4%hc$bhWdSQkzln
zRR-H}hU8wNDR+WVgJQVw=32B)Pd{7a>Y8r|vpOGF#TbJN?Te@9RnPyc1haC{`fQ`!
zCx!|*vUj6@7Jt1%@%@lvT7_<IsHGyg5tBv6BQ0DqE%HOAtWQIHeXFT(lwSK>YB1kF
zND~kj9}xII^?niA4EuH3##7P7`QcQs;uxig3V7ETi+f%5n(w3$17g_th!<rUbt!Ac
z46sjGT`n#z)vk-mfH+W*m7N#kBjjmY=x1zWMM&nleTdC9kZUxO%+T68-t?~D9nPu}
zrwj6d=;<JZ^_gY^Kwx5YjQzEW{UKUc%)2s0(8*EVA}IGD1KoHyzdtSX6_r-~bHDQV
zwT!%?YyX-TjE1%xa&{3=iuH+Zih<J75bW##7$r}uM0*+3K7-#QZx7cv7~GNY8(ZTY
z{9JbUBH7J48t{JB7tXngrDUs&L4d!q09}OCJVuQgm0Y6plHzVXsm$<RpjJxeZO<hh
zv=&<VQcVEL2^hw}5TY~_0PqK(w(%>F@mq}}s5k*#F>tF{%%T{duglv=4Y9MMvocm+
zHLe8{x^n&rpr7BR&kPSV9J7DqSn+BU$|pX{*zRg4te<c>^`DgN#`1BLtjW+>{;0;x
zqsHrcFQv_e3sB=0b$U;p^m5@9@}4`yq*E(`U8fx5P&=*dcvd?bP*Q?yX?S1q#Owww
zZK;#QvCONdkt5$3&(ySs#u{$KJqlohgX4qWn~V)<a9_1_c^rq$7CJ;wSOL|j`8?9@
z_VD!=0>}6H6Dx-S!4~B-YxF2p>3DB_2vFvQa&+D@nSrPE;}+T1tcQyewUN5{fN3cn
zKe<!P_rTvJ%KB-vs{rrEbPjjMYCIvc^W+1dJ=7fHv%gUWF~^$5s;V)7t#NOs)cK%!
zhbHIk+sj4nLE{z)=xh}IuglhVp_Hu9-rjBl_|=IThg6WEpq7J?GJ_JuBrZ_G^EHYA
z#*d^5Fn(QLfI#RjHQ=BE{F*n^;DRYop$GKMjVM4_lR-i1sk(YxjpKCj)RbZKE?Q9p
z<a$s|ig5wW=>p`EEDs(&RD)*bE6T!Bh~B>?MtcJ`0E67!qUmJ28-mj=uCDHr4z)Ib
ze+DJHWzt1h&?rPWTO1r56oTv;m;3k?q1FKjkf>;A&WWlcP*1EHNrH!g+S}Yz0Fb$$
z9syiF(6?Fc-2)i^arnD;Znd5pQ+j~H8Wwp!?KnDB?<xRVPDrVFWdJ{_3Amo3Q(@Ny
z7BE3T^9}+~-_<mmdamUF$n)3wzwyxOdSwLxZoB`<MEfR2?*nOI3DQ`AOk&xak_V*3
zC6MDi_EpkFv_NOY&O-0dpU!v`$ghAIE<b1cZ-ZI>Cvy!*>Oz%Fi99B|5BxT93;mhM
z^BxRcz|tOp3YXu0j0d9eW@n8bJbgC+Xk7Q---s-%tn9EW!%$fSDyEk&U-FxNy9{!l
zV%+HQKZ=r=ev?<l^&2<X*EVNcTeVAd)$+`MU+My*5MRFhqW19kSkXoNt;b?tI@IZb
zg-!f-LCF#bN?w2F=H|Aj<+Mw*JQ|x7I&8st1mcEj@%oU8<gk+9G(%uuAQ0_~0fm=$
z^y<|s#`<yz@5a_YI~jmxI0a1y?<b<d!ivz|_^=i3^XNe%^dif$=YyI_ARx8?L!z}l
zM@2=o0&1}l&@bc;sNIH1e^An|Z|uHUSU-)MsCC*j0<oz)J3E_r|00Z^<lMjiFY><}
zD(Vna@iw5aJUX5|D_7*I|29qc9~Jw;f76-Dw0kopXPeVOoB<`^C;detSYhCKwEsV)
z`LgXTb#-+hlrRAsls#h!`%m4;m-j93a#k)bWuP~$;$Sd{=*v=9Q&ZEgcgX|NKt3Ss
z>~Kd|u*6J=8RbGWXdViIj<>-|i*6`NBoN6#!czIaniG)t+1V9`3RFfxH;!!{5d9#F
z7WX*TcK7$I+`D%V$To$0$APDZR`c`oMSu;qI08X{5{QETqw>_*oLLJFOC)p}Rf_<l
zZ~gb@7jV9GiuPujeeEWyZL~6`*TM0|fcc1R%q=b!4Cct$0g*C)7c^oCLXx9bwMiQx
z!hKk)VXd-zccXFB7M%BqI_DhUh7Hgjkw~NrYBHVkE!+KeLj~#_>S;3tTAL|E<O4O#
zXX#^TwRz|Dcs4-AFZ`nyBpC<jZG89`4;AbGU8Ye3nqgv6l8#Ah+b>n{B)f|Lr9Jh0
zc<8fM(P4h;Da18Ed^r;cVf?osZ9Czi>%uePbY-}@w)XoOza6DAeUyUdw0jcZpYH;i
z=G{9INYItU?Z$({Y2wljX$T+zWW9sZ6A)uI902nOF$I9()J*mVU(ivt#T2Tc^uQF1
zAbld?mrc(#5D}@(K-2>A|1Drw@D&aPvH$P4S{{(>Ama7|((KOZ=qz=E8hf=TPk?8?
z1T53C<K|R-6@a+UN=aZrDG(LuRqeTH1R{75AP4jt+=@y^G$cii_SQ=d4!pobr@s;7
z!3u>XEN=o8a5BSxPmht25t4A@7*cx|Hh=&V*Wk9C4MGiIyxU8!gI|7lzD@r_A%GA2
z?0fDEetv$yyl-&-!~-hOVfpV52+3#*srf85$1>yMvW=R3hW`=JAVbh8ec*Wv@{@j*
zW#rQ+T2|0w0~8U5=?1ON&QARnzj}|gF%yquVDQW585g7~fS#4wLdi!X1!q{`=Tf8l
z(!gV&S*rtXN~!cJumA*|G(oAX0FtvN*i(<qnKICT1r#!N5s`^CfIJJ;KxULm1PU1|
z2S?9;F>I#ab{M1t0_W5c%9#dYuU$PG$aQp3t`kcUcFlKeSY|3SZZ!fCn2m#j!xImd
zb?+xJq!BD;_$ou3s_adKIUx37UUzh$B|>rars0cVNZ?duKR1<Q-7i9i@yhmpRM>M^
zx>UcIF{Gt~$b)cKl95l*2@cmi&Hd<6&3i~6S~yo%*JoReL&@%_o0?{K;Mh{8$zVZ8
zuc1V=G$=<uob~vX+pPaSD?h&`2r6rF|DKl5IrM=iYCwi8`a2>5(&P2>8VLhjnUaFy
z0|Imrqad>3ul3j0L=B)9QcF|w-|3$#dO$RK12L}&2^0v|UI9Q=#}4c7ZQ21jkM|=O
z&z9;uUxCs#F(&3NysNKoAsOHY!pQS6|C{Ffnd0#9kVeGKI@trr6#qT1+M?)TKmh=f
zpz3$y)9ss3*G0PMXtPmrz3Za(@)(G3pj8c&vo1ok7DxYMpT$(|p?4!R$~iCqz=3;1
zRnw)Ji2+o7ps^q>Hzp=V-G6+1T))`|1vZ%@wKpG7wvo3UJ^^Z_7VxM4NRM4C!W)J8
z9p`K=;7Uh80N&jlQb}5wtamK{ls#Z{7ab%1D{248ATom;mUVg$0>RKRc^XOMK60E^
z-c<CJkB@H{G+3xR01huM_vK4`<o>0L7wad}l?$NGk$n$9{ew(W>>dTS6OzhVa{i2E
zLDBTv+iRe^V}2tQ2o?~;oxm%)FCmc^$WI-jR)S;X)UPhY>~->7gjv@?n2i^-S%BU9
zGqZ=qV(ZR83f6p(dHdg_<^XID1|udR**))akUhg~22M{x2cFLRPi%^i65yMl-kyYq
zzI*miIE3UyeyhGT7pJiJT%e;t+b^(P|35Rupl?F<9K-p7$O{~!G<N3#41<8&j2f8G
z1IcYzrwaTQnb=QCPObw;=0uQBg9$FNf$kS5ZAp`hYl1Er8uG?z&}*&tYrPHC+tE?b
zhe3~>pMC$oUoN4Jyyw100<Ap`iv9pl*?<@eqzjEQEP?fygyGglCNYu6s{#7V@NkF0
zk*Sw|Xs+lju?%SG80DR%9tGM?j+~^>@Ij%30#rsiph!yg077vjowz=L(OmofGaHu%
zUT_+IQoTX%t*p_YGI}^_WbGlGCqxoDqiCFcyS(XfX5eu>kcfM@iUuKRBIx_#*khX=
z+hwnEw;2od$q2v>8EI>3=62BeJlmW(sn92^<!wykwWmtmSQ<0#Snt5@&Z*dqH0fC;
zVXUF0NzH9r6z|$@QOZQfvT?*X-eQ|RPBEW1`9eAj9B;V7ra*)`N}(agOtQIVO}47r
z8q<{-2l@=M?oHsTcDLn={=T+j72={+ws8;NJ+-scxx@_`I@CTG*8y&ZZGc(+BGfB3
zSoqJgww41K6@^<*$FAg&8zFH%K`y%A>^&@hQ;EW01D`1VC?>bUk82e#@#f;F(ZP4h
zM_-9Dwy~V~%kgzP$-UbDwYTryFV{wSV?y69)ZXn<g414^*4e(JPUzd6-Q~|ni&)v#
zMS9JX$#h0IayB8?X#F@r*r*`+p@>leH++KUw~oLSdoK9I?lB=&-|rQ}@_Uck)v^qy
z(eKM2Q=s6Ob=uh-M`q7WCOH|=rn<Eki<@s{7J6n&!+A&iE~Vt<`UF=IpIS?do_K$=
zl0!<h*7phr{l>*B?$OTTCr6of9YsI%au<=_?RQdU`gM!%Tg5UgV@Z(-w`11&?o$d!
zn4@&DLmzt^8nEthhQZx7Li0Sdm7?n?H;+Ota4=wnt|V;(KxRsejr|Dt{r%NW_qo5n
zwr6i$V@wZxLosb+>+QH@Cpu~;|2n~6m&P^LZcn6O4s%@k$YzNT<?2%CT1r)tzxSbh
zlFGP%xz$gm<a+Bcw_Lyo`K6)V&?UzsX*WT5+&yWe(}(Q0LHP}A3;LXjfpU=1%vB?J
zJF4~U8+aFN{p+m~Meq7oo4sl1Ff8f!4Ad`K3|)c3%fMyo5IcN9KH#G3<5dNkFt)>#
z)Kr57ao~Y~UK;%6+K@&pu`ybsF7<Yi+yVV8<)C8sH;+v}cX>_F)nz0%oTwIa0?(eQ
z<7az3TC@Cdxbd_Ad#Po#x;lvg9*-0i@a~8*qn=2uG9YIZ%V#$$AF15+toA$hrEwmd
zwuoBfp<TJ=TeWZ^gqTBmJ@qI+7_Z~h`;D+q^6)xeC@_yC`@Sg}_*LjvKnS&>NIj@>
zeziN}i_VFL2M5pKN<k)G2Y6-bDs;Pok|H6!tQ+W0AbFNUdOzr=O7a0Yz6|6)3;)eJ
zwp(z``VY!z0fIcOPV^Qn>xTaF@74YV{z0iGuXRnvo@S{W3EI-06M}|jX=C?yOMf<0
zrbUe7Fa>kz;iPu#@Rav^3iWvKIIf6|d?GtKjR3V4vOnvdwOZd<5~6yAo!?uzdbp%e
zDMapOLfub6ProiFyl;8X+^AVOp6k%6;CwR^=e)DCg93D|dhO?QziE&I(&pdo69L^A
zD0w<;CuxSD83bkjh+91R#|s6O>HB%eq;O({gI$vRY`|S5oRRId)%1SZ<`>G=c9HLH
zm+7U>BrxEPi_cv5g-Jp!?R^kRCvBJt8)iLowSHHpVL4H4Yq=4R^xGrlJzF9ap(PG-
zwQRU0q%Ga8q7MU)P_De^l|76mc*<}YJDS^@nwPJXcT686;#PQc9DPSr`&ZcuKpC*_
zi(k8T4PYd0SIR;#ecSQMJPymLx&1?Yn@OM1>KW$o2ylgBw*nonXxqw!VGUe%iMDj`
z(FnBmuFrm{-!K^{RQLF>oo}eUW$;<tRJ;JkL`ta~cVJ~t#Y_KJAO<|0T{9@5CvzMq
zTleLA<0*cwH3UyqAX2u&-3~sveR+fS--mxxQrI-RrS4Iakj#3g$E`WTP}1vfU{_}t
z%l;Us#x?ghrn^ZWLnYFis*-$qZFp36<$Bb8)^)Sx%;vl-RaPENVv^XGDRo8!(uy1w
zlca66d;J%UOsX0J#P_NWlWa4O=br5!bv}1P_b4G(e|3NF)bHb~ada3p=tj>Djyy2f
z)sIfLIO;arN=gx@(VRxxX^s&E7M-;E^Lv2|1*2hUi_u2DU+9Zn^C%i!N4t+IHhN|f
zOED6t3AY8CSIBefA1qaNmvh|+f;y#QVicHU3+?<)44JiNs$$Z}Z064TmOlBHX}<}1
zT?mh7A|@r(vbJ8LMuxtA{lk`X(WHh)DR}NR6)x_6YJ2|rzWIA%6>%z~qD(Re^uny*
z%5SQq_uiXYWys-Odczyy-(-!{^ZKiF8D<|X1dtw>8$g_L)~^>rxk<)x0b@%vPMGSV
z>TT)eSNq)?c2#greR<C8+`{XM5iG2g^%$l-eMI1T_=F^p)7wo}|2<=PL1e@>HogG|
zrxDlc>i1A02oRGKF)|gGEU4q$$4D?6uCJHXSEm2a0$qy$EF(5Z{D!YxzwR>fc=gZk
zxi+M))!e|=EOtI)*o8)!F+;MtCo!+F;Y5oZ*`RWUv+=0MTh3l^jx!tzkPA)AS$kQk
zUv%3lD)1qwu%sqy>c7VRAyyn8S9fD)Eq+5WjHY~>t>5rcbtqpwr~bL~;X2x)Id}7Y
zX2Wpb=q&bKt{Qg(9n)=E?e#Y0mMN8HgNzj2ugh+1po1Os4M3UA!()%S%*xV|O<cTr
zDvYa`jfdyc)34f~x9`v5qUU#a%*_g=3EQ(-m#N9<vH(@+j0!rYlu5duxF+K+8S@yv
z;=W~6%sGcl-oU>9GME>2lkTdTZDYCZjjQ=*)s2H|yb44WU^4VjJRk%rt<Cf7YM(^6
zpFyisjVEG7N~Lb=?&>m!(&01`5wbpRE!84=`o~cTz8HaC&&*4Rxp3RJ!?s#rKA7Cj
zarz-15&#%8-N!_Hb_+c=rdS(sJ();QG3z{d%nrah9v<GpGmILka$8Xy6$0DYmq@d-
zh0s5;D?WAY?G7gs?pQ-NaZPY*Mx*Mx)_9*m<zyCKrEcz|54?7U$zhhuf3Ik_cN8tE
zXgK6&>{I0CPrjTT(G&RGzN|_&ca^a5z0YURHY+5g3lJy>0v7X5e?Rk(0QjG#H6T*o
z0{UZJGR?HK9;KE&eW`72UHJ0VD|S)?a6_i-z5TOcJDt8_mq_ghE%I-Yr3RDJ<rSIJ
zCDDP$s>YiJIYfbJg^zENaC@qUq&!bM)M!y9fU-FMbE*hTOS2s{zpj4ccgpu8Wv-mK
zBgN)3bC1MAPLQd<q5;#SYHt2Vj~)pcP4A@iF}Z;%S02DrE5p!XTleV6Vip--xDq{i
zS4M{b#d|IHh_QCr&u6c`Ie5c{;nsw{v3Eu5lhY^r_vrp#E4K#Wn=@M#3-%cw7hAmb
zqf1X4-#Z%Q-bDJAO)b)=N0z=cBPJ@giqGWdId}T7pU;=`GtiNO@YYi{|0BtwNlHtd
zH%C2yF9FJt1}Jd)j&wn#BJNm{Ev1$C31!t!H~&Xd?7P*;O>!N5#TciUCw|hDzi;ZF
z&}6H&s$M(}Sy^y(8r{FtGNR+#iW%<QqH*n<#eSnaE3H_(-Hh9LhTZ5k=Inu(-mEC1
z%kR`9s7Af8eA(H#CB;HbcNq>0kt6u>aiHB4v^h}H(FKfjxQL{KCNlOL*)69>c1qqR
z#f~D2+Onb5>cP0zE)gz?1|Kv|Qc9+>Y3HdHSdc6_07qb)WGCnOyD}NtHmX;LtCKLu
zox32Hoc62=(2BVHe!FlcB1cbuDrGFT(@sr)#VQ)g(sx{cd*zeSolJqQVL!hf|L2UV
z2<?(#e>()Y<TtfGY8==MIxTV*nZye6N;I@oHFdWhay)u-7|PxwX$achSPN*CL$||S
z5{whLr9%J&dV+f*u!IxsE1nS-Zv3-3hYWn@uWKB2R{4}^|8p|b`U}e%to!@*ia=ne
z-<?yHmhe4IhofCH=BSlkdZXbUdLtbpwTzTYPghP4Drh`@$zMFUFd~n*S9{3%=Y5D|
zq>n$8Y+$giSj2}iYRP9q2Dgp1lue@U<EN~AeBI%$lhn%gBZ<9)6lyR8*A_ZP+9a3T
zH#0PDobiYU9A1cQ@_P5g55d<{Hoj9ISKCyUKHj>Feoxe}A9snSeJ7TN078S2&jMPi
z%_~61OUlA+sY9-amM0pHUK)a#lZqf`Ppu>KnoHrcCmO&4F&}$doiHbv3ro$S652^W
z6rYve68IvjQVVA39%UkL?8ha3LcTl~ox9nRxF&0<a}nc!6D8-OjH;nm9L;q)Wi0sR
zCOLot@D*P!_;jyRzN$-x^f#auV3D7r*3pwriqy+-(MZfb8OTxP?peHmKqd>#dHOy@
zY-HR|`Sd$J<-Eurep;7XWWs6XIt9MddXC1Hr+7s|w-S$>%!N?vOUNVfv#k5C^f9Mr
z447Nc0!ek1qpGbxR{Nk<iP3pm=%4r0BSH?2Q%&SUG+aa!I2dd0L=7FWjcx9>tMy3~
z`1h@dQ{5qb`0P-&^^wFSEh|qKy_yziHwR$UA%NpUZ|~>5#Ag*VY^ZFcxbXn=>P>9N
zukQP|v79!yzHH~o&B`*ycFL(>ij21-U%I?}r?!VYLd{zYNz|KyumD03x*fnl&=aq4
zoUyK+B*DyFB>(+!L@8COy>_ZSRf*knvjcmDf3Z{d$KI9tr4?TCXw;AF!k_PLsnNwR
z&mI(VpT{35;M2`j2bjGIGxRoj->x-v+^GhERw9zEwdd%ZF`sRAs4%dom@uyxH%CoO
zty0-JHx4tEl$kdwX{VwvHOI}&wcY+I7dt+^mlzq2ZLz(xrcr?Z2HX;uII9Zz`69xP
zl2UtJSe_9<;aw4e5K`Z%?NL+P8opoB&KALu;?ibN`SxrBd*`1dB7JE&oITg6#TlV%
zGh?W5@QQ+xQLn7LUH!xsm1fKC6!xy^*;dkN2Yc(1L78?ytX%WXlt4V?=<n;G3-H-`
z_jF0!55?WFd6kE+5Cs2bfbm@4&G@5JJ^dJtygWF&WCB;CAq~$bAF~*Jlx_ZMd2^p{
zJ3jWEEEP!E&%zeT`5Gy_@Xn{hA)@cIL|^_;u5$;W;agCa<TJy8J&mUT&5Un?gzs-$
zNiY#?4A&^dv(yQSKes3TO+yy(ozm(BMH{tZJ2CW4V|nf-mW!8C5qDINOYcohgl$9h
zr8PsT_VS9IVeNRC?nX*o@QVwuyz}py<KyRwp4xj*!c+zIdV%%>zE+5#6Zr2%08j@%
z{P(R{93f09=YMWcg<9A#*C1LUhxdQeK6;@_5nLn0gJ?Bt{>rK<ln>SxQgnch2J{V@
zho-Lq@!iWs3R9JXDCGg`wW)dy&>RPZE!SGmmi*5(QRzpiz0;1jLIC%(FeBmhH_Ol8
z-wtI2lUj#PJ56Ly&odOb#inWS@10F4FgA2!lizwEOa1t%tgLsdJp9)$gqw27r0slF
zakgHSB#?4p7^tB*OKl6f4AT$VulJkhqC}@w!JS%&s2yElTW*Rnb2I9Ah}zbJ1%XSt
zpU<j1XSdX}mghPoJ&&I8+PoehBD$OPGW)WJ+FB!(rp#4YHP2y`DXwT$$CI<9yd=A{
zq*D9BV{MMYF*Up^Z!c2DynjuDE=Uq}4Ges?VL=!d!wc8U*k2VuG6)$nfj=Mc&))^`
z+(`4y|NC2PCvAuJ-Z!Hh&@7vflvF@_<pMrS_{*VX{-~njE}-GogI2e9W`Hr92VFYT
zWxe38GP>D!KcVBwz1Iipara4|T)wwQ1E|$f<JJsa4og~gc6KwsEwhB<OYzgp$4HZz
z05<*telXKyM7wF4|K9i-;I=R$BQ<ujZ>~OZBO!`;ef7zj03c^>N%#qY2gwfcudJ+i
zfvcX$fSGoV8n&GuGmEwW7bC=W2c(s1Km%0&77&SeUH-^}Pfau)gDHpvU66-n!d5%&
zaogS9{T;nz255v%ROD_>09s*Qnk%?k8<&~cPe3jv1=_=3tz(9VseM_*LErM7LHB#F
zuilZuX8<q0l&QVnfoRtU#6fS+m3VvV?jvUAP~VOE(C0iqIBTcS&qYCB*=$i^;m`b3
z$Lvp^J~cOTzP_6MN<N4Itc3q3T9MlV7@%ppiygXQ&bnWwR~6Dh=lc$j;Q};LM8Iwi
zaWMGhd#|m7#_bm>I_L;~HmsH_vR(>kBNfalW`o)!E287BK6yKP>D)SRpT46LPXSlX
zLV*PG$$IcUOdDy`3-svTU!nAUuZeDm)V>0XM=LL{k%3{Cgi`V&2^Ks&Bm)$|y#a97
ziia%pRYQXa&|$%)p-bKGIqhEH&c}b$c^Ay*s?yRZk{-0KJOGx1t&LT@ir{LHfB68+
zG(C<4{NkD{!z+x4L}oO&Q3ZMtUrjq&0cH-UKRH#6j9~3|<O!kmjJ*PGveCS61j|=W
zXZsR+^~p&T9Qw)L{kX@rsi629sIpR_kUc%70Bd7cwDOb!pE@X80`1vWo{ulzlM#()
z^#CgYcD$^na=stXyFz9>Fp2YTW8%*(2fbT?pBo(dHkLu^`tsi0zeddazi#}@%WX9#
zSm59CG_wur9vVZ2_~DC9_Kp5CLN9|wi)=$Oo*nc47ljq8NqlRilW9%S1^&H~3cDYz
z($KTt8WT}}2%F7$AlgW9{@CBO@cglL|0BGrV3FrH{&HtcaxL!SWe~6hJA9s#3LtQZ
zgh>#-{2cw+BB(KTbpc~+$F7Pec5!z`+PGkNpU=Dds1<e%)T<BqeqLf4K5Jg3XslTx
zG$bf3ZvF22`3q^<Xv$&%9|1F7#42*@!3&Su<y$m=FUMSLQuvKimXid2a`|4sK7-Um
zBMHGHeAi&H=U2m6jY{yWaKgu<nSAt1WE63i$sP@#m5bu*5{eO<<A)M3$6t=~7^Y+?
zbE>_(MC7L*Xs>$%=O6LyZmTc4Vrcc)&ktP@5On3B^mNlr+2EL+heD{jh5PK{`*@lx
z)ZXDT6TS}2D)urD5`O_tu6CP-Dl6;s)D>osa>NR4&sJl>+9Ac3=ttOU@SD$x@QynS
zyix@OSAyI>uYM+N6g$<*;^rxCDdWBV<tqZdI2Nq`^UBY<+_@_&$cK(ekBl#n6}Ad6
zG`2KL!Zu@mRLcYj2HnF723>)h*SpA<a_T?3pOvZ_c$O-akgk`EICFd#Eko!)uRhAr
zOBjXcGfeNazLV2}l#98nDh=<@A=r%nVKn&Fu8Am(P>g7b*h_yf<RBf*??4;I4Tq*X
z2-AC2z0XqI+B-4t^)@}&#TmDrNJu9cDif1y<>oVbmd^3a^y&=b@`qHJAXgUxhZbT(
zsr$z_tbd-0To3-b#ekpt<T_rbG5neI5WOeiU*2B!rUCVSfq{+g&)y8YypQ$IPTUmb
zNu%yw>si5jZf7LT{X*cXo%HSBp7Lr5I<edx>eeY9w_i9}?Ms7b)rl9LeKIhfejpGF
zv)NQLJgdqLb0x%_8P0z8&%;XY1P5bVog98PWy+`1j1-(MimuU?uI>^$C-Y121YSGf
z<Dt9-zfVvD?CT#)$xz7{X#Ho)0aYGscC<S5CY_*nL_%`xaDpj<QM^R}`70aI=dgNw
z7qUZzPK<!OFS&oDT6LJ~E{e^TU|%?nZdUVovz?wjZc2J}Ql%vse&LpAiuESqdh!Fo
z7sO%iaOdm6=s|{joL(?8)jEfh=14GGvg5KtV8AEe&y^)3hGwJt!gp{Kkgps}$AcAB
z<ee^-Q?B+f<bO{mvM-k$sS55`UDfD<mrA}&*9$Xm^E_;9I$iu}xz9hW0w0x(|K`Ro
zUs@%6L+IT%{-y@)npS3l+d}!0Ne<PxLjgIF^1*BW4&Qq|eB8I)A`K5=iEtSz6V|Zj
z-^foYf;)_5vV|VM{&qw!8CK-bnYPyQJ&#tmt?&9&=P6<7v~+6@L#6DQoGC6o;MDXd
z%TJe|Z^Wwyscww>H-A4klDKGnyki&TW|;Hf-vge?jEA0ipm6`k;cd|>vFRWyxMt9#
zT43MdXMct+4|axpON3JP^Uv;ZhjioF5jE_hF#G}=X_<*|gv`e`SKxjN91JhtwZ>hR
z+B5gS%_YVE@c7b)hXZE(g-ZIu@RE-`tR<S@d)(z7HwXWK-q<KRH<^DueF*aOCgQZq
zVvkMJ2k-xU>_I3I!}8IeA}s5!Ot#^k1j7j4`q+Gv8Qfl;hERf8?cU}FVfis%*!^qF
z@hrj!S<_DvG2e_7d@Bo27gymO+1OFcq!V{iyjQ?&dGlT-7!J-^Ixw<0O1oV5RBl%T
z3ZcIRl@ywZK@2Z}Cp!T^pxouR`GgnBsnD@eQ`vib9h@F!;;`#mHwnz-H5*Frsd3CZ
zJf{m|E&T1d1oSc$MbVV}k2YccwMzNvv;E)aoM9BQkJTn5Dr8%gMJ9%h-%ri<Sred-
zo8l%FHw?BbMhxWwn*#V!d)JRLYEZ4~g9T9|n8j|#+MSgWq=#Elz)cX=f$JvFUbiEQ
zo|*R|zb>ozJ(gw9g^Z`wtdU9NWpQy6xIO7$`+z3$_m<ZHKWXh`<g}xT&FtLP)##3j
z$#ru}%Nn20)X^Q_Zy(^@<>p5A5{!0rDR59-0pwDdV@^ZE#FC1^H`9l<wzfRU&;Bed
z)Ew2@q=byFGF}gR{wTGxtLv-%O#K@Z7K*QnJ0Dqyfu5+#K}t?3p!DDt$@R+@-enf2
zG(={P+vgj%1_;qc5l&h=2HM%$zU`O;gXQm1-lZj*+dmHLnETLc32rr4=$DnX7`Yy6
zA`p?;EDiz2lf^~-qtsSugM-tD(=kwTlQH9!x4$+Zx;*A_eG-57euWZ@A>Xd`Y4(%f
zig(5z4?djW{rT=`&V^TxIjFiFzr@beOgnbWZ2{mAECs8bY@U>=6Fv!$q>6Y0S{dvg
zO=i#!yavE*w~VuY_ScK6kxw6{_BwRk2X~$;D=RM(zA7vZ@1D1HzFl1Wd*=|?Fsb{^
z<oZmRigpQQ$#8i8;)AP~Z#`D$r(p!IwW=)U=1y*;7d1dn!pFM0y5Y^x+0737M{^#H
zw@Io?wb=o(J-`J#`FIsfreG)xj9lBJ46J8b@zYIYQv8-(cyaNE-i1ejPun;C+1?K}
zxrB0NaP#<GwZZc7yhZs73cS<GFeR&~xa4O{0MTz9`n_&5v9+`7?CG)JJqZEZ_43PI
zruZTy>MQ9Pq&6I{(1t!Qu0^y@i1qf(yIdoF16Bq0f(mt$gmCNl#u<OQGBcQ0WQJYs
zB+;8E0QsqDXh?HVfn!rs++SP_x^|v`UYv+aTgAoKNeF>`gwp$-_G?hIA2c2KNq~pN
z%I8j3!au5BzRQ_eRWmSqgD4994?ZDX#(M@oI_6bZTqnTbTgB)`UwF<`5xmR3J|ntS
zGCK8BzJP_;vGz7F<?N<lez1kPIY0Z=%NMRs{>@`|y&BOD(@*SF$d@$gmq)$HW~npe
z0(<x6`KQjiayO$pwtSjpa@6Jo99dFtgvk^Z7vBe`j)c&&n+zq(aQ$O$?(5&be^WJQ
zNMw=+fj>lI`Tr30)lpe!&DWqHii9Wvk}7>E=>|~=Ns(>|LAtwDK)O36r9q^-yQI6j
zyWtz|eShm?E&p&WT|Cb@b0+re*;Ab%pB)|-N2oonp{7OzaKTuhDTbV!v>aUJ_;Rn1
zir`BuH#R-7|2zAj!uPep?~RbAf|MT<2xE$zt86YES_^ig2ZV4EV!(B(=SP}x80p-J
zF?cH{*D&`=w}{htW_{yXY6-^IBecQ14NGN^9l#Ut8(1@9vP1KD1DDSt(P&V-&6Y-w
zhc<uYKutyU2{KM0&rDvpMbgm0NjC*KIT!=-y@Mbq?MEE}Y2^p1cVv$lzwVTv-us3i
zq5FsHYq~ysZ;<ASWe(M{eSDjwera1xotYuWB4gE+zpxR!jQEu$s-~uv{gpI+%!q=L
zvQ_bN=g0xB7v4M@Ucu^E%Ce4Z3=NL&PbyJmW|6fdy<jIk5|>(%+b6<#u1qQ?=f9?h
zjvT<s$_j4@7%(9v*R14tXut~+0xDqWGa_fYn0B3)%*;h{BL;y+A9;u=2{E*pqPcht
zNyk!-$zNnZs_5w_ettC}x#4jt@g4)%H=e&<NM?k>6~N*3e*v~Ceu`E!J-+{A?{@{t
zx%vdO1~EeLy*jb~5tIgxqu=>l;H9q)TkHd`MrwBn9^%%O18Ef%->fWPs{##@04)HH
z6Wkf9ha#|M3iHyd&z$%`K4@?io3FZqxG&wuvw%4^EI!^JP5~Pg{_n*4k?gtQ;Mf>x
z5JG14ze#fPi}Lri3f9UDO{Vb6nbU_@BMQHu7mbNS2Y`{67QQpQhv$!853pz}D;j21
zunJ*vjVmV!=oKpVvCxqh7}8vwt;?)}Wx`>7fAb>6$yQ!oK4l>PklJW%Tfx$fhZrcI
z<Fn2#wTX(fo{e)N6SEq<6wtHOD#U}DqGe-P)j`QpP@1cn;~QM<ODu5;nQ^|9m*G`e
z5Ml7=k$<ndO3hS}?^UK^axLs}z-!%NavK~;UQLTT%9M#o*DM<QtK3ebhKz63KCt<>
zSiQ-l^vQjsAMakuDk!Ar<AXOC%XbBPP(JThh7Sqew!(JDgz%wPqtJsbE6K?3gT?Q`
zm_NM|VZk@Hd2e5xQ$Db^Iq0wXBA9iEC3oO%@eK37#n6{FpVS#ID>$osbsl1=2La=j
z=fTNTFr*11D`w@YLie8I`{PG@S&1mP-;1A?tVkbg)?#$xOAJ2iXYq$eD3md<CF<JC
zrSkbR`O~LQ>w2tZ9MY<fQQv??l$V!}ZrRAnrInNnQ&sgvEpGPbYdWZL5Z?Q8fGyw8
z@o%}uvN(`B)B5-D1~>4o?}jOEiC#uizL<sCqD48E(QcF~PrVUa3Y8DzQP092-UhRB
zHgEbUdFcdxOga{Zmw&bQo1Y&#Ha-<9;-mj#6E7+c7V$#xSxib*RowczL6v$wBX3c4
zr{0D@l=^JL&Iv){f+t)l#82Oo1OrT_>OAc>8X73-UL3f)4&Q9wlhiwL4nlCaqd<P;
zY|MQkb}72*d#PQBCz-*VYY6nKeASHnqW%m$a4p&h`lS;PcIk*k)}L<vWyLd*B(JpN
z34jo>axiyrRxq)3xIGoCNGrraNI{4};10!o^PE|#jsTbzL}VeyIkVBx5h`rd;K)c}
zXcS6gdjJ;jhbP}^tit;C^cEKvt-i>^<r9!E9ld;Bx&7qGu#@=qS7#-D;s2kbxc68f
zq=w`#&n($&N~&cAf+^)yHBP2osX`x%Ka|0;a;&11wy6-xXctGmVR|*XH_pA5lY^Lb
z``Z~r7&>$E<qQnYBh>5Z<7ax<<=g2Wd(ZCK`GriiM-zKB3K$#b1`omGZLt*ZDNv-v
z0opAV-ugF3=Wclk34g3EAe3qsE4R0`trlcicgv}$h}qcKcxjC2Fa^iQOT5Bs+BsR+
zv-{ZLzA=`<yTv&qeRv{~K2n_D<C+1b-lehxqQSZ=`(2Vd<aJ_X>9cZoP$=SE@%svw
z3f_4QpiG=oB;-zIq|=Rz6bYk$&V^(qIy!oBX(>27++)f_cz^}I32~qoTsEAYpWu0=
zB){0nVa1}$mNH`=gbd)qEF7HKY<NLzf9#DHTfWWpMT*`wLA9S=$;E|FNy%IJ_Nawk
z_yUL9#h67RB`2rL&XhYw{UcvFVOb4HXw8YNo^R_y%W}aMFS@xY<+EGl#6J`H9^%8%
za)ytZS9H-R9Egeo=4wyk1rq~6ivxc$^-neH34t@X{G8x){T7$^{a*&l)<Aa5eBfJv
z73S-s<HH<z4nh-)T-jR})zG|P&So?U2#I8LdWO`+-;;+Q<-*N4;zHEg+S-!x@&X`_
zn>1kuXA9pa<ET7MjE6Ym&z~>|h?4C=a4Ucf?Ot4gX=_U{M0O6Xg>QV)I~K_rDdzw0
zm8IR`K;ytJ!^1<z4!-YGgP<E6<$}ANdhLyos>hlmZ<v83oU4WzLP%TRV>+MGPnDI0
zU#jE&+U*0?_jVo8a%vwfw1`lYZijm!11zL6e>Nnbr3QX^)4-+@2Yu8UJv2dLporcl
zdG_RDZK5<Ei)VRxIawbc__zR-5&JfW3`-aeKy$3nd_f50O}IZlk%hx~gXJJZ4?#;`
zIRoXrgC<6_`pEUac~T*z`6vA=iOsY_(;+&Xhk#+lGbxIXI0P-;!6>kt3eL)cD5hPT
zY{!!jVPNW$42fQ5R+d5kaw=V#K0X9l>y2SX5V0YR+W4FH5XU!7uwF?Ijx;3X@$GNV
z{TQ2|$r9TQF<MrxdCSID+4W`c{XILv%Bkr!_wnif(6Qtv`Y|mq)7B8jX}RlAO=vr5
zt_maf)G}4Fa-V@+T?H~5@I=i3_p*eU2K<v-$s(TpHd1UH9E(o6QoONnYMJy|Gma=e
zXNQiYtgI}bTZC>z-7#0`8=bRq)R@uspIMRKh(E0k0Cf6M+c<cNYCt?i7gc{77RM;s
zLq$GwnDe9#1}er`IYiS<i6_#;a2H1NHPI!+Ag%!3MfJ}=>$^v64)wl^)VB4Xg2u;{
zA@Sz^>=P@VzoGas6>z5StM5o$PG3C4rH2nmJa@K|n30_P6$I7N;+T!cS?2dRU>;x*
zz%^YEE1=b=Aruk6QL3<_2APaos7|~7D{{*Mgf@UZX_fy_rSbA=|0Rq`pF6o=j20Jq
zzt!^y?VkUB%huDRwS2LCH`N(K5FODf8uJsikjxmMGR?cmi<}<LURrZxk&~Clk9D-c
zc6JzriHkG%C}W`cXi9$o5ub{V&e+N|`8dgS8ZiaoJ!ly1HM{k$FdrMOQxz(9-5))C
zLy+wJ9}H#{qDb2!BgNYc{5bRS+gTU-s@*s0uFfkFGT;nAn}rc*&|SLB^pU=R?)^|I
z(na(}j8DjX7292jPSMM?K^sA<^mqMtkAo)}4NdmIKYG@K>MRa@Q9UcqAocw9OGL%7
z`!iGYk7ljG!9hGnTI<CmKD9gU2tdYNFR0m&p(Vl)?y%yD?G}=h#DMC<jAQa=?3U>9
zuZ4DOZtxGup|Y~FP?oI?=f;}nz!Y^?0`J6WB|KSTUWZg##hs~2cHnd~A8&B(*4>Tf
zFn{TMemIv`xh-1m*{(AW1W!GX;}VZ%N^8HqgnH8mK-SWQ+jmfMe8)yNaJ(``(J}0F
z8itO%bYCNoPJrW`$P|EwVKiO!Oiu3Et52`9Cm^GSzea-&OlqNRB|y?hcMu^N`R3>t
z-K`77hX0Y>d06E!ZA?<+uaS{%?|>KoVvqgrfnkjI>Jx%5E`u%@x?Qz;0ROceg=E|p
zClyt7bxDwJks;2&oQ513PB@rUI2}o@5>#wjb(ji)`2^4_KCdb8QY(u?6BB3V?q2E;
zh4X$r)C+>j6JC<OzM%pnJuOzk)fNc}TN`6Vl#s1dR8-iM+b~HCK<1wep8JTLK(WaD
zE<va0gYR#fXi@VL)b_%RSBAy%7yE)!l%3-ty1^BLwwJi}j?CIlc+HIL+rxisp|u-j
zg6l6dVa#9ffogDD=Kl(S7hLZ-2z6n4#1Eg<XKf$@GAS6qVgkt<n%g;e#^>X#Dg7}y
zb$~J3&AOwM9rSP$@UDD+BAEmX(WOxs{qP^zV9jc^{u9taEZ}fARfMboS}5a_l5gR2
zfc~ZF{?!kmlomeGwOf0prq%@9IEQ+wuEL+-S5BsoU}Hq2M~W%paSgIg{2!8L=YU3|
ziE?=%bkEaPI;mZ=NT>6Yotbq$Nb<DVVr7`})}Z=<Y9_XOI7b{x{Z=R|_w4R_yn&i<
z(!^GIWR={N^ZCFASH!>VfOnMropOoR!%8i)t{`btbv$ke^&CVEAbh7)Rq-sDJFu2U
zAko@X#sdMj$YM#<0YL3pPSA9??O{;2!UTYOyAEVYPkftWE^T<HxR@M~yb<0rbo9wX
z1*~7%1*1^fFu0-f`Ex(a^xQ)^gxiO(3AY+E8DyjI^TQ7iV)QKj_u_bC*-R-2Wuyt&
z;+#<nPPRl_3bw5{jj4Exn14P?MD`u1@8f1zPUpVq@DonBMJoh**_8wZk#T&<)81&P
zK$>~!QX5QUKo6NayfH8k5wbVN#&jm7+{PZ*V0qxKy&C)Vo28HC|4oG+#lK6WAMB$D
zJAa;CP!I&O&>6A{|7+0DVaNt>ty95Vxg>q%=5HKcdT#DYwOVRZabexS=Y_fyL>qx`
zeg;WD$mO-?A$}Pw(s53FRiwid4LLBBUUG6n#1lS3DP0Ndk!L&A65Dfc&5t$$9GiZk
ztpx{$c)ImRCw97Ju9zR?vhQy$NR{aGc7tX3&5{uvW?DnSLU*mYfh>NA?zpkO{#PEA
z2M_IzmFTjKDdpU+TfFu?6hF@xwU7>0dMR;yp}3FYwxx%uEXI3^kSL;&@WlaB9KtD7
z7d%&o1L83(nav|I*dD$NHBovA_UVFw11KqMvHo9JM2(G&D@{4ZpN)}<^d43$Bay8=
zM)lY|b849Zov|F{Vjnmc896zR)&+f-Ubc(7sovvSQRj*#7kx^aF7dhUjB?=M$6fQ2
z?TCn3kx1qwOtl_fPlD}JSL2>M;PrnF?9IIeT(Yv#jR+_4=_6^<u*-T^3}xK?%s9kp
zaT-lZZIzGOs-ypBk`<tm_u$?)_{84rp_E|O<?C75AA4>aZNIX!wZTp&9X4`RVNM+l
zpXHa_2n{{G$<SAet5}Q3PC>|VpmaF(tmL(7su6Dg`}BEc*6kPMsu(m14-aqI0mlpJ
zHRPC3Gx`e=IEnKR>&;#4SZ<T!{w{E!L*cjK^~fL!A9nloVhWQ$$V!of*l9DEtr%WD
zIC2A-^xIk$Fc2tBsedO$-eHH7;LQstgtG6}zkZHzdCYctRUKr32W8u_K}`D8y~&BO
ztJ7jGP-X#or0vIm+g!XzJht0&VoTESQHp+y^ru@>etReN)2km)2yEGuQipz!mM)x#
zDO_&r>7jUZp9Q~fvmJCf{c;u{1`^QqCF$e)M$*oCuOz(CZsI#6uqDLNYsE!SJUUgC
zBzhJ1U?pxM-A0_^2~6YA^)RgqxIk&n&CQ)lP+$<Fn<Ms@UD2d{u-lBoZ)Bs{q!h3z
zm8n3_>n*l<Kpb4Y9C^5S>7f7HQBrQwEeei}HAK4unBCtBj_Q;VaAQm&4(%He1!G2s
z>qEW(qAl!Al$x6E6?b-Y_;q`Ro^*+pT*jK(9j>`*6OV?zI8pcPuD(#`C;Ggr(Xv;|
zotSKzKTT*pI+trgNgODntSov9F5OY7W=iQ?5;ZhK^3NADki21$2!-nJzrN-7FN~q>
zIy$_E0!fW?Uo68b1JoT;2S#Lz$zSi=@zANq%6<6oY11%`3Ka-b;0#ND?GuvycJSw2
z1_ju8jhkr6G8KyEQVq8hF&p1;6HK4?b8g(}yM5hKAA)AWfp=e0K_@2%Ggh~lol-p%
zWR&>+kdTbC-iKrXie>HER7Kj7){Qlf4=_NV!@K_OKL?ULbeQ7-Xec)7vsm4dV-<U6
z?XG|>Vnm_vx?y#h0T!6p=}&H59mo$kbE=V-Dw+uIU0E2eoq8Y6D!FH^-M*TT7-2Ke
zyjj4wc|$+9TzgH@%fsV;c{zN-fbai5+%0rq418Dpr$->!m%e7$orbHlYsUj|88R$z
z6zDEJA+>}-o6;-)b(_C`0cZRM{2!$LVCrgW@r)js<Xkk4*nzG6EV&Xhqa(w^k1@Pd
z4)L;wVm``q5eF81?u=lQ;~zl2gGis~KsU@R%y?MRURyCMva%Y|&Brc2yRKqNhK*XB
znF~)A=0j5wx#i>lcck(OIkvWx=)ZA>f}&y0ETx&?YkgATi=s`(!Scl+TzCcVVyS)J
zKx!u4OhbqvBO^1AEMQ87&BWHSxw?t<$HhU-@V9UfULU2NFJusZLwHd=-l&>4S}$t6
z^2|fo99TXcSFyR7#*TC48JV#Wi|uqp`6eJLCT9DP32+eNKop6AHt4E$x+u{7lzKlT
zwlo5L>qPFlUp!x!qIG6cG!Xx-X6^^G!OMrhqpjxOyOL5;ViS1xrUzp3#NAZLvgMK4
zC_>xsBQ?&MiI#5<`@DO3BAr<He_VjerIw9$&J`+D4E(>#$6TVN!52UMsg|c*`9_yZ
zz_Wqyj!;mleoSyN7>jY%$59-Iy?eX^F_$|pAKka=82s4(>VF@?2JQqt#HI~5psETA
zUabr7&xU!FI0)H(3ls5H|Dr?u$U&G6^P6%b-JPAxOwVA7%kW^3OHQtQIf9^_!<cLe
zW2#od*xIv3kFZ<0xzjhl%fM0vA{{i3@oG0>>#y;Lk9Bl(f}*4KjV=lABi;9EgsVxB
ztz1?8pT(+0iBmv`{}-vpNg)GQJVE#2uNZ-C%VF~HJrVd|%F5$k`bMy`hYWPapjPHf
z%x^X(gH%w}b%IVYU@CE2n0Pe3glizFvsbr%zgOvSEW_9&yQC!I_irXBisZ9d>0;~C
zeH-VV)6hJGDy6sP+Wt}e|N7N`XtIYy!8!q=fG|ykE&3pgiy4xJ**DT*IBF^YvF%QM
z)Gdkzh>*0vz({6f*{5}3(rRU^N$4nJCF<??puC=pIO<!|${(%j=M32+Ir^544o77H
zv>kn_Y_W9Y+Jk>zk|bY=K$`|yGgkL2rvj!-D6vto(Cz{Hd0X$!-o`qK0m#&0W{AYM
zd9=#DXs_E<3`Efg0QE#@-#9ACA`&^gIj)_N_;f>j$a_{GUs^<Yy%nD0az@1FdXDZ+
z#ENhT!~sTDq3t9WD+1OqVlVW!@87x^mQ3KCyGHKbnAFB*5IcmyP$FCL%AVb?++6b8
z{))3$h8>TI)Adt9MxU@t{BPye{{P0k@2ej$yzT;cLWK=RgBRWjV60@u6G-GVo%_Fq
zE#DI<E-p5jsi|%zXjreX+47m#VjI>LOzD?b$QxksuOul8>}U}8k>szCGu~`3VNm$J
zTH%GT%9_L9s=*qOcjc(bM0w{0D`A?`mM$CZWnBJFe0|Wv>$;=^Nl!M6U7~WO!<t<k
zlk1Hwo|{2tV;5qTQj=yQw$S*gv)ILTfbXQ`dLdNo*#V>oaHjTo{3_x93`%BI0I*+3
zHSy}djv2MgnSCmqm~ie;vm7CTV0R0ArWe2QNyZy{bmag@WwEjfAsLj(cLVvl(p8^a
zwGu%B%W94|&9T6BhueT-H7AAhbW&SpJ3vrQzUv|kUE}h524Bixv8$AP?Jf>K&lUb#
ziuew<qL$gVZi8P(yyNHX7XPd;EW2~x3@>LSPrz*!(&rU<G-_%ahw*jOgAwMBh(S95
zEEJ^X$1pQP|Kp<kki_yA`Zrq^l9$K%x3>W>Ga&SkMl#anSN|66eh`%0Pes6H^u>Bu
z1khf;95<J0J03|8O&zUqcJv*0IBZ|bVZ3<8v3AIIl=(uLwuMGm6|HVj&$M}YTnUGJ
z5uIsey!8iCgU(_F=YGSJRmaz^;t}zPGv^Gh%LlmAX;$@mI+&ZC_3RqonfXhVMglxY
zb7I&coeAwPH`?|7p4S-0w7|F=5KuHS<cn@WCC}n^57I-X*?;>2I!N!0$LbXS1pLp4
zuq);BiWF(frc6Ml0Ny^BJH!s?KtT9J5?h;Gw3I>&;rnF*Q8pFr_-F%iW`J!$dBoHv
zdWSMGzP|L@7_DZ0lkFt)g@c*Z*yhc*R3-J9s)`>^Z!W?n6V#$En^=WDAi3Sk#&K?4
zm#Wok^)yC!_Z9$ayMwOGLk=B&PjNoFLUUtMnN}Wh6{V|QB~|CSG(XPc(R;o6*>HCH
z`qYq@#LItFNmPyu4(^{_p=V}(659Um`J-1~Qv;?)A;--xfgJY*S6o&Wv^fA52A8tL
z7ux=PV6AD*>ao9&|8wep8$JG?+I5Szjz=TLW!$j6-zls4X)5?5i-0eW(ki25R(DDu
zpBVnVQ@o&3+(SA4aCc1(^YIaKA%{^EkqzxL3|e}WVKfx-9cRQkEZoG7a|V~DIh(n*
zj!$bv4@YZsnKQ(RglFzv>==Fcc0!Wa+p^^CV<{}f7tBA+LmWHM)9YK4^4kg}V~We2
zeaAj}hu9&(jH+RGgR+04ZSKACmoPfTuQ#DP_Uj}!$?%-$T-50xQHLt;Hp_TBQYny}
z>PYsuUaOErDeT?MtKpp!83hGl=p4lIZzP}ni)y`i=k@U3x4HzHku^kFh@x1u@T5o;
zKVmoE4>Mv@@f|U0(kjG=Z3v4)2EY}<@#CZ|>emp_<_o=ub$78ysWh&cr-U{qlAlw{
zs<k#8?0jlk4aIX9EIi#Xcwo12->wSjO1&h~=sAx6?;XU(%+m_U%HNR@rfXf;41U3m
z<PPmku*c=?eHz$&ANldl!WpK9J2$^|wY2;QVNG1ok3{}!HuIz^PkIkzr?U}PZFA1i
zlX>P<RtF@*j9zWoxleglCo)9G>c6rKklqu8$3v$V9vHTu{xBC6^gp3^+&BVk-K=to
zh`4Vn_y4~1-xEHj*zRuP-8J{1Zr!&7Ebt72MZSxWY3W2g$NI7=5Z<h%Z&63|Z%Wg6
zZwXo&FzrAv0^RbK!RqvTYk+$I<lCRG0d+^#ubiC4vQQ~9!*SU-wOI9ulx9{J`z2nd
z*p9K?9<AeNF89xPFFq7oGqZVx=2Yb^o~(cCo?BcPV_-V0`xb4rCMT41Hnq=Se>O&0
zetv!<T$>ypo$|D^c5w+;&usUF*Ke0A2H6M`_B*L0{Hpcyk$ebAJAGOe*f_5YCK*o+
zovZ9;XzAlz&^wo%JEWv;0(XdGo1^+u%J)M$xtww($R6L8{n)s;wr$1L>iN>edF3NT
z6I-{pNjA21%*+4bkN5ut35Vlf#SfQ4zmbn}sxdtF28gh*_VFii4Y8WFZ#T=>U%qTt
zBrSiC3x}bxet;JiD$ZR*0EPoX;ct*&O4%$=rqa>VBN)0q)UM4@5xcH`dyGfeJcQS|
zea67asYb1IirgCSL@)U5n353n@iEfg&ZvMj$7?F=rj=aex-3Qey>|UCF`nuAIm$=8
z$jDP?W^-SguH*@n#yDxOFG*XZ;<|pl4b*A$wB|Y4v^pnx6_i15|Hy42pTi%~Yg<g!
ze{Y-gMyx7gk3-it`k4$d(|Od2)$YMaC&zq2BgYnIhQ;-wFCK^4r*ZWIUhu|YVPPL+
zWe4+~E&!C)|HZ1bTQl}8gBF;K-}CIEVW};;|1%k|ENVgAzXF&}agvn4w8+ZBV#&-h
z(223~Nd~HdzmLVaHXvD*k#c#txy?FE(oixWX`CltflsorVT14rdDE;#xgYIl*I1Dm
zXPi^%x-gE@E-}jWLWd_ql75);>C^fSE>T>1@=`}km&-*pNkYdLZ$15+w#~GPxC^-n
zM>jFB$@iS}_^+N%p*mQMEfV;r>(-v5WuGS0&g}o#<3j%Cl6SN5`EXmN4NeIusmu=H
z><n4R><}iXg&4`zRYIG)B8EO+>h0e}I3yfI$>{94;dUZpCvtkO*X`&_BaA`Wsk_ou
zo?mobx$~U#8~63G?LBn-uxs`tM1XQ3{{l7MiO)Mfp{#%Ys7{9|O*Yr?0VG`T1#kO|
zD{0JZ^3y8+3x60-P(Ausa;G|IX?(QA3kOhw05xqJB)bvc4o&!p=r`N0!&L0>WF-Bz
zhnSVMhrc=kXeh)MXdU=be}w=<#BF1ClY%*5JO4oEuvc#}nmw5XyG<W|hL;b;u<6|y
zX7*~#jPH>K?ZT>dPu*kbQ}gzW?+x|u?or2;=_|ifIXVc`7$|ZiNxxa)d)=s8R5b6!
zFlO3^m~%R3h=I@cSWvobfn?-vA2|d6+Hd?c0yeihGS`;XScHg9r;D*oLwMCE4D+-6
zgdOJ!{0nZxyqu<=TpJAhoFYS+HSFURoeNf{{zMC$*~hs~+hgM%FD3I|d521Jl}}~V
z{nsXb2e{<GxCx{$8%K9$Z7L_Q-2ky771_VIs!JHz^&SrXk55VTGT2=NTYj~T`i*Xp
zeJNr|x;|t<;FJK)f{ud4K*mfppl4vMX4~Z}>DG+3>S+vkXu#SAQigatxW5jF;(qnk
z($ZMSDV|!rW6tI`0j495vNBT^yL;QW|ArH7h_1=-{2<}K2qQ3ciAzp)_7|7drjNDY
zu)SW=;Mnx0Rkc<lr*SzAq1IDxe`v}hSj*SUK`43YLK1j+rH|wEDv>qYMXN99s%&nY
zAJObIIOS~Do1OEh#S+18yQB5~#oN(1n@J4r@yeNfn0sEVXNX|l9~&F=Y+U3dBGed+
zya?b8x!OO7s@3H^a8>Hn!;Ndr%&%(69!Q-1d9l6HtDxR~`ryepd1K0f9gor2Zx3MX
zrsYaNas}tNWRaGJ8sDt-0;mS}L&gs}J^zR3d|!R!aNM?LRkHC~0Vo>;S1sz&*w8}o
zNoL`y^J`sDb=)fc-tgjjFVGdin^gijYlNV5GHKR`4j!HTbTl36l?C(qPkpPxzq7Lb
zA?3Kkahde?bpAqP)w!#ldL=jh05O3??AOGf_{~YD5_%PHZkos){_~)_UPR5pJ3a4p
zd_EG(-9+AzkG<=~8|-1X(a1Tvg>erXBi(A_p2cK2_M-K|WGMaYk?r!wjm(z#XhBb!
z&;d!F*tOo9C$%#ub^IjGsBYXlQKE8V80xhrpOb&ye{y5bS!8mF*&3_HW4Di0Rcv$j
z=5jr@$+GI&f+<ISdPcFsaYJmBoj*{5n8Eb=cR)EmWrt@hw#QH8GalA0Mq<Z9tR4dX
z1WU>UDOew%d|AOREFpmobq{FRmzL8xPX!MdNEki>_{(2BME$G!*VM!RNYRgolFQSo
z2j)N_V`0DZM%-nD{&0ui&19<L;IA)3nogcpp<<}g3&^KzBr((~mMk_UBKNO(rL=bX
zL$gR+=i@ydDiu7JE~^7lk{K8E=0s$6UWHFe09RIf4Uli}5QE{>|CVcq5)r!dEi<H;
zTv3=SwFs?*ei=G6bIJ^9JI>zJk4RM)K96Fx799KL<5|>+xSSh27}!>O9pQ-ibejcG
zMR6*^*$k4}`6$iSCqCb!4(1*L@5KCcHwBSuQ|Wu>`eL><KjV?l1fm(InjgLk8M8)Q
zy|gfi9#XPi0%yVqN!(i&Y|Sq6jQxY!nR~MVZ!zcQ@@wV-q#`ZG>$K$M<U^$sr8!fD
zZF}suj7?PjAh~s7;tQ2@_e=?s2pMy9VKMZ~W$;TZyJnAAjAS1u{xN&e{LRa3Ld~{6
zO)QckQQHutP~fDLpI)vXUKK%HC=zuj4MTdrCH5b!`cEYw9bndTW`7)Cd^fGVipb6{
z_nAnD6R9JZZyb@|y}p>qKziGq`3a>jTTg1(c&e%O#v$P<@AL_0v!FM3va~vhEuwun
zr>ue;5HS?1b<q{n%Zu#3Hf#NaT=16YFP^~j47x$^(;@Wh{8~6BJxwQAl^z$`I#8<M
z<d1a1LWRBDnKxrz8ZsrHsHO8sigX)Pqy;6_`NTAwpZbn<L=@M)D!Xy#Il5F6Qo)}s
z;56rm8{MKOw4t=|dg#^dJ++;xF(a_JWR>O^sXl6qK@n%h7+7P0S&i{b`;U2#_sT)L
z!cLLY-lK)lAomMIvw7-_XRR3cj6O=tiD^j+e7rR?U1F{r(FuY4jhl(`maQ3L_O_SN
zRkekFsIFQccD&!MGH&T3@YT+$VyyL}vl6cyD$F3wUhJb*&E20OwrARli4kv_Hy3ri
zbYE!o(_C~%v8X4HG;0bLPG_YlcfPQJHb{B71WGizce9qzF#~kgrucRqlNFQvhnVzl
zLr1}?!CSj*)|1g{D_Ei&U$5JTj#lGXBKuA%2ooGkoy-)@q~Y$M1-O+`sQOJAr2S3~
zxix8M*H~_<>^wVBpj8uZr(qr|`TP)uB(y=S2?0hVcqd&hwp#2R+y+0e(QtDUz)ie=
z{{Z^euLWV)Q$tAC24m_TdWueSX`I<InTz7@ABfEtGKJR6H}<ncMiGP6X)JV-WcU<#
z7>`g<?t6SY;(t=AbU>`##z(4HssHrigrujg=B?T2$C(T}XGSlE2<H~86rsDj%MJEc
z8{^+jrtL$ES|IiF-fbfF&iql2${>~Kik8t6jag?+;?jY9fjKwYz&Ba6Xo$+B8fIU$
zd62l$UZA`uoI?1Q6pKSyoC3#m`s+e{lcCL$K%=5`%{Rf7z`L^{wtTZ;h=PQFQ&5C!
zS-fy>&K{jiJG0wfZ<!L>oX$zqB+?av_9q-WWXO(u<q0(<5;Hr$s@Z>rKhJdjUt(aB
z7CBKeX0)a5Z`9)VOlNuY8CTX1J3`_bb7E(QlO@<mpaXp%RZu_~H7e0X5dp;#P)6Z&
zvc74IJG_*Ub-K7aBjM^(mNlP=d7fZ5{{U8GNTycctEdwrsA2MlTuOTtp7Y<dsP=^t
z>uI$8W)7SO0i#j6*Ix(MT-wJ{!)X*O76xNKr2JFUq5Lv@NT;lbTIsjnW*mJL`$erc
zudkN*JzS!?1%~N>#)!7ZoSXE7N6|06ran(PSZ@f|g7T?zo7v|=vo%#5$E>m18#9Ce
z-cubu#bX>jLUs&6>0bxKT5Qp024`9rT%X7!y?pcKY3F{wS18h28mVHwk;`H;(K(2a
zpvALAABb?Rv)*CQl3vqAu-LRo;5@QJh%lP&AXtg(;k&U<c<XpH8+y=Ne>S@y<ce{+
zBmT0G1DB+LL@L3G5J92#=0UCO$S3C5F-4Wu)R7tFI$gE{3H@g&4kxn}*?A3HHq2`c
z4>=7d%W#m{(19NXWB<~MildwRlhBj~su*yv3f_=CCQD2FuK@U>Koreev9Znl%(gf|
z@CUO&*Yi8q_kAPj=0I@p1s>Y|NbPm;7~)3&*s`+zR!WgfSI>UL7X0ZLNSt$iwi1@5
z10iN-r{de&^eb-jr7vM5BrX6Qw+6AF>nv2Mp^Ud9J|;tb3yMz>khj3*kaiGq6wBlG
znnreJY<}8+!Ts~Kn|2_#mxha?kiQn?L~tXoYl3_7YmUb+@iKcc1iuOmi}1^DjrDAh
z@U$8$@B|DbAlCIc3++#bD4dyJ*`9t=XNohy7}Y}(tGS8FcTw#6i9+bXP3r4XePMaN
zYH7FfZG`jUk2kDGRd}40lj<}fzZ_8Q_d;r_ozeMZE@|6&jr&S>!9GIW`T@e{gr>nF
z=4JBqK02-&LYJ%@rXI^d6j3D`0!mz5{8lx2nnv9&q}d<Syya%AkLXGG{JvG-?kbJp
zTo)T5hqJ4;Bb!!rUgUQzd*~%D9WwXM#$>Ypp{5#P@rO*>$H(XR^z7)&Nl4lgXlwEs
zMXWaHAzaD#J9d;VB=6t=K+&WQ{1hx;E-KRMQ(I1kL8EK|^AwNc4|@=m+pFWhDZg@d
zE$8@e{GdM{p<=>#j_imLWEIP9%*19nRq^>DV6X^X(ac^D^DQUMTjyxT?!V@Fth;pg
zB-C+}nz|k<Mwdsj)X?n^7TGarULr~I?deq5Zu9_$LFxSHqq^%3D-SV)VU3DsZC(H-
zZdF&z5Y3N!-_X__5f1ng>c2J@onBUYT$Z_T#@ZS+ksN+18prF?(+I$`xucE2`Hi2L
zGuSeq^83j2jj1R<J;#-DT#vZA-3mX#Aw3ltPT*euUypamnoMfZN&096R*$%C!VZQf
zED<^6Ef3DhCOC>upKo7Z(w%>%L?s9tl>6DTz(I#|Z+YoczH35Cf@?EX&!eg{0|fXt
zVm_9?{b#A{4u<R+u=Y3ZhPXOce{GF$7EA9IolVa^{W3o$K5lw{Cuc_FdDebIN1{?$
z4blidvG(<PvC*$A-A@s*m@{_Yf=PwUHU>A&%$GWJt9#L*;)8_y5r!85FD1Z1s2<8Z
z<P7nL&mn<%MqVel@t-&QBk&3@6{y@rm3!Gk!DA^vZFnTJK;0Tad+B}ch4H!nAxU4Z
zBfC*a(|H`cIAEQhr#r(;UK}WR!pL6{=-z{;@7mE@`gW7l`Cc!Cd4F|^xM(BU)fv&1
z7H47h8|HG^LcqeTrdMC>$!iDP!@0#qDb1Vf#xnB;I&2;56h{|^W9mr~@;tNC8tt;3
zkY|MKhq4%(>^?R2SM9$)Kfx5v%$aN%_oq2K_Y#~um@)kE&er6+V_T5+Xq7SP1P%MM
z)(Y#j?)tX@^D7}FXZx+aAFw^<z6pIC&HYh|+uf>Ltj!i}e{;b8U}X5i@xaNU&5HT-
z?AyYqE>+jIOs9_(p8SU-9r7A<UbS}okwkVZ0&<_%+urH#=a-mdP+&KZ`0tmdV7_%-
zno+HEzpA9j>K~G<{>$O=T0kMs;pPdXSKoa-x5Q>>+0)NYHtM$3%5B$}_KEF7Y}vzh
z^BG&suG+rcpR{6C$-90sRFB}xK-lY>k=XFJ*|+>=ScSuNk`tCam8@3>G;0}6mIaHD
z{hn|JNV0Y&@s~qX{8Z*x?BetHcT$J{n_c_j0h+GSusaoM@l}u}Hg^K8#T<~?9@}z)
zp+`R=Znh;h&?ybGl$BcoK9C+LwW*pnKc!W#;#j%{-je%mkGpca;#N!@_7XIJ+HU#8
zYao010Mn%80zq_u1u#*H&<8->${V=@eN4q%`8V!O&)Km3c8@YP+BZYG0ov1>t9<D1
z(!8)IFICO8n>0t|)6h!dDxS}uV4^yqRakJE?7c|jia$n*Pmj}A>v|Jb3z)vBvn#9!
zMd-3y)91Tcm*FAC^sTuD*s7uS<^;+8gPRCP6X>SDyV?H6Y1LL}G+v11+U7eopCqhq
z>Uqc;q>aH5&Ww@RblUGapZp~M^ClMV{y1fChU&pQzxdJ;dq&#f(&iI}7ZXz+&rJDx
z*t2UF4^w(Sk7Y3^DO7lt^2La$Ng(&+E;;dS69;TplbFPYzGexQGgsO=?AmM?nGnT0
z93&!C;rQXzO6K#=WU19Ne6y}D1<qt!DFhQfD<lNdd@-G=X)ptih1U6ksLb5n!O6-)
zLtUp6QkwKQDHx^DNsyP5li6LB`4ZF8LX%N&wBA{Ene@|1rRFZW-Ew7kPG;86orT_4
z=GS*raUQDB(3qn*`SC8GwOr51s>d^XePJ0-K2ejC118!Bxgic5e<7e8QThCZaVX_v
z<|4&f-1r-M7_@m*%Loh-O`w^e5po2b9js)a`M@TkPSAkv0^$0Yq2b%<Tj0}0^SjoV
zbHmiDOiwhMF+LC!qM7yffmg>23w<IOG%(0&#?=2mIgR;vxqpX-mR1LrmQru&#|+O-
zbFZbFqD@Hq4!MWbZwMdM>h-P8Rb2-v+|JlO<FcX(ief_UPE?OH0u>>rn=M;xi843s
z>l2(M5@!*48?9I8b$2lzJ@jaMR_wU^<NVVh<VTZRKtl0L*QXZHu~uV^J{G?|^r{Wo
zYSVmo6HGX&$8m8yorva0kwNT)VZiZ5a7cdG=rSRVhJSG@ttk5Jdf~EGFFmQn2ID%A
z*oG;a-9`BGmd%1<qDk|jp7ny)kjArCLL1x+LUVfU0Ipa)byvR;zUby_!A3=m5>IrR
z$b)4i+k)RtcX2|aV}<=2N&F$`Wi<`FZNx<{G@1}fN8|p=-di5C#fOQ&c`TlC|7|H6
zZtLNevFgF*2u>d*O`%lZYD3fB6WwahDWm1!d<C<Uey>7P7h*9s(==BTHVxCMp52?8
zu3rh>{;G~T6vO9H*dD!|C9_Uu`0UP?&jzHp0va7pnp_8la%#0upPUeMH94act?&E(
z7#!N$>try$WNdxqI$fn(y;A`s9mu>3715xjZ1=dA@wNt_87}_btaT=bWLPV=-*G#G
zOJ~)~d$PM()XM~dd_6n|U19FWS0^*QoeZ^cA2V~PVUtY;Y=ZKi-r2KL{PP71^ky2M
z#I-0dgVnE5tcEy2_irf?@a&7mjL@#F_2R8@V2R-iY*f}kr%D*CxV1sVk8*P1dvT@4
z6P_arE6WE7KT8@q8Ti9x6x4mw^(W01&><q+rPFW?+YFPD6kv>|is5)r)Gc%z^ulhP
z2e@}0u;S!S!E%7%N0QX$v$J%5M3$)c0~dueF5b=_s3%0|J`J`THJ|JMvfuxHsWVo@
zd6vY_WV6WOB!GEw!uuoTgx{N4NJ3|RwX^)NDO2^Rwbo)l>BO=_U}bkXBwFagC;8RV
zVrJFuaSHWFyxm`o)0qqAAI-M+I)zCr{@EDLUrg2RUFam?*9Mn>7c>%U;Yv3oz#y3&
z??F)WqH|lTp{CVr1~=dGA|pTOD&JR0rP}|&1+hTc_0bg8?))D`C8ffrZZVk(yW=s|
zs=V{B5qd6XR1l2z?atPd4<0i-Er8<6jJBoVtCLA)FALp_e2@JF#`0=(K9fruA8iba
z$x5eKa@kw)fwHaL;O{0iF=7~_fpb$8zb%BxNsMThmJ;O;B0WeLKb^ZmF{4;$Pt$t1
z-b0GXS2ss0ygyrg@<CpHIHwoz=371=(Dh)t0~j$E=Un}TF1$bUpe?jVzz7dpTO-yp
z4yMj;J%6{s2I$~w0{aoTwzwRG7~mONy(tsE<Y)s$G!3kxW&D{N2=x}3OmU^K9J-?r
zw}A{b6o{B0Y^xx8K@sYGYevBNWc=I_MsbmxwQhnaZ&-ZNQUDVh1kLZrPX3PzU{dB2
zsax~4T}Nw@b0Lc7OzZsVQ;K1vZuOR5mVqf|ii|0g{&Ve}#B%}M+p{3yt}o*z3yg8<
zwQ94&O=OxHO&~-8!SWf*JFR$``dvgO+G8tU85$Vs#QdBNgk&2Os_$ska0wAw(9byW
zMv{oR&i1(rf0(pJn8{J|<X;r_7qnlV^1A=d38}&5W#N69*6Vmhaq?uL)##DPGt=64
zHoC-Y+24NQzj%yZlw$Bus<$?fUTNw_7M3ilpB&>sY3t3q{5)sxEa~L_SDLkOaF-#?
zC$0j|AAedF?md{nOqj#8+?h`zz8iIU+=#ZN+_Z0O*xy$W(?wgk>O$)ieNh`1IHdFW
ztmJXcc!hfsezShz-=XmKZyIF36IxF;IumEEet2w%DO$|g1=yKmpkR}wXJR-nlP_4G
zSy;UMa_%CGjG#BHP~$lFCseV94r|lklTJp3UKvM{%5_h#2unxX$`cK;Oa0Ris`3x@
z&^CKDs~vWY?UC7C_c;l!EvQSK4xjDRo)9MjL-}H}l;3v7c7{0P7xFg&02N_B-+Itm
zO7088Am#rE2gx2VP89kLGqgR7X(*c!nn<8i0IcSANDCl=oSSys`3!#j7L0M*-?so+
z;MjeR>hG{*wq?nJ5i}HZdlRNstw9b0IA6E3|IpDF>U9M9WE(?De*Q!eGU&On3Ohd4
zgXsx4&0t}Y%0yz`g*fZAIYNugPupd?$mx>R!C0ginpY(|nOSJ&n+5Yt-@0_~k?@Au
zPCKJ~QMpIGHyZD|TS{y9#H+EsH(S0xLn?7Fv=Q--T^#UH7oRKdcwN_+OjU(UPkEYM
zp!=4eXbA5ujgyPV?z5cbHO;ziM;2w~(PXH|+oHyug~=Pwu9a?Uf7!WfrbskOYlFeS
z<ri2J)5_@;=hZmhlo0$jv5`TuMek<rvd~t$bi7b$Q&CuQepw-%>FB#5Rog#xy2Wz^
zTxFk^{d1<Pe12Ye?nMsdEiq)Mb;dKBF8Jp!7ipXz=KM3$h8f9QKbeSoaM^*07dG+)
z-G3-tg~Pf|fRKA55#i8k<<j-Ihnp+Q;A*CKuq##4fgv@h?~Si4^1IagUs(iSe~{f=
zQZ7h|q``St4lFQ%AdHt1Tz?%g$zOOJs#8SkPz?0$1iHK9iGJ{znc?kKRx%vb+i^bT
z)KlDWTAtcj0rpz|KXtJu6*jX2S6<Hw*KOGndNnV8X&SY4Q83n-t4;hqXzP>5+F9_=
zezVP@v&v&iU1RX{skM+FZrl#vxU(CQ21y17lce2)p|CY1^<Os!gXg)l9PwXkvtz1o
z6YtD^&}7UWV196NID6(l!J+=q^JMVZJ9~SrCrDa6yj1*=qJbt;UxsMiTU&VBovR(x
zhOFwLwWOqU7MR1S^9t{s`yu<I(oydpw?C85L6&|NHW!qw?=}14GJmRb+vrJ={RjLG
zI*?<9UIx!DEj0t-zbU`=TA4w3wYMDIb*Xs;<cybY1#fi|VVg2Yplj@45>8xZWoHw_
z51JG!!^|udwoj}w-tlTbsnw1P?u4!QO7%%>NOamlUP5;e%@Uf!?}v9$c-!2r?{YZ5
zdRQdq*14@Y;#DcFtc-fe>uzQLeQWkWiZonwZM2ZoCf2_RnP9haL~|pab&H?y+1#m9
z_5H?XYgHDrf{EPIva=%7o8CXhdk6?b)B5$R(T#W~QzEjcPrJXyFdhW&oEfaL(>fbG
z9o?$3+v%)wxj|B|IC)Bee*WET=f1L6v98&LCQs1m^S9%e+GdrSi*k#9DSi`mzg2mJ
zI~Z1auc|Xy?V!I*-%PIJWw?mF_N<E2*+iJbOtE*MWU{>uO$(ZXd-~jiKi7MLIB{c=
z(E`G;oKNm2o)NZxE~gb66E5A=MU(g<mZ{!K$;)-H5p>;yFGA3}VzwP!XoZ1Jk)aI#
z;)7&ZSikGoKNQv*UFD*54^fWpesb1jvk^#2?rGa}sg5%oi-{LYihUcU`sX2v2-<GO
zNk}+YMZ<T=_um~<v$%%_TVok8iAj$q%FAfNbBOqUtc;U~oOdbq@G?hqRB%P0ySVXc
z6I1Zh_!1o~?|z!$BZS=xS-p8sgfuiX81l_b_=Bh$j^EQW!yk9vzq|ceym!m#hXsik
zfqdswxAYs`BJ<5rMjSfjCLn*}@j9?TVRWVuGchuvY3o$D>{DA?%WS*+68E9Nw1`zp
zbF&8w3Ye|qAAKTrIn}HI@oERxkD1tM9|2nZSy6!>FKIiqI-2W444o=Vo&!~cw_bul
zajF8Uu_BA8US%AGl%evu%8a^QcpyjbU!D_?Pf~e^7M5oKbhVY<#8%y+7vZ8eAw&ne
zbB%@1h(`3NXzRpMGy(WSz`j0Yed>7q`epEk8fVh`ck*k_*PpIr=H=PAL;&#w3F(8X
zs-NhRz{mOL_f8zVUMEt~;yigzhw>s=WHhoQHk#Bm+0sa%ByMVzkRh_ezOc3S)S+|G
zVZ|<c=EAAwI@ICi%3fun&RKohAiiOoalBUL2c>NC+*d<mIIB-ZT1R_wxT}``+Ur;(
z?3}#J_#L9er_YL_M@-K3645`K(!RoiuS^dwx>qWrKwiv32_tjKzRF2~;BT?qpHoe~
zZoOX0(JB9i$kr$J^`ptdMnZ-x{rc07L54Smor~+8{hyYHvCm&Ejue;C5S{UAE3@AH
zG@V(5u-VnU()NTj7=hfu>$#oQl&;MZPSDA`%%D<k8@<tvYGklrAA1qe$@S67*giDD
zSiNU;^f~WfgnCIiSbgreZv2C>tJnS(Gf4SE)4tdF!5l$;k=0yUTpquSlF}OXj4i&v
zqhFckJSv5Xn;OE{BA|*5KkPqBOWD`f*1i-g->wpcyMg&GJy3p*<F*Ze1ztcN%;WBb
z0Sh~z8?P5X#l>BKkt%H5;NW2Bzzq9REiEsHL`B&(GQ5VG2@^(-E9W#Kd3kyF_V?e=
zM?%91D&^$#^kpAOJDW-(Al-;lRL&XOkG%aSJgJ`pjb!*b;ccW#E+WSty6}!ScWZN1
zkzd1V6zVpHI-OS^l@t-()K$Js=~A+>MZo6Fo29B>YXd|o6{zgSujjUWOMZvIzS#tq
zbN!A|({}|?FwIlcC_1To3(Ft1m`;{qWy%6=-h5b7xz>gE>+?RIYxVhLuHcXwUv@q*
zo!0t9fr6B5x`i{%8PZ^pFM9&mPtKiQD@i<XD4OQnnJS#3&idzN3QPAG&gRoM7eB~`
z6IC~iWZJNK9Cv)XCO`6fZRaBH?4M;xZFpx=ItJ*?9V6q3N7gW${8<%Je^0#P+q1B+
z*17rMK=FKMsOfvte$R0;%hmDIE;bsW>}dQV?|ckm?eU)}JPD(BrU-fu-dqfCwwVuV
zFew{vcF}yAi<Z?rtclEbuuf6jDNEOoOq)G!MnhQ7VOzvah(zEi(j-7M$@{frp?8%k
zc@Q!CQ}#C(Z)hUQVf~C;cR-v|%-MMX4{N+8%?Ky*aC36;<5@d%>TxbF_xJg~?*FC9
zE_irHt>Fz`e-*U~6BFn8<`X0Jim<0*k?Kc|;LKqccC%dI4E>=4FIy~nxF|VlYOOm}
z(7*2G-Sr)JclTTT1)8Xm@m$R?(E)P*WW2mn`jpPR$|Xkc9UQnQCF5Q{LIN#;Ggwdo
zF6Ym|l2A&2AIuEFx@OPk_kjt(4E2{#AhC0!VT1@7&tB7$kx61ZR%}#ceE$}GJtL;*
zK7(TPm!h~+`z9Wi@#6HYCb(_w{XdWQ>)tLr3eLdL*_$k4V0pv~5l{CZX9$5%wK@@n
zqE<*m>yCza{CR1~^?S0%zatr5Ea$%e7aavFVg_-uL4zZ?(%1iBqn;mG8W9;r^u_n`
zx4(HA4A_e++2aQ{4o(jmFPwjfQ>^T{YwVASq1b63GBw<JJnCqHTzgqb=>4ui?M6Q!
z_OE|tEPp2Z&4e@M{@Ho^Q4fF3#lx-%Zj^=39$JoU8-lv;DaMwRdP-g;nkj!T(J|cD
z)r<9Q@G$31)H-f4-H2bweX;Y(+MAS1uSiL$#FE;E+RpyVhsuiNkN>>CkJ>aP`2Og7
z<SGfSL_sBuXtbYzgm#qkB{pWRqVICHlH!&PUsc8vsbc*Hbz}RYtrhGhjom{}bLGW~
zCB3>$^w#TBB%gSFtafjd;z4yA(@ps$FXcZbUK>Rv#A1rdF}X|p!+XAvmB2nl+p%WD
zG4TK&Y4(db_2UOsf9$yvxA?oHD5Vy5vUFVqun<hGjCvau0)A|q_ctmXE?|!M)!{sS
zDlI2xJ+x4Y4==m_%R7ot&V#Xc*Z9e=AA>qF7S<VooxMG5RS^SeUt5XFZJM7SGov^y
zDPZ_PYqjrZ)%E$t?T9iIQzIKEU?U&Z)VfInc^YojwqU$Y(8=jiL)p~Iie-Cydt}G%
zX_zQEzW+T;%s>aMG?=)BzLQYOUWB3}$oggd!RXwU%Pv}DD~Lacl2lpgn$e&B%{hlP
z=LEuH;H>_ps>z7odxE?s`2CL$Wlb+HSl`XoBvocM%JyDvP<i>mVyINI>_2<EySD?c
z@R=ZmQ8NG8A_*<^AQ3Wv^{QCh$(l|U+govyWPhC*h+;h`oQk@7&#13-p;Wd^<S=!>
zBcDBHpC`P16LeVhRxyXdB|-7~yO%%56}z_&i5BITIW7dx-(eB-h4FtZ;h2ClN4t`z
zeRHd5J@;k{Wm>x<=KRh^xjMt%LdjSZJ+>?IYW!&Az*q*=uQuBfL1eZUeKbMgl?t!+
z2e5K@3w*x*cH**GubXyu9CMsSc-zYVO(3(uc3r$DOkw}BWvj`RJ+O(yMN>lnOP!t`
zyM`r!RLXB&=Y6#0wt21T3tA4m$|=cQ^+2JN`tEN33x}imtN05@a0^{GNfDYH3=wuO
zc20WsYSEn^xe;*&90jOWyDOTali2ReWApVFEhi^1#cw~1APp6f?$BX5I1qRd9-N)6
zdD>ykQ)a{}j4FmVokM-```|<S(&c&EC%%i-M#E(U9Bgck(y}Q@BQ^qHNr{j>=HcOK
zTJ4{CXI1#*U!&1Sx_b}pOBnOi68M+B3TN27V84-Dnl~n#DZ3Al&F!D*zO$9?Sli~~
zjre><;6KjJ&)>3zKfQuU{5T2>_DTxB=l6L|E;$^I2S4Z)d{i}|U>TotiinjKHCHD;
znX)B;9E5U972o*6dbTJtQIiF4jpG+H$_Nozo-KQj$dBzw&n>ozna3hvd4tPbsECv<
zrC~C?0d>jwb5Mv2ybQ<cpL4OmG|@WUor5o-mz%ZQc8o@FgzkEIjm$Kb=D4`@w8{Mm
z4TqJ*(_UV-#EvW;3_?{AuLw-UBnxxO*H79RHfGYwUWQT<U;30AYLHgb`nMDC&0k1c
z##>g8Udbgy@rRh~@qTC{R3Rk!a`Zf~m*~Bw+=~}6j}Wx(%-`Qb{cO7~x&3|>!!7%7
zcyK(4b`J^vu4*DKqVON1h9ukQ{f!kTUKg&f&kJ-;USgyZ=}4^;8~C#hMsGwhS))!>
zaH1}XrhO#F=ZW&Vhe<(4=YK*`^*-mXN6ud+8-F4J|Mmv2%H#PI`MDJNV71(ZrEcyh
z=LW3!{$*U2vK0VR1zrXheEGmXoEmS7<D+Ip@f47&xIj<0&dJ4d`YV3~*f@)RPm=!%
zk4+%yWIi0hDiCdHg74&KNaqKM=$?bEC*?Gwj2iJn9Sj*VQG>b6rbZJUMDBhDM7&zb
zKaTg(a|1oUp<wi7TCUKzup<dPXJ?Oux!4K(W>6Orc!V@LHFYD6Y<QFZQs{pn312`}
zmGE{U7Dz|JqVob+4FwCg14Pr3L7yE~M&g)NwLx|Rui0U840(ovUMQ>zWEyIffur?D
zna-j5I(<;?Tx*>Y&p%_epZUiTfyL@y1}wMnf9^8RXmo}&ciz8$e^%q>9Jglf`q1TA
zKwWtk!!+-b{p-6o>ou~;TpB+()a>kyHuEn5K3oAfk&=q6DVkXp<UZ=$G^0AjIB&)-
zn)vr=i8cGn&hKG4_C89*;hrK+_#vAPankAXrS}g}tyAry-DCe$c@2IA!OY`N6VoKm
z_$@I0!L|K*cb=MxdxgiObiz!|ki$;*Dd~m8@^osUxGOWVo4?-hp746wkKkBeZp>1W
z_Bh_i5@}Jg$J6T<#4^KDdA8FE4z)d9bTgldH6CAo8^_eWDR&N&)nK~kjrqP#W~sg;
zRK-qrIP~(el`uz<I-i>IIfq;BL<xlUME&tU)SU{sx_?Jz%uEBf4^KZolggbDm$XLj
zpXsT$9WNz|j3}y2x}PTA;@ivXC6toM-kK3gNgCN6BuXYb{H#4IFa52^qu8F6SKB9a
zm#428Ao)<Vsz2LK_I{gLU)S*yGvsG(!<O%}Q6%@HK%S<_>uTwCh6juC*8t+=p>hgk
zhnKit8xe4ZQ9f2F*j?2BT^x{WnQv><8!?ofZ<mXUh2#z+9vMpQxkC9PAO~S?ZeE@y
z7=07@2sz~6YFUId3gG-1rvvocKQNK{xi3ioB0!PB&_5JX30*MRbt}Nxzh-qih6YvA
z<Qn;mJiU-8HmV0RX2yRC2v%G~x<zN_p{j_!%?-9QE?$vB0&4Ld-@H0s4J$?S>hs>X
zIK0@YIiSC~I6?UY@Qy65;fDYJW9lu#vJASlVFW~4x<R@_=|;M{rKB6AySqa|T0pu%
zy1S&iLAtx7-r@Q7v-fu#0{&2LxM!}Jwa#^hyg-7K#DkgylLv`fq3WwXe~$OG!EQ5F
zD(~uaEo(5#Q%KV869ZLChzMzCt0o=fKV-^>VaY(}rmP|KcX7iyXE8PbRn-ZFCM$!y
zO@6ynHk&hp@8Td<PCam=T&N*Vh`je4k>A88dj0VCpJEM@0G1>=4arJI_7o-&s9qUK
z`_g@Ac=PGaT1ci-{Nntv$sQ0V79YfQ7arsG4j3E%BAGvYY#y~PXt|l3t=%DEP#o6b
zN_QGx*eSX(__{<wzs$3?Kp7|Dn)|XOqw|zfWBrgKz?te6D4*;jBWB0V=ER1Ph<)*g
zWmB`o$S(H)0wfe(-m6b>I^KQc`*+m~J?~4kRF`uwg3YC>CU>Y-_JhCVS*;_3$K_)6
z-G?*pIb@tw&$xoc-y*beU>g{WhqKA-yS<O8lYk%|AHfXh9b&<+I~U2>e9>!(;b8E&
z3&M%QhTtP(nkgQd<~~{2HrUs&qajL8iHJ1c|LE-c^+c%nAX4PP)#2vP{^6J7rDJD6
zF-^BgsD}qUHsI96oN{Dqoq<^t^WS3DCKenSpM0Jl>7KtF?(FpWzeNC%fRlt83mKdC
zmbM%m93X%zeaQ>d8{8SMRv5b#D^a4^3bCoFlFjV$mlHi07u8M#Yf1y_ZMlOIV_O;Z
zYj4*9B)D;po!M_y%G48W!z(3}^=~4Iag2WmM9_hH;_7H7LSU-Vfof+oBO>3JlyA-^
zXE!zKO1ma5Cu2$S@m!rW%Ppe?Hw!&}IwO^ZDs~tH7K?873t^a!SghSww(iUy(S020
zEv1GU)@pxKZyiTJtdfTvrIRJ78`XGk7Vv($!44l{J}4t0{f$I5YQ28gYlt&`%oe<Q
zOO)6ogrS6s>n09yc7ymgS3K#>_<=yxXq8e}!kkCz2sZxVE>7#jEn&X(Kgi~9G{qb_
z-4g5>%f4zpk{Wf<W&zM}A<W2UtAFe7J!g!L(D0Rl5Hked+507ys4<%y{Wuip5Xe@c
zbu;}d!fw6PH*cD`ojmR64+6r`;AEcu9l|_ix&O5@VAU0IcZZ18Qy$J?DFPUq$;g>s
zOZs$h&X<tOj0uZw)5ZGVNLN(CTD_q==dZnWn+02n>dYU)^?_p+jZ#maM~pCmXMZKi
z++%qJ+-7I=?>7YN^$xfqdML`tm!e}Hu2w0A5xRn!xS(g9YX!P|z+86DJq%f)27ga+
z+1lTy*&$pJ5%DAcPcv*E0LnyNmiEtOY7Q4$gG&uLJj02ltxR}qW~_Luz~Cow2v-#d
z0CE8C0?Y!$tUHxy6C?3&SqtQhe^#)4`|HXK2FCh)%8?+FZTz>Xw65NBc1;;;32bcz
z6C=@ema1bt@*P`DHx++$YqJXn=Niz8h7ml&S_`E7wJMR56+}geY-|ba$}&NTorg1&
zV`nCx#@fkow98wHb%71HiAt8IKGA)HV^K@6$`v!dM{Z%sDC$6(x-8mv??tckj2FYp
z3#G1ZYh?cx>oCtOppP!pSnIj_j${oDJt<IXUo1m&jO?~;tYV2@5ap;R$mQ#Ebc{Ln
z@i)^5ccu=mHk}?FjZ!taZM}a)|C;V6SvI`Y2z38ieU~(Dv41&~(RvB0Vc@zYFAAb*
zY2IVFDLC!5;?S?0R@Bh#@M=9VoZ68M3=Y=4ZPY;kj|>q$Pm+$u?fJrV(ZDu5c!Tk-
z(s*95t4o|bo%V3x*2Z4(E^?T=bMu(y{S|UF-|!}G$QA7gPqqjw@3x~4JpL<gZ%ZWw
zPI-~X|IEDui0YlpL(&Ji^e#$R`VY%lvScuq_pL5dG^{@iTK!hsf+D}P`tj-@-5|u}
zfG+7wy=dN!A0O`9$#Nr%Wh)vGY#tpQ(S7;*LdZDF)AYl|kp+P{;(y_?;QjGzC?L%O
zU@ADgr$F%Hf0qBcv~}B_3)Fxh3w3>C!%rx?!=g54?y#t|)G<7-{Ss~eaJM^-?P0#&
z>YetF-Sq<~R*y>kK<eFfNyzy4!qFk~uY>%~M;Z_YeUT*sCNomnxoL_`MnwFd7*J_T
zD))-dE|5E|3bV)Ni|#@=`bX$u@iKDF_UdM`Cal+a0(+N5b#D}RwV(ZDxl&r9A3R#4
zW9kiu)ipFM?GXqJ@+--Mp!$<gx7IKOMkppw_OJscw1v}tcxYA~qcdN`<jf>roQnHw
zKPgJ1LT6`6q>exE<y$=}XU{XrUmV75UnBJPIc7^$Y+H5WnpoYZa$vVhl@%r786s^r
zF-sLoP2d)%a9L8K3#8s9TT19eW!|oYsFheN@PO#c8sjO-Sc!qMa(xLQf~GmWuOpR$
zR_zMXgBJCdH#e;bdq|O%EB?8gV%+`s^jJd6gnv0ZMiK!^ab~ZCe`VV?u+J<`q>;5J
z0lBOVXYZ2RcNBW{i;pJ>b!2nD>f2i;>(!0))6El|A9bbzHjIpW*U1(NI$-W10uoAO
z2m6~t<q81BKU=Nj^{O|&4CnmsO+SEi(TlBj28cAwZWl&C^;}q9p5N4z3QT6sfU7nz
zj^F@zVhK;{pCvKFvQ`Q-I7UWB(57<Luj?=-fkIGKRTT#@7~dI@o`EJ;wuJpit1e)S
z=pTxgC<EJ!SPPjD-j(b^OR~eTI*CTv<IR%ob@UVPOS!9SFVRioXTq*24^D+@K8KC$
zc+2&#i(qk?$3OnhRA3L~1|iPzQN_&}$>tyD&r%ZKZ=l7g;xfGE-5$3l^c0&KP%#$<
z7qD>Muv54~sONU7U9XpT`Rs|<)ABMK!Vpd+EKUS15WhGRX?2w3CV#ENizCifR?*{T
z^Qg>}fWJV@7gjNy&D7pb7iLsG(~e2LHz~>WEnBcK>D|a9;P#2t*L%O!epnE%m@a4I
zhRh~LI8ikr^Sou~e6lF1q%?`d_q`7d3(L7T^eDbJBIWA5zh$kKtUZRHYw=13gBjPm
z+ydUx61Tw<=|tJInySJ#I$QA1!BkBrK@fJ`TH*l`Mk{Y74d3ZTNkBhDq4(3hyO(&`
zYRAuf<mS6~n>mtRCiKHfqZ_*ih4q*l`+iciWEDyty=HT@z=wg1ifXs1Q94Qx6mL};
z3e*?|t+`r(dr;K>b1pmXjr#$0$k|?Q%ms*|p5xg$I2ekjkbXARJ_46AKo%3FoP@bh
z06?866(p6{Wm6BJM=)XoYGok!g9&&8a6qB#&qXRZ3e@Fzjj#Gyev(Xh>$iM;SCm!q
zX#MO2lx=#qFkW<{yE*W%pE=Xyw5K?`x|t}B+regm_5XYr!(Yy^vs}7Qop91nxH54U
ztmJ9+J|nm#a>k%bfHH0<J;D7cAoaKL!y*&zbC(drqcGaM^4>?j!gAWJp14F-Tb9^I
z|L-EA+&vmbA|A*`Y2Et{TDUBZFqtRMcmKo`i$_Su`$^imAk)sTbK(#WpV`NF?~r^5
z{=wx5$x5ms`=v@`9^8L2DTy=O*9vuK@;w^N<6XJt>U(_rc*}Mut80wBG-kbNc0+Pw
za%3{-_g3RAzFOMbAZ@=P(I7D0VNSu&6VT(oI}2DsgKJH#^uCJ|=d?S*Vc8wk->LfE
zAdwd+|MI40BsfjUuUSd5Q(AtrL~Y7rP>ME6%u|KdguqGUuQsMVA{L#WP5D;K899Ui
zRn~TZ&V$D78L>9wQWE>NFImN54my#qHf>&?)y7Gs0{@cFk9wXpQRNRv?NC)!$@Xn{
z5xNyrOiVFnXJ_r{Oh&AD8Tp8rnVCN3$Z>~c@Z|pi0OHy|12?+Cu`w#pm4eJtu)@I8
zft-?(`|{f}c7xgInafcS6SlNJF;HS<_yB$8J~a3Qw#}$Ok;)JV_5<!i5F~K=E&NMy
zKELQHD2>N(Vfc7UI~(Q776h-x^|y`}#b@v7!O+X?@w{IXV8ejO8oTOscs-`uf?{X~
z$6xgv$aX06wdQ8eVH6!_Dm7|KFK_CHB!T85u2b!o>629mjbpVfDYkMVZ|rjUXc1a4
zU#aw-5x?@<GjyB@GTJ6^h9}3yeqU`Lb8$70b#@sBNz7dhh1(bcdli=7RF@EhzuLP(
zQ0rFkUoOB*43V`sn072hkB|*8B?-teZBH~)q5KLX9iR)Jow&8Xcy!KLMfj<yfM8GO
zV8&6z<)`b{1JgS}P9D8{K!8XI1s-XCk&bA>w6A2Or2Yl{0<s{`)F#dNYUaRXaN9oL
z^PWPv?J+))F6nZ*fy&N75v|<5Q>$!8L-EsFHh0fyi)9RZvGUb{Xaud(4*fm|&fba3
zwCAs*gDlI~n4QXp^-7*3P1Ag&kr|Emjz|dl`wI?;VKKrK$>%$DIuMWCju0*{)sec2
zsAk#;bJ~~J<NJ;n7#JwvOzAcR)K;^hZ)yK~Nb8BvCdS9ceijny%FrDm>@=y-uhNJe
z-sULd-0rn@#lEBk&n;HOTflpovFJ5p<#xyfzGC}R_3s&)>USRF9NWC?X-Ar=?7#Nf
zSUKaj^IkBsM_@MVKnQ}d(N}C}=9l@y<;8*tv)G~S0To(qFFu61N(592LE^(?x4qHy
z-rlK@ALUT6mb;-6`?VA*SVrrAPofrwKeYdt#V{o-kSxiJwj1;zR-VPiezU>A)v#;I
z%8v>SO|@PGx3lFZI=bm``xF=>551b1teb8~%6-;fp7J}qIYx81?~{9nirUjT4NNYw
zdLmV>`%~I}dz9&B6&Jzayo}twW~p<3T?@Dk28MV<q|R~kBOjTjIAqC|1^g#d(U3C)
zLS<+{bCILc^fSKX{OW3bTU|HbBPLj}UMoz7(sDZDIaa?Lm)8*ow5bN0r|na1!ta>d
z@BA4rlYb@+7u9OSLTY^=>{|6FZR7V$AL1sfwQcFhY=Lywp+-YM23g?&|0X`M{s4nT
zp@Nrz@j%Yiqxno#^Az$wnf;+AvMF{~<dz}tg=phG7Q8&QG&xc6*k8QdxU?1*!>7oo
zjTvpZ1HUuS3PTVP5#`WyGl>1~FlhidSK%2`pMgY&)!><Bu5#1SZ!8VeZH%*OP1HZY
zf`WWgR?FGQD=(T?cPOLGF{7zI**~eE;$!&o(dXHd+vz~0-TMg^9^PQ3$>~oz7rL#j
zEhHr5oyk8(m<k;^)jl`ox)bMcGC`DqJ8!JaRET1FwHE(0&R$gCu9ffT_}=+$YXxu4
zVyvY%;;dQCXTBfRG{c1-es*(aW>0&Y@(0x%)@CyLn2!H%%3K}yNy=gbw;`K9v4UEr
zq{ZR~$D=#Wl*W*BRC0kVN{mF)v*U{boCP2F>^CE!NX-$E{$%8RSLD+InGKcXlL*mD
zn(BRC8GX6V=UelQ*VD7AC?5c>I9dNmf1-rE*YRgde&j`}{pSuN7Q-sdv+GOaE5qFV
zrFy2-)zv6>O7Iz6E5O-l&ASaPtS3c&>PK5HV8Mgo8-(NGY?SUwb9UK`!_FUEO0T;&
zPf2CYp7KY|kJ>X0@)`RCF9$l-_eZPE8854j{2_vLvmfb6qTfOh#&xs%J<GH-la=fJ
z`Aivy;@P=H(mD6;eGr*DhO4IjLOYo1ytQv6coF+!T)*@>Q{|q0ViGMp#rQ8~-zfNM
ztk@v}BwgQu$K{Ag_sq=o%L4(qAWHfUV*RkFG-(`X{hZF0$^}iq*Xl22U^keOK5h+u
z&v*jWQwFzFj&w3(;fw|7nE2sj-$>{+$5n=i^OO&pkX;{Ez*;Qg1*P%$>9{M}Ir_9Z
znUVgb=4?Sf{n+oC6>38M!cOSBMR>W-zauSY{SY?!$O0{1_vWXVR$x~z5_V#U0}<2d
zH`-daQr!psfOqoX)UEcJt9P`zJPRU%q`<`G<v+9Rf+7r(2ajas?to`F-Pm^>&(!a@
zT1B(TlWp55+I*1bH__{DzWJrA7pUbYZGj>6R7|JTAuPy$54R)0gh_bYuQ`M#CfYM0
zyE~ZOy(CGSk&2TOr$|N0#>OVMPy0RSBfosXYinz}dw5t=F!}}1B&*f70Ptm^#mym!
z{sq%FX(Y{olH%>Paa`syb|!hph~ju%^{7O@oAkic%$a0NdCP2eMokt(gZC#W-vD{u
zAz$MX%})&ilAV5~Bb+fplw`!~@yhFKz;qfqMJ6~VW=!<~*54P@yf*c@Y9+Qyb^R|+
z|JS!_m-WRF!vZPxC19#v0xLz`s<9U);4}m+2%6co32oREnc4NcMR3a2EP$QZJ2e%-
zQXp?wv!KyvPk~M;^Ak+;!&IbpW2g*Drw!O~?2w$StXWjmh->FV+k7CJ1-#RGe7vNb
z4$S|!6H0N2D|uQQ)7U&G*L6u6+e5=8>}2%K(81P*F!_BtrWxQC`euX=-6=Oz#a5_I
z7?&JzOIJFma|A+{3JXVFJ;aG&A@eYIGcIjb*)l@M>jsRIq4ibFv?d*piIv@AmED5_
ze-Npt?0#2vge+@=1`RDFWZGsPT!BoQXl*%cUI~>+Ky>&4yl0CS&e><5U_?S8Ahymu
ztx+l>2XkUzK?&^4Y{T)FM4<(LwzLp{ahR3W5z_W8IRypkhYvA;D`Dfs^F3ZDHVxm6
zJ0#I<+j2}%ztFIAA0wZTdo<+&9u4Y|^WzPwmGhf`-tZEk=us=m_4YL=g@-<bfjp&W
z9Ik}b2z7hgUt~>YwM$|52rx+Mxs{bDV5O4PyZ`>i?ge^F5D_oxSlS%*zbz&u;3t?(
z{(=MGuqjXk%#|iy`D83rwVqViE3#_RV}-*bz#${|88LI^Ks~ozrSoG}P=|0ztW^fZ
zK0{>~s(L^Fuqd7|`IUnl_W9%>THRBcmRSyUxrxp>E%D${1_L-wl&o%?6MW$a_$6oy
zi8B9-;o<$Gm{Lr^0B3DbVc>~M;`YxH+MeMhMI$!y7Nn&)pS~qq9;*V-0!ii3i77MW
zVWw<hr2pJqVX~)p!xXE@Y>8*VNXY+LQIhj>i@r)xCp5Y-&4l*-tj+U#Kaw(PMy~F1
zK9<S_dgAXqrCrDDMkoC97IYJ8=J#`!FHIW{xOjL6j;M~HL*su*3mY*3vhQNWrBTNI
z?rzjX;Eb1>-$xP-rUV`OBZO?1RQ@X~XUZi-wVW1&6v;{Ge3p!a!U3_;zW}<dtg9~G
zQo$kKA0^~1*du1yvWJD*2L=b6S3MHke{^~-nOeTpFyi_?K6?IAD!SxmR06i%y#|6s
zD(sw;_LiuJ#Q#H@G6LDR3xK(wq1(r>X+RV!lzk^o9wET_4Fb;=2|ft2QG1kcrTNtq
zT!oav&g=umYGw}-ZqGW~@ZRNAy_}uV#37A9TCn_KT{7i&iP**_JDX#8(uZh>&*)ha
z=BZ>BKRmwCkr29*5r3`M8xp>RkIE~if0q^O3ZZhsk7GgihO^OyB+NxXPU1>uOGoWJ
z4bvepZ7sI_eI7&j#>VmB5Z|96qz@No9ZaE!w(sn@wu22Ih0#EyEHzs&yuW{HF2_D+
zLW@633eVYeV7&6T$eefwSTs1b61a#em^W`U`@O08H3APG(ZSsUlC*PAt-pRHK0iMn
z-3uxx*k!0qE-x(nSyy+%6g6{rJzC`~xZd7M;ATkMnC^i2<z@#ve(zL#VtA9r-Hlh^
zy8VMsZK{C%1$9M<VQ=j7RYWu?z-;LG9gyGbfgU|#^q0N71~p4eCPah&8y7w>q+w!u
z-~Cqc9kLy;E(FzF4rdcoDC+-I^=W?qM-~ndv1Mww!T511M;xESDrej4xU$*&EF#Ka
z%_ZXO><U`2;}^&J!KwIV6&ZfyuttsFxL?4AycV<$8CfB>H~ie!{yN$*%5x*8EU1oq
z9n1Z*p*b8n>4}}ZzxewDNX4`RBKx{)TOBJ@fu%J(XJ^Qw+r=ael4zzbBL&7wA(czk
z>D+Dbw(Zn^O!$TnP>Vtfo9C;o@iwxpCFNEm^v^Y==$L8fc|y{f1)5UPgh&^sShG3U
znhfCK<zhW8Fez(@!esn>eMbz0cF-e#sj6>n3P8=|nI#`aLgDBhi5M7=4h{~^wNyMm
zKdWOvCASZ(S%JC-NV6Im8&NSZ0>6~mTxqDQs~1eA6-2~43%*&#Bp9gJWrnqL$t^99
zcC^McoU<WSR~Ie}!GGVwsxOw6=~_ZS>h#%)6`h-6!gSx3_x!;G0$$HO)k+6RXNRHj
z9^3N`nBxMCo@^@XYaI=GDy>ryp3|lOAHIFQ*{?UzD~AI8no8B;#sV~!?ND^h*WgYW
zG!mjse9m3|Jp@m?$G2~11kW{bORitcR_%0wNUp@C`1Ft6b@2DzbwUQ8_#vu1>knx?
zrAXSnY67w#H1NWji*U=rhOs3p=jMw^)ql|M#OgY1Y8B$$2pU_SpI}4H$2!Yu_PCZ6
z$!DoPCRI;5LX2jtN@&+SR4rL=Z4**a_HC47;r#flJ3%AVwd5xyt1@K+jk~0v9WUiP
z>rG4Npisy+$RH<%Ld@qt#STZT{PokmAY}EAx`E)9(BGR-NVX~(R=FbaBBYHuoM_M+
zN$|fS_O9JQw*YEqa$@53RD)%kdf*ioAVO;Q#k7KmJybVB-JouP^YRMEprv)7xOnH<
zx><se!BqPYq^6LkbAGU8Wn`RAWguXr1(D5IHPKu-LGRh3vi~TA)PLE=kgi7q`d(o2
z#Lv%9ohYS9up;<>S1|#vYo%YriYQ1KP2sR5is8GNx7M9G20Ts@SV-E$gAbP%`AQ|!
zQiJ*A6yIY-`ms2$T`;})-BYe#G9@2b#@KPbw;i~2!5~|s@Kvj*8&0Z`CC|9*I7CEl
zToq?Y5b&TvWHqp-uLuMMX66p?SI33^dN*#8A*81onMMf<A*+Bmllb-yhHLqR9F2xK
z2*ISrFZUfbv(c3c26j5h;UkHsCn*gh|L`-Ez~&~ttn>3H`-u4S>zHm1>No8J?CFvV
zpG%^1i>Qa!dp3G;kP-o#Pl38h%Ss|3N5Q)VFYLBfXO>=|<?T9ffTf5dhlK=YIdYc}
zfd8RX^b6d9%`4t3ruH?U_uF&i2BUpgI5_jK>k7c0uJbDxasmbiKvRF)PNZ+NKScvZ
zim4Ijr`2dnvz&8-?eTY#G!<L@-wWJK`}wa-SjyqwvYD`WTd=K=3Q6J$1XJSm|G0HI
zF<h^nu^Jrh9aEKU%sxrdtpJ`rB{OJ*gvWpG(Hfur_rir~OpRE1fqeOW>;%Eqgcpp6
z+f{tq-o?WlNHRr;>|d02i)QCj5j>Img>SnE_czhKQ9xt0IwO_YJdTt)y-4L!oVKf^
z6H?gEi)qwlNW!HP?H12ZvnsyCv2CUlI(q3*BKbeJ9zW$QaglNa6+I{xAMPGPi`H&m
zRVA9WkSP?g1&aMPzTGKGu&g<MA6}rHCVbm=s-&zVVt%d}F0D+>D4RE`5aeTo27mfl
zDqh{Ndap{Fe7=k_Rw}M;Kia-)8OHyV^u{02iYM7nS|IvqOAhd&AivgXgocOP@5F<l
zhCX}R8{oj5Q(le?762eCjEu-(BK|70t!1QeKlAg0udW;ceWD6@cr`SDT$Rc%Wq_*y
zM&}u8gKPCqo#i^s)b#XQGwOQAmd=cnB~hp)vxZemcFNRy7<eKqdbXskyj;q<orCer
zTS_z_H$MF@Eb*s~n4)(v%;%YMGJx{@Ix4-eto>335_l$mNq+YciC%jl{ckgYYL0<z
ziS(rW2ba-wdngg;UC=Qwdg{hZ98U2!IonThGF~oEVZ5f)U*zV6k#fa%Xc=m`8lj`C
zLr%`TEiT74Q8y|bXY(_J_P5n5+LboNR|g4QgV=<#?Yo)k`0kt4P*zUf9H}9^HGi<X
zhVQZXM5@lIyBd%iKVNeofJktTBzixXfRRB#Odl6M>3DG|M;u2-LoY9kjw4;@gTr5o
znXr}d>u#P_28$23VAfA2`scSdFDMCKMBS+=kr(&o>Mcw{lE^#vFK@8fCUtWac`vZ?
zv){;}I&uM<#_Pp0H9D{g4NXmG=;#4L*~H}JfgK&6K@dQ?^MQLS5#Y!wm$g>}uGkb5
z7QzB#QijrDb{!O_CD_*BA~XWj5s1f!fcH3+R%{Spn%)1)MkiKBR3ww3Vr$AP<VS<0
zz$MX9H}<>eN>+-f!IftFpOq-SfV}+>ZfMIcmKz6V-hiRmBQZ(%xn=V|SScyTKn$?T
zR{r;>*KNf7=B0FRV*Bp?zRGfrvHFIP!tI*fzu8fgwt{J^G?;rxdGgFUdHSp8O~o)A
z!3B;4c>|ke*iE|OVM<jFw1ccH-yAnYR=zc7?S}YY_o%lWbmU1xq~g(veG^Ws<^m8e
zg>@Lay5XGe{{j^?<3?tqvh0}M*eDwVJMuHk7wtH~ZOL(}aKU_QcQasrVIT9=*-p@Q
zoK)SzppXHOJewB@-EOlB`M>xZ&q*c7d)xOu=&v+FPPy2U;@>k~B%6|2ui`FA>y*!{
zlg4RJX!P{=hwSa`x%c`SGvk6&1c3G%o14O}uAFdiaNw8ZvP4~x0tXCp3ZlHcJUDyG
zX^2-}_Q)Wb#DlBIsIrvvw$+c9md8Twp&l;2YO_i@U{`_nb=tJ=!oA<QRCWwt9?NRt
zTmhZmrYz;cSdl;Z|9^wQNmxb>W7nGn(^s^xA~hg)YX4Qu=XOw3U^91N!v%myRy?NZ
zgZtedERQLSI!)un#?sMD8Y55$yR4j<jruH|PefKYe8r_AS$dz(I77DUJ%psCyq_~e
zA3|cQE=;>Zcx5O4_2QHaDVh_+k=NRO>`Rqeej->pAy>jhOCd;G7-)Q=YdCFeOvn%r
z6P(rW%HjE0L1Xw5ToG!?0h=}=q$tQNDr$Om79&({-_OSxJun*clWt5%D^^JI1nxq~
zMIq4zCM6AqmzBETC%8qH148EVIlyzTWPNgcf28S#$Lj5_VoN);QL^r>m1v4O1_pNf
zFyPWq05d({F5KMQJax71K48QQz!*^10WRF~@-jG0R8&>7=MM4N!~twi@9N>fLHq8c
zc23uDGUP-27u~@e!6#RzL%AU&C&Z|jyr6a_Mcqt|BWF-&R`XQ&GH5nJ{5e?qUh)0|
zRhiOo)nBC!eL~<Uzhs$llCKPcJU~9i#>Phdb<4})e-HIvr~p}@#}1c)M2hcykvUy$
zwY`qFp);w$6+}@enR&r+2rhZu>@BUWpAIczn|0CNJ4XclC8ILqb5jZUw9sBg{B6K&
zw;hw;8>iNIIuTR9w77?bF>djWMa3kYn7jxZ;%x{!nakXCT0c)QMbSg@CjKf~Pz3(7
z>|&+uInP%9ac^&T$lp@3Wi}&Hur$J|pAv(L%!WsmRli}K{%EGI7!@YO2$26uzOdvi
zqMjbzBN|j{?Se7p8QwD!H$}cjTeY3uYy2k!sl2z&hp=whVIb?p){8_#C+a^PJO_@0
zn_Qeykq8Pjc%W_eG^O*DOn*t{DP=2AkDIZA9)C?G&M{ZiVi>^Cw3ur|(mVIgVFdK_
zCN8>f#3Q7u$kq##P<IbUeyELT^1DJ5BPjO!RZaZPsEK73CXAt~Ti@Mc`E9juk*_0e
z2=E^<+C(%wJolxsr}psw@*(g#K^|)cQyg#r-`LIaX*-zB1jibP`0?gRb{W2>N@dYT
ziV_Q`s9@sJ^$B^0^}J=pwCsb@&Uniih0SPCG@+RGUHS~&`k;20kL?#zifqsNIfQFp
zeYWSEm{7dCqE>o9ua(foPS}ctn0x2nC8lyds$}oJr?HV<P&sPF$uGUBP}ck*^t?bQ
zfgv>+{T%FJ21(Ru=lZZ3A??Q4bU{GR&LV313<1XD`3BzyQ67c9Jw3XR+vr};nyfw|
zTzJc~hX%r_8f(nHgc|(42o6Rv-hBgjKn*)j5n*ACH%M(Tw26QX^2%AwRsa_x?ECj%
z*l?VW%LD7b{wK&15kO>s;u}8gXkTpEz{4A8x<oN4vMCu{m%7=TM=vi))wdJ@xRBOR
zgf~0=Qc^T5)2Ig<av8lFU)%~V`~VJS-}v})_|M{n{J*8)B&=9e6NZxp5Dy#nN4Ex)
zG`0fd4CKjxfB69brEJI__+=a&9b4;;1e~XiMsD%1MjpHvo}q#{P%s!bP^SsF{KZSj
zEY`6m64I9|-bm<N^L+}?`{nxLPU|4Ii9Ll35j2#MdO_gK;E5h46D_WQNR)lv&|BB-
zY(dxQf&})W;Qc+B%2uWbAzDN>Cwf+Sx5B%Ag(eTUS~nUAGIWJoAA-w4KBOL0c7(c3
zo^QDSECMX6=u`Zyu0#~3>uEUN-w2}|)${&yndsOw=usq%BTgPxq)p@?xNpP*Gt?+7
zaR?g=O#c<Rk-+4;%fe>=0N3AcyvpvY-G2V0nM(~(t{-f*6XAC<dAou11r*o%=0lBX
z)i>?Jnwq%a&=wGAKXk#m*ZS|Xz6GOMZ?lr2+hx6-Wi$GXVl<OaOa5$nYdiPqTKr^3
zts-m1LNqO1W-6qg<A-q8_aN!BJhKQ9N~jb>9bG)NG|bz|e6v(6<s`dln%GgByUZTc
zmp%)21)W$a;|VU>@Tu4#Oo`T~IPv<90!f5}q07P%`E2f6*4l-FswPWnt|lz=lJbUD
zHuSJGnC{=$ijhCYzv?7;9mYBlLB#)?K+J9;@NgpuzYB)JSVy6&@6KH6?R^{22|6l!
zv4rxi)LV5BZU^G`#J0QcJv_)Xtf3p4n4fP)JjZLvCh#igZjN#A{(#5S)%uVT!a0ga
zE+m$iNW1M{e%aHh5|FsaP<AgZJ%BmTi{<}qB0y3LnXxc_e!L$Yn==GoI2@yeLy)VC
z2GL%}+V00{C5dkluwL1Z0DuL6a<M5#dV2bw-{w=Hk&#<v^{YZ<YCLnL{~=<Vv<?UY
zJXoIZ8iP)~+)71cBZsSdTh8Rg5oB$IYCD;?axLAWOqa+9cpJaF*_bKP>h`ZvO`QLO
z1ubmc>Re9=m;U7uQ&zkc!B<@yADQ7U@b5I#kLxdyGnj=2Xo_#ebNQZm+R9G5wN-@%
zpV<oib7^?1k&{EA69g3tg}Od{SoB5X2@dzv)=l%Huf~PA2{(YQfiwU1{T~7rG`$vf
zwrBH(%W%`asnJs(n;Ev0J&NyAB3Hgr(MKFaoG3AEZS8^K;pU<)ng2Al;P-3!o5TT(
zpxgPq=Bpg2h@{>cSMY@(YS`Zp#FQ;aWXwn@eIx#~ir1pbNI=JEENa1BYl0xls{5+B
z;x0+tWq<B0-hf(&PmBN)(5<z3cUKmmYH)123tPtfza<#W;Mv6xP4;Qi#6g5<yZakK
zM`_@5Il=*0LRj<~VEO;9xU7lhDa9ou1Ol^nU<&wnw;2O+Tq=QL*c1r|G>=9`<UGzt
z>p(8*G)e_4#HZe73ED&F<`vB!moM<9qnW)tCtV0-#0qit(Utv~Hsck0G1el_hq*Av
z6zvr~51nyl>I@>4nsM6S{2k8Ui+Ol5on%M<ldw<3AIZE!`Cc9$xn{DCHW`@H-JUzf
zkMP-PNDyLKYbA=goUrfF0cY&U@3KEGs9!NTG4ilv$}Y?TW>R_ydjx+cyTO0HDvwYU
zZ2~cDvGw^)3GUM13UzS%#J4}yMH1WaM@ebey^UB(MbnLWi~<5L`POC41g*OA0DtK0
z>;!KJA`H}%`)xk4zz1ttq!TF$veaG{(~c(yBeN0-nu;Eynrybzg_N|hlxp}cKEBY)
z^PRKJ7__|eVk*)xRkN@MXT~RkocY+4kybvB)vLohI)G&9>JkCPQA2aHo{#SU(*Mqh
z#VTD+SLwXlG-a~*GeDSuvaW8sT!dT!I9~~Y#|I?9V@pctKt0U)8f-yB11E?ANS6|M
zN=D3@0eS}^pdp;h13t!$muripE<E*fguJe?^{d|Kc}jyEdSk>SBpboIAJb%h<Gnyb
zBo>*w_HbAdpG9sPOK7jeHP2O8lg*b-r^`~>Z%I#Qa7$UM6|=mT^tgdltBKn>UF*_N
z^(Ri;K*NZh9)T)yfzZ8;A@(?M051<B%r4Vw3vPbOa|NFI8<Eqh_v3xP&8b|96`gYt
zL;oXIYA&uA<6f}gRQ3SZyLXHCTkE!5Y}QGk+uB9d7`FCH1CrJu1Tq0<2jD4jdi`Ss
zo~)n}<EmfMp{HMJbTBm;NxiNg%bPm{v1s3(9~`CNAy&GCHCa4SK+XHf<d`xk@7J4;
z)uNE!n_cqq+L;7TZ1!YK<Pwu=*m#oLq}A=@U8%DiPlbHb==Kv;Ju59gSTv?ft~6JZ
zu%y|^^WWFcdDfsn`vLR$R95r9kvI{{y#IYX>;qmm`&^mb^bdY)F1HKdNf!7jQtgb`
z?A04&8moszKp-O{g925hHe&%OF(V1M;(-#Czexe-l2oZE!8FS&=`}#fM!P$-=Nbk2
zO^S@Ou;jVV+vb%~eA~88ruIKjMW2)T;nMjsMI(|J&aSVG_hPeO6X2uJq`}Bo5ald9
z9Qx<_y5b|xt=V{m(Jr!Q6WedYIimR}kB}4MxSvvHqN<B<yarkZ#q_-<%&6yE@8%<6
zFcy`r+|2hPYKn&C1Q%MMR2MxV3YLIh<SWleg%+IU5<?Lr^mx#IO{#0T-uH*Whii>!
zZuQwnga-ghw8A3yi@SaY3pr~Yq1e(~x#6|K`OZAcPUC$!6+Qh?)9uMvU`U9vg2J~M
zgFLjbtzd+dlZBey;5j?8KSxvO8TbW#E&t^L42lHTg(F&Hx`YH(y5D_Xe!e2SQEMD$
zAWg7A&6YWsP06D%cTBGj1Qkg0B|OejF<0&v@^Q-z`}I+-c^xEx$^oKdfKVH$&j-c-
zLB3i2wN$BR%~&t4=k!<{c4UAA9iM=ShQ|Es@R4(;8E|*qI5?03gaM5f*K^gejosZ4
z71}(I7v<~gTXORNXioRX^=k*UI*~Ff7T_6wP3LM&iTV5CppYRUMQzkln*Zxq{4xED
z&(VsDtBUczv-8uGRTlFE1vGAd6`CAU>Lg4|#Bhwba{ut0@`jS~lJdj0!)KLBy3i6*
z*|(Tgxy0c=5hN792mvGbUvkW?nYS^61UQr+lvpuq%xQa)*I_3iB`nk9<e@PO^Yr}R
z<6M{sDXXMB5GG>vB_a`-|GV0^D{G;<-I}!MvW8nqv|W{X(BDMsZ`To>T6%%;EmC%Y
zPi_<JP!wATR{i=>@orm)HHC1zZa;LRC2hikPehH8)V#0#&9|0k`MdyKRaH!oO#z~(
zRA@1cSr|#I-QM_FAI36|304>Mf6z{X^K2&AT*IPFS6`pv7)O+Slo54)4lS1D_}Iho
zlX#D+`pJWp@g2m&?P<mJb@2#rUrhWg7EeAc?}bG8K(tEyf_>KyYa}jtRjXzm$NJ@N
z*EDwK`o}|}I0w~Vhnaic16r(KD|Wi?Pagl`&mQr7uD@dSPG=2HX15`^?43)QE#axZ
zJ`kfzW81jL<p0o~l%#xQ5B{)}xwx3<v4;|FW2)Te=7pJR`9e>;F!uWYQhuW7`w#{j
z3k}p0*~oG;LRCOmOAAa4d2CW5#n4~c5C3@^zW)dmyU$OjK9>VzF$A}q;y&nvAnE`b
z62ysp{fIMucwIpLwOWJzx7h@Ncu>R@ma`G=<1-5P_e^Q3moV@F2hlz}jPO0nWFVG=
z*iWpiGby3R_i}6(sJp;#uj}*V#2rQO#<$=XhL^j$(8oh2I^pl>c&$YOLIaw>z)<iz
z#xYuc-GU?$7pjvI{ABq^p;FqWb~D<3f$f01DIZQ$0`Dcn4*sdq4T&Tg7^S_-=`5W)
zeA}__2n;w2-R+s718`2=NH&~cZM1`m4<W{TlUawnnKt;+CSo46{yOldg#6q2f))8>
z!&Mwn4c8hszxe2?LaX}QQMp9iy}tA&s=BG`u=!x+PEcwDkP5CnXTuVck{~}eNmIeJ
zwI-wifkUHkF2cW6ur~QOyvZE(&qJ~wyn?B4$QQo;l#P#iJg{KLAnN&m?4YS;$u2`M
z(ztJqsCma9;c{g1H+;<8g56=@nBQ`D9hgi2yL>Z{%F*RR7Z{9MkwXmYPdwdOq(SYx
zJDLGbbl^DNZfNs8GjlXsIOK01Fo23DpRqjg)2%QCHda<kby0xqG}ko3P`}*8P~XGB
z+yK&z64XYj2G6)NkXv<Q<O(dO)A}q5zUMP2LlknV$SO@u2U;U<6aMnJF}}iaLgFuo
z@Di{gvps&f`+HcCaC0c@?p!c@_X#4wW+l4S)L`kb`sGPNX$=9bkIqLs!s+v)k*8(5
zASXu|nTB)dctx6ooE!*gh~+roC9j1MzC{@S&DpY}G|4Bw>AIu)%Dyb*Tz)$SGE|j{
zU?1-=|ItwqFGM%-YxccNXd}LF8J!TaTovY{wkxA-bQBOV`b8<u@Ju@_(x&OZ(-{<~
z*D9dz)1RYMl53op`wVYLRC;-NP&$`aK>peNct*YI;2P1>gNC6Di+^OKV17HlvG)(Q
z+gsjNeRRtkQI_9mAE|37w4{d2Eam@hmjr*zG4w!8?e=>SCkvS{x+#IsaQq_0Yijxp
z1DL7~jE%isN9`%KL!&07pV7c$63ESoiO~D}cscf}bOmV-#M}j+lZ-)dog46FHk&EN
ze9Z#g*#Y@CiTGJSWi-+1aqIq$3hEv5Yt*46tbeq4JUM9b06fvY7BjZ1JewU>`^=7m
zNcA*w4VE-iq-#Dt7YPx0D=~hRl1#rjC!JJjNt|Z@w-k#T8TWHP3Dlkbss}5fJ(r|)
z_1`+CCu8igJ~^3r`TL!b-%Tu^Aih1?^2rwTnrhRw|BKDpC3?HQXj!qH8^aIz6*$;D
z?h3qg)>E{ZhnSClgG_77v%h{P5#Iw|IbpI7BsKx2<jZ|p;Eh$+zdyS}m;M4?VNHkk
zVF=w~UC?v%!Y{Wj=%XB0gNKFXa-ITYJhNoAhj1|HiyZcFGLdpSF~&6(Q`y4rwj`jh
z*{0-Vzd?WN7Ct+ti)vfSd3rdTHF9;xmp4#-_HG*W!j~VJkL2fbGo8<pVCnTKI=jqP
zT3xFR87*nwoy}bN;q>gL^6&AtyFL-bcX!9_D2Dd&8q+Z`f!!vBA{CJ03xaXH+g}=1
zSiJZ@ynXu?B*0}PZu_=V{LezU7Y_?nF}2;1f{e_5yLwz;Z;&e<qvpNV?7{@T&kpc6
zR_+7pcHrn@us@j>NyL{5^rN~S$82CBGu^Z5$t7PS!a&Ly`(7iUiw1N8(;y}Py!n_f
zuZkj$258&fa7<Qfb9L{v_IiIxzGf;b7wS`DJ{V*<oQh>VnUFuTaGROc%!vEvm*qy9
zlzVo$6U7U>Hq<1hYy^sRdw0*4Dq`reeR8?~9xBd`mZiYCJ2*TXF}%nJ!n{C6X=bhE
z93_D|a)_jxH6dob?q}fSyBuW=$|sZU8=G!B9T?gGxTPGvJAurQ4JKRiBf6_6coi?k
zcPonM=Uzlzjan1YKJ*)08;K5z@{i(I*v*92&8v6z92?&_K@zat5~D_?ua<2qX2o2@
zVk)BY%lFE5F^B0Y{9z^@XS8sUfs<(>Tk4(}X#t3ipNf}bg4p&CZhRID?~WE&GeEfN
z{Zi1A+uw%|gRVA=Gky@sCUpkeg@=3Wx7H9%=?<ebT&J6H<YB<_LBEzLlV#hH`&AAO
z61;1_B$oaHZ(^C+=IsLwBO@gl+4=C>>=O}Ra&XweA_2mRKqYu|zU)G&s;gfnw?A$N
zzL}gbIBtr$dF5~eR+v~6o~VTdt#W}}9{51wJ_QhA=G7t%kOBh~YS0Ko+<@3ua7RhO
zWs2o04xJs~(VN$PA&u*<*!F9&Z!+PKo*z}+=t}UMhD@oJAt&u;Gl78R%rLVSH#USA
z^LxV4GO0Q)mkcZ}*#j<42w3z#!Rr9$r}CByDR!IXfzG*Hgd5;Y@g3Br>^2|X3p{>g
zPm3Dke~>XxMMW>CCqaYHv6)1h`b~A#a`Dm4&F#0jgX_fH{NiUoIyVCUk+WLsCwpp~
zjJS>#xy<@y{kTB-d?49xcer9{F{-4M#3QIgc$%kgbKLv;suBw{<W%CjYBRH5a(S~I
zOAlz#ms;!B?vfm7ceSPITc*0v(IcsTbSUd%C9{(k_ioGHx-K?H%W=+wl7$w!EV%QY
zm4T){%YT}Aa`Sx5Dan|g_LabnESU7uEa>)xnZ+L(7zjm(h5&PEAcquB+WxXDm^xze
z8ZP$JExb$rzmJqVkjXZl_u(gEjfyLm%m9-#mg`kpF!lrwlpN07zN1VUdlV2(%~<+?
z)J2oB&G)=#pIgVu*sIWZz|})PEQPD^5eEc)^}dzn&Nyi>b_JQ6cgx<}F6{z=qc~Zg
z!UDT~s!>5hkH*`}jHFT?)zDTzNT)f6ZSwf^ddj^=_Nq<?wz1y-_Nlf%f#M(d^hJBD
zfbRjaEyyI%YD`8Tf4Yj8rZO$T`MQjV@4Z`t@V%GcJa7BQ0p#Y%m13A3M+-G6d>-tQ
zl9E8`x7kuo6E78ItceQIdCD)caw}D+$s>u1jP|9O;5he`>WHPqp79{No-)!B29fNl
zTYR88l73i1wnm|~>9}F;pVr3pgPX<pR};@Fyc#v>1-_|Rh!VN}5_P0bOxTOQS-q6=
zZ4MPzFPvIDL<^)e<fQQ-lZ}HL>v=wF(lTo`Z+h;5af%O>bBb?$c~C;x-*QEP<|zld
zMx1ALO%WJGnX&%8eXx9Lfpa2*#{gXnP;BgsWsLzS76ce#U}1sj8ah}<w6NH;G-*&Y
zGy?Vf*p+#vS`}K$uV3Wgt_9Aqz{rC_!Jd-<lxx885*8W~ZtWY>m*25T$5Xy$wQ2NJ
zSR$&)s``@hR?7-;dZ7$-bo$Fx93LFgZs0aYTit#vI*UY$94uRZ+yNqrUkwd7vfmN6
z@xWCQK!9xj&58-)F>NmAkWa%+ZqYw0=S>)SFN2jk2tA`U(li-Lpc#n!Sbk)(UPyj-
zzhw5zN~8ZUJ{O_*p5V=5?(oU-!2MDjcR$B7tVsci_ALoSO3H*&f=Y~+N`<9>*8S5&
zhkGA($(x33hj%94jd-}O##3taFSn%THF_J*73~xSU4k)``}2<6#ZTUh7KxpDFjw2>
z5llawWFJnwaYv}HMRqN$Q}C>z)@l!{w30dMKsfSd-KGFh?H5_FZI2gg(Jy1={9U*I
zQvqj!!0V0?j<rZ~Nx&YXUYnMR7Xk08W}OwTJTRib<sW7d&r_o5R02njT8Zi)Gst)V
zgmD@srUJl(jp-M5a$=KS(;0G7W`yie-R?~=lAEvC{7R8z62=id^KSSfn>>Ag2`#5@
zUQ)-PvPvjFThp_Gg@r+f6R*$nWfnLrx0+vyB6Wbv+^cJEXmjdc{4r4}Dd9;X-Fa*W
z^vTtQXmUxQd=R7oLCI8haq}QD`TTNBfNYp~G%Q0g%ZF_*=0FVx9_7muV=)OwZ;ZMs
zYD*Er#jGt{v;@Tu!edw*q7Mv%785ydQFJ_Ud`+x5mBINuWA?BJ;@uZhmE=xKaa-(X
zP}EL&e9wz^Q|@Q*h}vBtMg`XVY-bVe!QFi_KCJP{o>M>uXQUlQZ{YS_xY^4Ee(9gV
zZgfO_*8{h7KM1~5ktMGzki%p?^gy1@_20t-1n=2hzvwhMh8-f$0+W1hHv2oW(8LYR
zGBqV-Wl1TiE;EqAV+IQo0Fp0G+8%7UudV!gv2-2^9u+VpWtGuxwmCKY^31<<xn6fN
z|EW!wzo(@gFCd5cEGSO#(rgDCPks*6AhU<=Fo-z5^uq}PZZ6ArIxZ`{0MUti(tTX>
z{&B|p63p~*^&XM71}Sv(C56kuq^y5$BKIAi`xV4z2V;Z!p4Z)pgy+WTp`t4KRfPK)
z&N?*keZ>sLFmqfp6REZi@@IX05(p&|b(KxDIa%%uEJhh^*!pZ_B3_A)1o=TTUEdGA
zd+w$$k7z5PbF$jU>@ORwHXaaN9X|Q!UNbC7y{|Z)=3)!tN`7xL(hw)}V1-FD)+Tdm
zB6~KP+L=JKLVZcle5hw=Ia`W5l1c&duWO$Q5)f^n1q%yaJp_TAQ1HIhcCF27@wept
z3W#)4QnII%J1#Alu>g1ksEfc;|H|h98?g&)M3uUt;%|m>vR35^no2>nk_Nt8^kp3h
znVgKnzT_LU1G^7lhMErD-`a4!G0R)ISgmIxf#Y1|$CI#9?M5_Tum^wtzB7ylh)fhd
z2k9_#-dt_pOE=Y!p<mS6&tnhVIv3~o-;chNIXZbt=R<U1))f)qIS#SWx*wCi5;sv!
z<{2WNdKhSJN&EoW#qu=Qh%f;F6cu{gPArKQPoDinH|Tc@JomM5(|VJ;j*B6WT9Ll6
ziketJwopNYk0*A9fST9Z@tw)Y*Qu?lLRDHRKHcCtTj8d`p5rX9?pV>&8uzs~JI^f{
zmPy_c56NpkyAC)1jUIDx@L|F=91hzmWCi!?8?GEgKl8$wM*mNzn|ac7$jI2VM{1lo
z0es;HoaS`;|Mo__+N8ZE4+FZj=^g!;H<;%mw%g#;^Lk<y$-Z~7bx?iyIICo^TW)VC
zQ%4yoE~Kv_?sOF%C?vu5eSzco-Uc!;kyT#a@a(4`y>>%vKyNP+w`uOg0I;))Bd1}c
zj^1Bs=BTF6oiHnWzGaoFu$0wwrbJaYUhb-t97d>GxrBrOBrMP#*feawY};kXC0E+|
zju*QqAQ8RABlJ5`fm|rsOG*Y89X^MR(Q<<w;7$N1CrLZIPr$EfqEU(LG={w%He`KW
zcR(UN&565FyZrlV2>oyiNW2Ag|GD*`2b*<QY|tso(MpVl@;6_K1>t+|N9r$2r3r4W
zte@}&gY{3ndpvk<Kb|Z*6t=i<pFEsyW$3FbZd_07MwtcE(GNTF+)j@O#FK{saP@ma
z!d;N>x1CT%XHd8KE64}HApK3VZPN!ch|jD6j>olg%qB8dT2FvJ&T0WSaXM41S~4&o
zlV=!{WVbm=`$2L99g18&5x<qKYQ^3fDGF*v?zg0~chjz17ypM6^$HW9$ft5@13g&O
zFrtZx2~mx*Mx}|LQvIvD8AZc3Px`I4-}M4hf;FkERxTkI`ctjTda5JxY5NO5ZDKgM
zgPxyn+Z_(e#GP^wU|`FhAdWYObLMoW$h&-2Htc)o${Qdh<AI+L*peXgGZ*()YLMQx
z-W-WjV=RkgJo)yud11&0_t5tO+{;E?HshOT_X}*+q))3F1L}qq{pE8tM*TMKqaGpn
zWBVPsQ13owFKyIaBDZHrW3_~+m34tygtJC&_-4|cT?R+^g#-JOcURAM?m5FE>x|q!
zU??tv-Ta-HS>pBG8;1yZ_kITqV<2-lxB)m_P+A%ZtU3WTL{eT}pxe~-Mp{meavgyp
zvbop`$2vdoT!A=H%tyt9;){54Wm$Vu>Lk*k-OiEz`QP5F<0b5@F>zuPV&Fb&mhl5r
z$B?ny_iqq0a<{F}Dal7sx#JhCfCQ3z5slxLM|B~DYCXi|5w^Xw-_5~eOQCYd-Y7mk
zzF|M#3<}29nfqu3ul3nB=_j5iSz+!%wZ?vW=hv?wp#dbVf#H+5loUL$gZe(`=n6D?
z7Fa7fw<PHs7frz+=CtLk3js(MWZ|L>D83<QW$6<xRu|;NSM;rR<ej#R%}z2uFTe1D
zi-o+`?B_#E{etp_!j0+N;r-1{9B+#Wp_cT}HyB8qO!rKR_LH53VVLutTLH-NHm^<7
zC349ekShxkBG7HM<w4Llpc0F%o&GPL756JYf3nHR(%3ufn<t1X>s8z!#Rz>L-*N&0
zQ$wQvuqc%JQ^2<=$xX!X!opvU>UAg}9E{kxc@D}Xpb!NZhJvoH92i7`oElL2gKz(h
zJEK9H5hT}uT-dC!5YWA2bv*&^8HkVX@xq&ovEs?AT+a<aq$&7mzJ8Pdr7j9AHNxIM
zPyT5Owwak3zyw6mOez|(5i~ls@AB{=xjS?xm;9Uc@>p=u7-x(d)1Ujt;d0r71|r+(
zLpBksQlr?)=5XS50iaX`$`weP9fJ`x=zc|pZUo5@-#L@i9%$ck6h)IRNDk!U&8Gen
z0X5gZ-{xRx0@U@q60yXMP(gkPoUd$!ID;hpy%Wb=d%M4njy>wjms@moM%~fT3n`O~
zwGfwJz25DJ<Tx!iPS?r<NBTjcAyw(5@B9y{i&RT?Yd%y{?k=XHa*Pl(G+{`>a=|ke
zc|hAx@T-&&2a%Rxv)3mCPcC5|4(wwg!e!4(5Lq3z5WT4hAVC`Lw6zdmZ!lRUOttE@
zt<|^eRK!dKFREZ^*(h4H9n7@#{^uuh809n&I3pR~0Q|N;;UETf>x01aGJ$I^z3!RB
z!qN#*UTb!I<d9A1u~4Ownb+b7cyKE{T@jkOBZ(y$1K(v}rcVXTcCklNjlWj?c+HMJ
zUP=brBQ7H&%Ew;)7bA8LhxMNGr$q-Q(uOA<;IJ?ZATp&xyTr8Z4E*)<OG_`pL+cT=
zt}(eN{CFWMb#|BEz*Zf!3vQcH90{MA1!GjPM}G2s+Jn4N<#}`GHYEFs;ORx-`ibTJ
zjUH>KTro1Tzb%(}aKr@Q>ztESe*%#XZuJ@8v)%byD>??l{ZT#w$41G?L$`bRWac6d
z$F_+(%q+To-dX>$>%6IbGf<V8I2hcU8G;fBG$){_14SmZ1eGB4e+~$RcOYQ=x-AY%
zt`s!KoLjAShS^_nYQ=PhXTvx=<G#WcC0Rd3(@suK0#XQ|#7rM7-neL#KBUnP*FN{u
zy$NkdFZ{!9q?r#^4ks#PjwXC1f#`mW%*Uqv$r%iHC6tuVE>1?w7M=TRT<DyeGbG6U
zhgAt(bk%j;C_r<25H4)C9OkFtYWmx93AE^*%Mg-zEW{NL5c}2V{a%Czw9F4!<dWf9
zP%_bhmTn;yE!VJd+ck-E3T0e3425f_rbAnURttCx3@ty6VGfZ+n*R~D9I7G8^NsB>
zW>g+tBl>WBvDiJkcRji2)P6V#u}b$uKaFY#PUf)Qt~#B)><*l86(*S_q|W&{B>93h
z@}vD4+vZ>~#h9!-TfK&Ii>LUm|5gCJv+6`IHhj`9M?0y=*7vSstD7)*7PWXl54_B!
z3XmK?@{lXwF}aNTzJ2>($DP}0V72lod*grL2PfiLHf}a{o5v68kvtqW^caY9nS#Zl
z<QPI;D(`rFO`iUJ`#RzRM`7AeDme)kP>92Y(Q)5-2qBZsS+$D$=0AZ;Fd%?b<AYZ4
z%XSL%?9B7MQjy9l7qV27MJTOZu=QAL&-r}gvdxqC;Defil9E5UB=1=V)Fbfle?co}
zKszHT^TaeCQoiP%`yy)vE^xcD6X48g>V9(sw^_4Z_PDoxIThyl6pe-<87~4Rn{F>p
zi`aTmBBZ=G^M>?Vbuzez390|DwXcq<YW>>XScrfkASD6{N(#~)f*_sJC0$B)8(4re
zNOzZ%G$<)u0@Bjm-FGg}`F(ecJMR7CyZ4UaIm3;6Z`NAx`_6dgGoLB_Seeb#-Q+--
z)Yc7ku+FZE`dXb6b1gm~5$ZC=wc~yZzvAv<PMGxnUXC2E<keF*K6x5nuuzeBlKGMV
zJ!@=qIQuqXb;;YCvBM$ye#SGyze8wj%biU-c;d#t265(};`IrfUQ1?X<|Q^L9Wfg!
zybs#P;IV)OJUR{93ecv@K;Wx*H0M}ZFtY04mAm|XetXikJdg?o2D-Ms{wj#ej4lCz
zEzVokG!Wy?^C+5Xx_6T}EX#8Cagg<;#SDs{c3<Dtl`IE7%i=-j-?iDWQu<Jqw5`?@
zV5KHzJea4ADzjRA{Cuprxm~fLb7ImOV#>4HJN?gpEDN*G^V*+xhy+u`xzaJCz!RP9
z=6m@Yw}{0!k4sUFh#l}`Z{f9k`*GYwO6C*_8ev?{8x+u5hbB+(O*EIcYAowAYG-*c
z02~L*GaAr%LmALqzhRM&sC~l|_Osx960L{zqW|?nr6z?1`^}j@n1_e*7}`2+zJ_ne
zoMf8P60)N1s^_+>{fe`?wl|IOj_6m`<*~V%jv8}nr?|n362WSbA<Z)ZwJE3c+Y(-T
zw=8+{(brc;?GB8vd=4CS@5)*I{@Ki_^VBhY54L@U{mQdt3&R36)1G_7)*?V9h>Xm<
z)C$Vt;M{|uE3XkZmMOL09OQ~%Chvt;-!UfMRD9~l4WYt_?s)JzS?u0FSp-feuq4{1
zORMU;mTe=C8>n+Yk-bDrlnY;Wv@Dzvn=bd=PleO~UD?{`2aAiLq&&~&u@tpIOYyqD
zn23u@Ig3}u-J{)QbfJd%CgY+zMfdGiP7JnFQNq*1P{GQ*ocJ;Js`Lvy)$7j1)bZ}Y
zAg8s!!AaU<|7YP)$v5svxRIY2rQrDqYXINYRn%nZ`StIjgxYT*u3Q_91UMbASF+j6
zV5*yWetAp|jki>39``Kg*Xk^D!?@+O%6*-q{dGAVPXW|?scF-^8z#!x^=Bi)@0yF=
zZpXWG<b=9HcD|9An#SD@Ya=J}YSJ{gde+8L0wbE=dF%7_`x0>rS2*IzSM-n0oo+h0
zpJBH-&RU-q2B_~W%j$R@8g-kNy_}6;IB*~|B4e5sIPwE*O)Oh+bI0fnQE)%W$hqiX
zM&YksBiU3%81H~dC4m5q+lQePw5+Tlun+@B<sba-#G0AYM6~%dPe1;L@$oHOI;;Cp
zQUf$w{N3zTIj?8iB3QTMD$Dkw^xSLLcpWV#q_*=pdgi0Hr#ZuMt<L&MuqXdaqFS_{
zNmJqLgNUj>87UeT(W8VP76%Z5Fjj2{e|tG`o!uYs9|21#<1oov*nD+8-U`V*IJVrO
zx4HUVA}GCzE|pgWg0N;UkOB<WpuNf$aDLD}jHfUfnR-nt5iFm|My;jumP*2zuSfA4
z$e?ZfHVB{{!6xOx0dwgrFmUsgHcXENzREMz=B{KO4G_hcZuU0-nHrSVSPeG4-Int~
z-P!0+I(o-kRTHMFR-M5_?tEL72UaJay9{^Ob%829B07wBjHiKieM$wUCG9ICzQ?M$
zMnfMg7WXGab4c9yMV-c48?Bf`?@NS+hU>FLJ|gi)a747zqr$WF^PVjq7p;Cq-|A91
z<C9Na81L#C&e)^Iif;QCXQwtgX=p^@`T<}7(_6u`BQ~DhA6QPV)jErorWvW_;@W`O
z^({gmmY|XiAOp|+_+3GkP4jC=%xx`ZWMZNeSQ}5M;~dlx{q2!a>oX3|Vn)5n;ZNPa
zG_Zr72lnstec!Tpy4Cp7sb?{!)4{ymnl7?q?(G$9U{%4fsJq!gzSL_b4Y<%NaZAhM
zUAC^S4+7z(vEnXw0g!+iXHy4eR!3_Xeb~2tWYZTiu$uK!q3UKuZJK#BDjio3;60&#
zGQ7^uFh)qTxxd3UB693tM;d-%{v7=e1#oq0k=(&7?Em2cggnIy`N(KS>!)*uktudj
z$t|w!m@@A!&h?O#L7PcUDVlv(G*feqo~F2lVH-+yrv#+*J35ek8xY717acS?ZJt1{
zyyKvhbls`+DG}+)RSp0RrlQPFHm2p=H-4Zml=O(|JvMGjRNPbL1>KMN_L?OYQXQu0
z0F8Oc{K1g$@U)Y)UDsu07Q(iBE_`uUavN`be3Frs1;-(w-yF>SJuaTh4fJ#fo?oXv
z3siyhik%NRl3eZ7T}&qD;rtA>%QoLUHd1>P<Sp3b59;N(Y_|`fNj%UfQuo_SbgpF1
z&k$oP+v4?3481y>deN?Q?tl8`-U|7X$^wbt1_;Xni2X=u!wZ6p)YVPeYL{Xc-iV~5
zdYWzkJN3+F&sGg=RWA<Ob4W!c(QDPxmQ2ArVKP<WNCY1oPn^rbL=Y$y=kHb+kX}b;
zs@}nc?s5arXPde8L~4F*aS+oxzsc{p%DDh4hm<|WYM-zWfCkIuF@DQ<sl3W4^ea25
z>$iQmso5{<cDThGv#>ZV%<`Sr#s|HU6T_a*W2(7J^zmzTaZFpG2HM6{8`B`k2VirO
zmJUh<U@|65(_2kpU2?XWCg@!4ekGTkJQxHfyu~m{BIH4_#o*`bcX%2+@*vT;{yw?W
z$PS_xC_lj{@A>WiMQ-Lv>o1VnLYe^$1z_AYTrmG8kOzk9p7I2r7qhbScqp6MHQQ~w
z1b4&UpC5{;=PI-jKV+GKZo9UZBPC%DJh2QC!SDw-e})plGrFFKqAr#cIDt}9QbrMN
z4q~0!;YNVVwRt1&Ssm8pkaZUS<|R)xB1jXAHl$zh%SBf<U5WAG!!VJ=K!&!$6>BcT
zB-TPh2MtX1mj!J-0b7~M&z?~M7h98&i79_~Q=ICa)mVixSx{_l3^~QV#rU(s{o~_!
z>Wj8^Me3e&g}z2YWMiqxE79yCd!12tFK+#558wL27;D&d>eTA}_s1PkcfHt+-@fQy
zoY}=|qA{pZXPvPRQHbN+u^r;Hcvp4anQ3brEVI^Em06`f^GN-^#8w#0OQ^GGYikQp
z&DYV+AEd;;)-qCR6#?^C9;}a+EA2V+VT?esQC?W6C}91^TT5B_{?mImxk<0L0r+DJ
zTL|(kNQY`Cm?FZ$`j_pr18T?CN={2R1EdZAvL+sLE!!{s{-|3!Ue`L*rdDOj?1O&I
zWid74XGcrp`aZo3X3gQ;ZJ0&0(3|SJHL6B1V<@>&rNactAXjYXTdSjsuTT&y0xS(v
zsGluwizMQ>?XQU@_v4-zT^#2;mr!=Z96gVFoigytBFU4%;AUkW0Q-jOV(EQ18R@B3
zY-7Tug}3KAK&$GFVv<vFB*=_ayE%nF)Xno78?Qu{cxO5^_M_5j*1PZW_a<6{A=O*j
zr6%sB3vQ|kO2Yd?8828uUTD>>7_gXEd`HuQ5T%~?+w{0{_Vi~rlX$Wdi?IMyO0?ff
z)yoUHFI2>;&YA0Zil?`kYN_0Qf1TFR1P~oSpR8pmv?ueSE$l|NFeMr|U@#@x26E*w
zOJpNMOA(j#9~5LSE1<PGR_O$g<?FWDEIR_)J@rr{V&6vFR)VNo*YVJT5w^oslUT?X
zSqz%69M{J?KHQ}ITs{Vs2U1SzOH;SexngAzs1@jcFvu4k+gFn=GUR<Q;tq72Z*X&S
zL!^yRW$P!O!}fz0g;o;+(Bt^Cr6uv}*TAVMec(9_Rl6b6XAj&3;PmU{M?jeJHhK(T
zDx8huAfw3lWUfG|li7GL7gV3<>QKvD7%~=g!hRfXK~R1f3W9Q@93Xd^xosmmFq~g&
ztY6A#YTmj3JyvNkzQ>PgdoReCWoz%-9fl(Z!nVBHaQruqG5ppJPHvC?KEk|DmS33v
z6g^JP&X<^&JI`hE9Gm#(KroBZODb=-qrJ55MSV>(Zk*Wcz3$7WUvFb$S&qGKr@Bp^
zqxRz<+S2^ja&9iMsFBfqD0#yGdJyG~k}m{~B`rOD{h~cx$D<j8cF|xXVu(LRR)Jmv
zFb>J+@5dSBHz|<x{I#_TF#P8d-;?Q2HGG5x+FPISYA<Il0p=7Co8ZyV($FO2<X{2B
zHn`jf0sG%F>ZJ>4smWgkms3F<6{rji9myX-?<@s6=_A_0taD(+r62U!Sn0&WEY8<I
zC}h(^V3&HH9nSOU$}!+MCLKy5xjDd=eim3gMJ95{?}7P)%e!Rzht$i1G5pA*N)&&q
zYk5a5s^#ayc+XoMOQSrmR#S~oG0}0?zDIeW6aA{VA5A^cAa%QR%b@46E4z&9-hSP!
z2%M|$57{{7h>MnY?IomUH`h6640uYSq>40bW|NFskHl4C!xc;4n^jj9Eq}wPa#;<l
zs<yrx@<94Iw4Y(D8+f#^$Ig&Q22(*~TF5XMCHjfoVB^poGUye@O4Df%l8>l=vXr8C
zUGvjIGVhCs()(Iqiz5E&cacdzuOY<)A}qyc&l-?T2=a}vG_8hveW_r~9fmvg3;tMh
znd(X%FBouFG_0u+FlwnSm;HDSY8vrdD<Ty5P2|`6Xzxc_+D#buwlZCPTmPcgh>svP
z`~myeJi;#;yA{TD;`ze$&K3VCrY65XEe?V=dG19g1yil9I+5MT|F~{?FclW@^@Xae
zmT{(PpM9src<A^xq2k<T@bU+%*cx}$>kk=>1w8k)A>ILn2WRL~LR20+Yiq{`mo}Z3
zem$Rd_5<lN^_1t2KGaXjM=}3i-D4dN+6)l?2YaB1xdkKbV2lm#oPRn)8{mGQe0{^p
z@1S2Nm6sfc*eg*!<MC>U+(0H)(`RLUr(&3InK~mIZp7ob@)JN`(}A4GUdNNrESIrT
z={wDFi+HeP<levaoI;hYx69`@Cb8?wTo>Q7aYmb2$X%cqN|-OnekZ?Ti^p{jr>fM;
zYL6!?gYRBj??Zm<5=s4ETVbtVWOJKtU)x-lVw{Q@9M|8Ii=Lii^7zUdL48I0=PP0*
z8}v6R#Mo8a?dLWf?1O>*Sz(3RKx~IoPhIsZq(jJzEvk$`k)d*a*A)g5+{66{%25Rx
zxk*Z6#1#KYmAE3>L_Iw{V+TVq-q!;JNP4}Q7~{j|@c4MI*1bFQ5!qHv`jCcdxh|%r
zFXZ61eCHj+zMSCGD9r1IySgG#>|rJVo0=s*fGznILYNr*gQ;F&z?t`bQq&2(#PXwh
z_H3-KBb@=RwVAO+lgEzwXhB74y28{Z%Wg=1;H&D-bStiR5Gbj0q^XF`i~M?RfY<2;
z9+TDZweMf@Vx4Wr<WO4uHQ1Ur!T=K$K{u)ew}pS}R=<XErj|*^XhKL(C_XlpfjU+G
zU^bf*WZuyo1f`ZC)~^5V>COab6xalIE-MToLU16J9JL_pdh%;N2kTzvZC{uTz6f~m
zOnB^Tk50qH5F;o!^BpZD+iXiC3c0}kjQd)<wPg$Rr!LdyFsKv&QYgh2x$RrR370!0
zjZ|8EmS3KtlD*7+V_BMbpdfZ@{q~LUuUu|4{81@(XVJckF;UXS!Pwrnd7LO1+FGkp
zzC?PsZRST&JebRJRDU6$*4lfgGOXY$gPnG><ChYWikdwFK~kCrjYJRcN^=lheQL-d
zSA$Wh)5Wiwn;6+BUFu@Bec;qn5<NbAL=UWQfB+v>@i<uuU)fMor%LaWkd^I}@KQ0D
z{TxiC_lPKTB?Gk1K|!T&QP%vtuI1)GJ}Blc_%V>|eB^#$9nPrT3gubyfjsRtLS6Tp
zrDHJM`6*;YKVDzvJKgE40?hIRp4P>r$WTBEkl|s*oiQyi%Q1(KiS&9%WMs=yfA;hq
zHMz;-?!H8@prBhPFE1|ytdQ6uF_R)C{#P}5Mvf-KJ^8|#Zn)V^I&zQ0@T_R0TW&^1
zK4o0$LQD7UrK}kAUmX4@86Q7eX%|+|PDM{r@=0CSEB%s$k1U~aZa5Sa=lm`5o&!-e
z_r4LsSb6j&y+F20oWGno3Ez%sa8ukhbQDZiHiid+U?I?r01F1q63K;yxY_;Jo}6<8
z%<LbBrIBMt7TGfh#53hdyy-B;+xG9g@NWs1H6<az+s_Znb!Xufl>}(X0RMtua54v^
z6#KJPSuIDo0Okfn;v*h49p09ax)`*mDP}?s9tLMOGMVUjy*9tkl|DU+p>1WNHhyYq
zswjyeuOOp;>8;AN|LThyfjS<$%AezC`v&6PW*MKgue6~Tz_#bU^*)+ftDEH_a4Y(5
zbI=($R+E`fe7_U#C8Hyge7Pj$9?rEQvsSgUk~QiFo+p=mWFvYvJM#kaiguQiMGgu!
zDArfo6AdI;Ug|`3MYtTJ4&1HK@Tg@CA#keNoQuXF7Ob6+gR<_A=a3$39G_l#^Vm?m
zY1J_#I@+z*$N>fS1PJ_d5RvAIZ51#ePHw`gdHe3)cqQg8G9YDdWvF#sAU@z0y^8gZ
z%Y-xr-~`xUE9A|j!4`x-1;)tsn6mAx426K!$X0;Y$s3FOYK}D`s8a!B=1x)l&?3+~
zMM3@PBj*XWLgq7Q&q>o+CSl*<vUa>8Ug@-sAp7<*0XJ}2a|l|^3q?2`R{a?2TSu!U
z_qj1P-q$k>6=<ppaTr5t&2`?<IP;9>w}KAvZSOTcf8G}wy)aTmi?Xts`0=fm<>=Ju
z=EEEUS2l@xAO9YQcq|h3#=HgnsyA1z8!ivzjyg?vn6gDp>qxzLJRnc=1Ufkm4-aSN
z=i`4Z=k=RIK?-T8Km^o%8-+m}Q%fSYQD0&1-f3J7-$5?VSvs^T0@Bf^Kl3vPd<6>&
z%fd)0jPP6NzN=Acd8agY=*5+!365y#2LL5CCyDxgE?)<@5NOr+@855J`Lz1Q5DO4Y
z@Hii|vHW<xks|dIVC&-Ios>Nr4uJBgs}?P>3JSCPv`G&8<OG>L+lg!~qT~&Y4gc6y
z(q`s7fvx@dNuak~uVVRT%x9af>I!Xw5vevaBj51c-7eEsfZ*cSJLctH7NLGh7TPFe
zW94a8`NgOEiD=?ojhjf~rWd<z@;zG0)U2pGS%U2#*jrMUt<Cl^6d?gKI<x*_9}-kN
zluy$<uA$hZ4Q9e@YvxL=Ak3UC9T`iQZz6d}MNMtMofI9{Jgq<?1c=+L6>-vbFa-Xd
zwYm&f-QIn~`Wx%UTt=)5)Sedya&H5P%zD(|>BJxQncH*B22C5S^!cPt(N$G^$gHI@
za;O}E(r{AJh0>AON1VxdkX3k=&Wb>%2TUoB={1O1!&0QeO`PYBCAK-l*;^TFOXs}a
zY%{SM8WFBIYkyDTtKoq%e%F^z^Jlz~9zztQX9p~u08s+?#W=e^M}V1X$+YiP9d6xW
z!ga4rTzs^*=r&<h`K9xlXwPc_0G5t#_*{66gPd-5Rz@@P`X%S7W9-qvT29WmzMx%E
z^m4ZBTgb_~X*Gk*`Ss-phh_QhKp$=uS&J!Ka8S@=;4hnj7se%RtYCEIBZI}}h<?b(
z^@*SGK!%h4uo{P$3ZXf0VTuR8g@&oLXiNQ*gG~t0p|t?&+oUik)lw5GWI2h7s;<>>
zfQ=NvFNxx`(plz!xrV%@qqb6`zcUn)fsqdJDm-hQd_w53mrWN*?@R1Amn)O??V3-{
z$yq5IISJ>Em5q_5HZJ0xOe-xQy+Z*xA00jY=ApafZEFbq;Sk7GT-2b#om^^TYMyf}
znEdt;sX{WEWoEppc9J%>GP<S8bn2ZwQmPQwSESA(motJ>8gD7n`&&g(EV5ZDsztnX
z=b6q#VeIBu*W?<L%645?guJ}6zi9ewo&bMO#xuieE<sE3TkJpTnj{P&kGDBo+ci<&
z14Q3H64>QK>CUgC|EW1%o1Tld5?GbFu5V=PSXn*L4wPXSSP8WuK)7hrd82@@`}`3=
zkYKeIEap@&(&K;r4fN}Xw_uX(22g(+{YCHHq>@mS90mcD78vV6p9?eI$RQ8Q-@gzH
z(YzthUz@e2K~hn_XlTllN}|rZ1k#4rprHXNa~lg-HDJcl(a|AA(fRE=td)=yoa1cM
zK~_aWN2h4{?si`gnj+Xw2nWBfI^tb_kKrQ>Hq>1pp2cMD&yP2uVPOi2io%dj7sfw-
zlwo)&J3D*RnGaTJwwfOD4J8$Vno)jto+Tbsr-aO7Z9Dg54Vl4W{ED>HhS+$SbP?gC
z-?MYkuP-8OVBAL*LgZ(KJj3BSs(5959M-6jwpN#uO6(a`<-T<F;+5^rZP&mT-N~xi
z<$Gf{4%40so}O2JAyC2WcxZe!T6?3BkuyN+SA5AaW|)B-j8p}PpV?-rp^t#HQE~z7
zlT5m7ECj6T!q)iL-@}npBw<#|9l&EeAvx}yc3ysjkskM-XKV{84l87;c{NAo8pWos
zy5<eWt6Z>db4Zu>6pgGRg9Cu+#NxVRiU^AWAEInN&TYwqPa+u9<Z<ohJ70F&8K_f!
z7&kpW^}O)Pd+TD4RbPq5L!j#Wi}fX3mi5ZM7jzY3RsaV8zl4r@Nr@y#9Sm9YN;GcU
z<B_-FDyyiFl}2l<RCBpT11fakl~A3l;8UoQN=Nb$x2(<-WpdI_t4nZ5j*<oMyk}L)
z6AIMXp3vK#i4HB$baR-FU-q2YuD&d}yw;=5!#tnf0bDMF)mOq)LiZ&Qbas&O6vx}n
z{B`c{*ZG=he#yV4CqJ%xrSRY+ytKvT^nzKLYol)@gPl(*P1BZ;VV96s?w%$6F@&Tv
zI3xstTrCgdAeWrpq-bN&$R)_`*GMlZgh&B)@ZkvV^4(bC|B5ewg2Q*RQI8aOLCB)F
zudjSA%AyRKSf{_a0WJxnm<6DJ^&UHW7|g-#h5xSHVI2@%HUw_({kB^O=YAXN?lkyy
z%e!=uUnlB$p6=diBWnPB7`AVy;~LOD!7>k8#0OKcmzfD^v)}@j_CO{o0mi3uda92M
zyr88`75Su1-1aqU=<6~CQUbVKe|G8n2ieubkthfJJ3T?N7gYR6WdjBYaX$Z~B2F?g
zFqfD*A?P@$crJKH=a8v%Z>ckHYhfxURx!jZBBo)<1HEjzitNPW4=U;x+w%`US@zpw
zT4rdKNOpMGK>SH%JRdKt<Xyd<7Gp#&<0xQjm6NY~OdHnsr*~aNz3tz{)C6iG?zB_5
zkU{Y3$+mQS^N=}=N<xMhIgASo>@pHCG+GkfHo%UYhz3>;BzQ|z9za__zd^?5V}ieG
zSay<1s4l1%%JOYcq488VfA;j|AdLYFjh*+LuH9vW$|o21bX!}wtvriUtAOTbbbL{k
za8=jyjX+2A)>sGUh3CNQTN6CzJ9E_YiaWHy3&_)!t2E@jqqO246&)a!T9?u<zj`B~
z-?YqYfB_lZHOwLg1wa@Um-=hx2_IQNZ<=O$AE>!5g|cVKKo=FnSuh*@j~L$W!N}@e
ziQa)LcV?U<+aDtx7MBux8YDNNLju;SB-{-{0UFS<2DR>k1V?LSp|<DYECe=FO~b<>
zG>+v%x~tEcsGhHN-$h9u9$<blAD_-M%OtYs+^j1N-}pVei(1JYGHDJq9IigOLOHIB
z_Fa&%21V>_zv}w#Tbur@e?5|iR%?jT;3f`}|DEI#&0%E>T72nzT)m6-Y%hmHVU`(y
zv!bAfjS)bMbp>#bXHn@OckzNl^UWJd9J`=SHU)iZ1h^7T2H>T1M8?KubNyS1lL6Rd
zAn8IA5_lxg51$_QYpy^ivm)zh0gcQB>?<m>>n7qV^hPI%m`672SynEb$JQ{!Xx?$M
zr{p$`eFZQpz2Il&d6pCJgt+{4lJ(qN?nC;tp?oIbNhk1DELH!kKVsWLox6ITsll?F
z)$~W@3z|>HjgH}}y@&mrXx$6gd`I0PTC6XA(^d`;g-ND{cK4h7n1ITD^Yo8--*23p
zL|8~WI>r>Jx0;tk2~h}fkqWHF9-b}yTMida{_x~pXn44yj?US8vv%u30L-pGWvj%s
z7_<DVrwt!3V|bF&f0W7L*-MpH%^c6)_khkowbZJ<NbsfpkG(s7y!)g^d`?GHig6yq
za*rCY+XNrf>)*X_U>>D}bkWp!9HG7%kNUqgj0Du8w_1Xvh>5%%=t1vwlAFk{6~U%G
zIozHmkQ4}OH!e~XR-Jg!6_i_PfXyxI$UK`&3|-%hggW`93Uk=k9R4Ls;w(MHxeBFq
zVJ=c=)&PA>An6LG4*5^;s)vK;kyKaC;T8hk5~O($QT6pQdTjm5tUB3ht12|32eON5
zQ-rqHL;ksk-*ftc#tLFtbZolpJ7f1bj$;`gMV4D`+;262HbXMT(rC+AURJ|Nw*zx%
zxeE=qTjX623Xf_3dE3fxCPN`5Jw1NQqR+ep2?$F^ZvR5TKyWC6U-c1|toO$=IwOI4
z%Jil)Y!(0TNZci{fAN-Inr3F1!H1AfQ~%a5{b0?NQ;aP*rW4H>fPwd*J6j#^$b{`;
zlEs~uF>K*yR*c3XddzayYjBi1WRCYTIqiN#OVOxis_peWwK-;ql%81|Y36qtZ&@MK
zz4ZJ~BO$9>XH1_W4|L4~Q3jx%D<=BjW-LoJX^BAt5QDg;>4609l{b%Rr?b22x*;o;
zAwf>&<BZZig{Av3+|V;Kb#}aFu)({&DXQ~|B~G91>-IN{462WQ^K)CH=N!{{wCS3>
zdiY_tZjFBN(9S(}-}52daOPVXtBjb}b)F1Ge+oGjq>_ldFC?V5@ktvCT>&A6Y~g;k
z>Y#(atO^6~{j+gdrl5bVg3i654=l{u2d5P-b(cW@yPn+3Pl{D@dlGvvNg-2S3K#A_
z@c3wdE>%Wz@qD6DZI@?w_~mGKCjYkJK_JvyLt<lNzse=|h#MO}NKH+Jn1tQ6B>}oa
zBo}sEPbF5{0PNg!hTWb4U9K@}_CVSI(KvSja_BNj$(P8u{V#$l&y=*gLEik~w&kBf
z`n=~$Z!y&3JwJ|jL|JQgS&ll(qAU7~A(k?2k1Jx(bG*s-cB%Yo0OPfourP;qkE~~O
zQ5-Kn^_Uu-pPf?Qp=#8ypoXncpeE~@Crg3zPx$o8RNjOovhz171!zY=z&F4B5w*zn
zFE(*?_g<^VnHmgBlZs$!p5EN>lH+d_tkG=a*UcBr{d5J3F`hhoNO}U3Y04^qPZ+rB
zII-!DP@DyhUUJXxKi(tHAJGgYHPH-W{-s*KX21Ui3Q0iUYtnpS!A1f~C2%U_Y3y4E
z>u~>L1^7Q=^^%7`SPtW-bGP>je5|9{%xv3VB4ZtObmDMBV?Jo)Hd&U2#Ker6&RQnQ
zUj;fm1x|<OqCF2l!|Jy_KoSJVut2<A@reL-XPlGH-xP7;5<FE0<-Z)5=5{K0H^ioL
zK%4_WRY4)B`xf&zfoCACtCC&PBRTf@x*ybbjf={A23gIqAV`7{lp3D#5B`ZCE+uxJ
zc-FFTir_3wbtj5|(}$~5w<wSL+iPU-hzQh)VAX9l{yc^OmM|eeJsj8)z#kOjjvu~2
z8Pf7A!68UC{iygP+~}}4Cvw;A{=Y!rl61QuE=k>@Jp?o&uw8*@-8{w-PWy?S3MqO5
z2X1VJgpLSnYS-1v?v;YH_Qe!vrR!V7N9&cq`M0X!@St`1?YnDN-ay?4?~1sPR_Qmj
zH?M^#x)WTUz7j$KK-X?}lzUl>xgdN3hQdF8q-H^zFtOe|YkRk?kfxp&<G*ea)+EO7
zU8u>DJx^+-md6Dupdt<~D}T-e26&{j$t-Dd1q%@toTMrpt5F9ie~+v}*nU?+F}axc
zobo?a76;sqPr&M!9I?GEilLGKX+BtZ2uBW`Vr2&ta{R@TfxSSRYST>FR)|XkDFj87
zKMRS!iLz?NyI%7WfX5+TRkS6d&&2c*hF3W%%j;7bT~wP(f9nkL)H?a9?G`w|R<wE5
z)$_PJvy2-Wgbv(m48p^Ggzx@6SGu%JtcYw!6Kl<91q1l2<Tr5;P`mE{p`qbtl7Bms
zA7|hx=<GKqC}vJwh)aHx$Ql$2lwC*+A=cIMMlh@Iw2o^JLNKazycDtTbQ%{JV;3+n
z=-+qnFJTP*516|IP<HJN^OA~zKh%3&Ma<K<^c}tqMcaq&@J*?Hvf2ImkBFwCcu#cT
z>L@g1Ga%d^dSNo_wx{{vKil95d<8a%f7(EA>2E<V-$-CA0@n`sHb`{I3}^Mn^Mm^K
za+}IAo7_DE3vmqp<$z7IzfsL(3S>J4LM!U`9F?Fu8Ay&8Kf0}$Q9R-DG;d*<p;kC_
zuSCNi<cIPH&-La;;9J41!9ju)7Nmn`ZEs$^^h9c1N$8TeE9)dN3PpZJ;>qJ@NKJ<#
zCh#X&tUz<yS;O2^+1OfH<1;Tg{W08A&ss%Uf@MF#tC!$1y)q6tAR{gv!NwW<0An@&
z@j6}^gVXl^?$MaXz+3gfB;10+22lNhk8BJXxXo&yNFolj!2N+vlV;0OyFM9Cn;ua@
z+5o8Nh9TMv(2_;w{09x{C#Vo=IT*RPG3CqD<WaAsqtp2-QwjQK)MIO12+$O#1F0l1
zyg&C}ntZY@!61U)mT^M%vrw7>=YvHV>is+c9Dgc_Bmg|tHYbWGN1U2Tm`RConxoiZ
ziEdJ`v;1R$>xkK4L@d}ptE$Z_5WjQV&21n<AirO{K{LV!Lx-UW0#RL=Sf7U!xZ(Mu
z?sPeFC|^KMk7Q?1K@>PUFcAApit)tiE9AuL)voLjTDBhx1=YiswHc`|q*Q!_k;4|R
zJPK#!Ya>v!1o(%ip09r-^a|@F_aqE1TdY1w?^_IQ?_GBp4+4beY!0F)0^j?SB_B#O
zZ(Ml;6=i0^%$uj4g6e-uZsFU<r|!qAr3~5?_|TpxmaGRkl5{MO@}UX{1tQq#HY83k
zxCV|d&<KM{$6b8|JdPk;0!b=ZevMq&GBYw}{Yz5e40_+a(Xh|L)fsTy)K=1V)b3lS
zPp6%hl2Y8Er-7sojxpAidK;@9kYC&yoiOmW-_fEh8gd_Hkpu?8Bcg`+ZBU4mn~prI
zIXx5jdwCnT6?Akyw-|oH)Y9h^Ij^mD3;0}Ean64hDZvsFN%2G|P^iHIEmcb+0sYc#
z#^wsHYnMZXu3r7r`r79vgEBYS6Uu9X3K=>0VFdJo;=(!46~aGUy;_(5#O<zz6$O!{
zPtw&HVV2VQlOOG0TBw5Cv{+h>O9mK%Pgh*_jOO(CCss~ZR&N{8Ffyt@-hCl$)M2q#
z6jGPakyRCU_bQNZNQO8eH7(72vd#z12vFyaFU%rP{H=ggn&eqmc=6*GyE5|G=Og%*
ztp*Q;ww(FU9tQ<5(Ljvtur`?0=5d@B%z^$+?_i%JnDI<7Ff`P@<ua$c(hcB4VFU`(
z9I0}nP!FF;N`8Wx#l_xnuXB$v*X7(Zp+(rc70^9@Ve7#1Er3Iyj?e!5I|}3-`!!^t
z7G}f|y$jUsr+0YlGuS7jn5t7aC)`<SPN6-|!BR<Se&pg$!kuUF*9bBDLZu&sNIy^t
zxgHoJn@)QDJ^bG*@$%x-4!3SCC)MvA|Ij0`_^A4&m)GBmKx_i{z!KeI4($@($)9-o
zSruHFDAH8s>=CWMNr8syz67NxxQfi;;%CrJip8aJDpq(mnHv!8UoZiu`1z058ru+-
zClo?T>2*0*?a%G7jF)I{)vkIk)xF^0%tW=UI<_s^A6-n&n&`>;VM=o}Q5qcIHv4JO
z`}_3fLBvux70^m5$8^>cx(F$N$kAZ$#diOaI12S7(kjRYo!njC<7fwi%Mti_3-FN8
z%;0Ju%-oIAH-zS1U<ZL9%oC9P&tmR^nV0#1`wLRM1S}Yh64f&|a;e0;TS-CTZgfW)
z;Nr$5=-mO}P;D#66mE+`Y<}kbvs7HDzXJFt=O*||M~4F2YIm5Ku|xm{p<`l-&?EX3
z5TKf^l1swp(O0^E%k7@ebu?TQN&(to(vy<z&CbmHBHFND9ah!_X^HQ2%1=L&D&*@_
zFL_Y18UDn1sP<I|+E{gCC;9?!t&=1#;pZ&*-u+dKk<~UEtVaYw77rhA1$#P+Jyhgu
z5G24k+@sm@5S#RrZR0#`pS1pIS^B;tN8XN$6m~?&$m$-`*qu-JD{@t`8MVqN-q-gq
z>^FU4yFi6PCCvbgF<fjG0#K6DvTj{SNJxLFmF_KkeABwO7(8BQRXQ%Sw>lzO)5m8c
zr5`|(iwxXU*0!5)ITy4rZ=nm7JoM3?i1wIZD+<Ram3W28{x*4)F{17KB(QH0>r1Dq
z2UEH{gSHNN&3X_M*Jq!bc%>t`P<TDJuV`p^GPJrE=;2s{$}B7lM$uln8i+!D2<{R*
z*MbruiJ+%@cL7xG&p-4+YE|O8Yo3~(o(;tCC;^X(3aDZwu;f?oX_UYg>^xk?qd2tG
znx2}1(jIk4%gp)E{??fMv5vpT10Pi8X4l=rIC9%^E%<Ym=_XKf^WMZxlQ{Ylb_sFu
z4Rk99?QRN^tbEzQ-;4Goo7z&+MVQ#GH&ES4#P67Ody`BDlQT(J3_pN5rj<DJZY~X0
z>yKc{<q2p8!YyU#$iskUC4=wE4&>=aSEOrbbs-QOV)!mk$QU67+^EOdKg!1}&*v4C
zyp?zD_Nt>u;1R$y#`6z`N;AiPDdb{fb=;@jP7e=i@GJ)N`uA2xmbV7RAB1Gmytw82
z*bh#Vw+7gg`q~O@8p}}Z`h%X^+ehO<o}lc><9_hf61td6pbpcs>wAk@F7DUF`I!?u
z9hpI0FOTZ+<>)AtSSF%RK0WAwOG5|3Q-BitLDnT%JC;OwY?~H0*%D4OM6~fdPHQ<}
z(igxXVOh9z5F;BlqaVa^leB%vccJGL2HA=O9;n<E#}A+Wa0rd=J4@O5=AfBE!fB~h
zoQ-D#w+s$I0mzt|hQ}7RGe2+Usuc)BLuH~IN#>$|<OLK8B@~JS_k==u6C%IigE|1*
z4ADUx4mI!3G<)mt`$hG;Qq^!j`cP<m3g4~8K%x4-e!AmL@L72!OYaiAM4_sXoq|Gr
z_eK5l()kJoEY>EL;tDJ<@D+2%p&ERjn%V+t!6cw;eeMi#I&e(WfabFV5*I-o#=iP5
znL1`|X*oGn*Q@vAu<GT3rIIEcdC#|TN^m@IPv96DyBQdA3fe(k0lC9TD=q$81O#+~
zg1Ssq8?Oad6;~soRLR8&IaS6B4Bijq>*fas<6;A>3K|L9Or1ABt-pmCo)Rosq1=cD
zkq-F};|nP7w1!81hgbA_Ax^P6*$nVv1N*0-qGAlv5X9sjJ3)Fx?;!Q|3!_&hfn#(4
zW5fJkzkW?KxYr9BLfq33^>Bb;hQU&L4C;jqpn6nr?kqAA9UJ=+h=G*79cEwQbWL~i
zuCo;u7GCJ}$~k@4Q+4#+1OkhuDwl0P_*W>V_Bzi+n)t!kf%8oq4{BG@<MYY1ceIWu
z6vbSOpqCD>)7mp*<1{FK8iGH8x%0DgbCMMY5l%%z3`IjZWo2AIsH7P@Xa-={9LNPx
zEJmNErl*ZcG>|v14i!BGJLqqjKSat~>Q1kcBM&U6i@l8fo|!JYuc+WOR0E0;#p{#@
z>=oeO!(&eZx1^*vo&QR0IY3-L(H@1uu><->f2I=k`3|eZ5-`(%QbhwN=itD3cd_p&
zG}KGV$n@9wU?!^$@gV$ExYV#OUl!zwG;)I(*R1+Rhc%;NC}Nu-iy1_SM<5QgE?4{P
z7;Hq84EaHffQJTIz*8u5@1#4djjBVTJ2+~Bz5DOp^!{SE^ov&8p<fxz3?RN>#?MU<
zuVE8C`tHR#^eZ(bg+@?NkYW}5#vnAq(Pj4!FYJMS?lVZcFeSa97bmJ+Tv$}p$8*1O
zf|i>4DYQRea;Ze@NqssL$XME0?5lt#_+1!yVh-KibhNZRz#t0^3QBBfXbAM69<|(`
zSSy>pe$FcX@#DvDaE{0BG^Me-AkK~AL1Q!ecoWe(1*y*8Q`TIFKk4mN?1<s+iy!~8
z2O58>(9@ilrLTqnMN!NKV!6I%Y<t_EeFQSHfI+D1zUkw+>8ncvSM^TM6bPlRJ3UWN
zU+H|P!4ot|{w*ZKq;&;_DtilUhr$r+Tob1Uch+KPzLnj7`UP}i%$xj(m&PhfL91<+
ze8}{LTlCab31#GK;diH_UA*$knGZkSNooDf66nPN7S{CnPByIBeJUy%etvBLN)_P(
zi_T@A;87y?Gxx6(EIB{Kli;zP`QW;{bU%ca!-*MmM!*9H!4TZnme$b*ZrMj53zJ!U
zx@twjYV1eVGQIip5tctOs~8BV=oeXe$pAl?N}Kdn)_Bv~w{HoCsDp$TuA)%-vbMH`
zM|-Q;(Dp(6E)ne4B8<3IfKJWg>;Vds&X}GwndqHgN^<=NgXePt<4X^uza)U12h(tw
zO)|)6WI$<r5v<$*bP<L{MGdcv=!QU(v)1X+ZXf(;*0B=)D9g(yOxC}rXJ#%C-IGYc
z(Yb)qV0+`QZxCKT*$vE!5|3j?P)d;n8b&^plL=SN*?-3IyA^}{<+-?!GDM1nLocu`
z*P&9JTUS>{(iIDx6E@RLJ}KCVJ79TD54UDDYTU~ObD86v81`P-+fe?7)BAF~N*%_O
zRIGj*VY8Wf;>9bx8Jkg~kdt!0KcH##5JS#a|03Af1W`(%3*B|Q#zF6dqzf)0rjrTr
z{wUPOJt4#*|94(~2n22b9X~(+aDO)-80;=vGv39!%Y$wD1wUtPeR5(Z&h|==`tv<+
z6>g&i;gGLhgYPfK2gmRrgm$<e2@9^h1-v~*)<nDwve3JN)Hs&N)<!*S2gN<2RXUUm
zfzk@RPFjGjcL1WIz&Ygu!fgipg(IM8U);jF!}APUA<KY<o(A0lK;BYW8n3Q|!UCav
zjhx`=GhlKAb{H!Gd=(ZQoef{B23!DaF5P}wNBE1fnAkfiT4rVi;8X`wWdf@miPK-V
zE=<(c&b3G41@3J&lUw}$st1d*G*pxh@&l<*Y)sgPB~w8hAjIK$WmDJqp&<&$V`}`a
zSXci(P;Z*Uojy4S;BuJ)JkR5@m4eu?<fk_vHOML}`yJo4Xm1X^uNnpVKFHPrhXIiw
z4J)e>$ZBrP(=#wU153FJT&L`u9BITaL2K=99VXKy>C9|_MoZ}YM!XwUW>u9IDA@(t
z+eki_>7pvvVnia0b+^401-RwiZjl>E+zlk%esFrl5FmN)$AZ0s*r}g=P%F9q{f*^@
zk9bG3OkSDb5wpOr2UTOdy@I%7A$+2Di9N|rA@PucGL|BwH(;TGER;E7RjWPS9B|`j
zd|+S#947Fuh;0Qm_@#l|=rd^EKJMZKehokHji6CEWaxzooII$%LP+M+as<J?x7Tvt
zXTXLRK2pbZLlGSQAsyC?9(*tuIy4s$vPuK3A7xN#3GX*o78ZUDtPGfFodO%!+w?B7
zRX@B&E*|z>)FVCN?gAJT06uvM;={9IAZMuO4tTpm(RdIVOJ@ZqV571D56GL(u~Hj2
zWY{8@j6A@g23llV<yr+j#h7`?66~k_b5Q(B>WJpFfJE-`eA#kdC6qE#K;-OeKVAks
za#Fl$p!F(&YO!KD50hu0#6nf#nlnKrL{g6r!AYJdjUfzb2}Q8;Z?<-U_}{ae+W}LO
zsU(0w`ncJ$V)4Ow7c^ysQTdZ{D?%3jXLnimTq77k4Zpcwa1iyr1P<>qBv!%r<m6Qu
z*5Er*eEXK%8QEp>z>g1X``B7hF1|m~UK&-7SK69AhL6*|ufMqg83_xP2xNCx&G=uv
zqxk0$fh!UZmW|=MY5xMYL$Em+c9uU=85hVWb{7N)9A(^94z=GTPF|`xJF+<MQq^_{
zgCO)n)={(I+2r}qU^p1mq#q7w_g*346~#5=@`3ZR2tmwsp1pLzQwhl8LL2onw+}&F
zn^9sI$lO9l(Ck_ak`Gt<nPIesDaIP*)6bV)oDBn!7Wpi=$!kBIChIQ094c%)cY_Dj
z4-;+A&0Um{T!(YiTI9{r)wd-7>_93iD#SBI33}D+t&P#wKbo;EFV=6o<^h##u!1ht
zoe`m-DWJBAScEf$D68626+|+)uhqWhW(FM4!#>1_$MQNIMf^E~(16YJ#2Ipp1V;w=
zWx!Gj*CVR1o_q}g9&6t2y^XUCK(<d~w;1>dK4sR{3`~1#XQv9xO{*Tpe2e+_qC!D@
zO9MVS5F#g#38@R5XmAQOieP@<#5KFD6f_q@hBD?dAB$vu5KF?cflI%+!KWm1<px>K
zh-IY=aIp5R@Oi;f!>Km~2?t33K};t&oJlwDCZ*V=60l8oVB+uR9F=Ss`;i2g0{><M
z9@?R0)utG%u{aRA6(Er7&(r38`TOf*gxLxUyga5uK`Pyo2PU(}kf_OB!Z%6#21SgR
zA5kLgK{{Su-eKj(eSIT_ECo_lw~%}S#dZ-|BOZiEKr(kteBYT5BHdrmYK)s2g;wB$
zta<{mxVU(C47VL*>KBk-9sEp1B@Ixl4A{mb952*DEv`0P`Fk{s5ot1wLc@<<m#YA3
zn3t50=mD$3SRR1u?faj|92equlfeH|{Qkd$^Z)-}(hCgVf}1M9syBVOcyy(P|9+?g
znm18j{Ua>>*L43pJAWghrmGU`qLpMQ@88Y%yzn^xZ~Wu`bf%vI9516#rsgqT2i1Ok
z>2*4v%T;w7&*OF9SI3Bt91A)v&t%@wV@el!f0u;MRQq?^#}t@9Ce2unw`V)&R4vM?
zc!a$3Xg%5TbfPQvChhX~4^PyiTONrF9m|HT+rO<Q4e3bP+g{aUqQ*y_Sw<5LH`DaW
z{v=aI72k=`lgFa7W?>QJ*lyfmx6<o%#8U3$@(~81G8?ps_gL=b{xVuvxJ)_R|AAY~
z=kEUg<T|Gvmd9+A#0n;I>;6v>ql7}~{PoSKKYwl1|Ndn|AZS5Y0Q7}8VTk&m2RIY@
zBxB+HD!6#mobGADoH$77eTs+`n89DFGlY}qLTA(d{B3}xK+ditVc+X;HUe-B$lrof
zT0*GNW79#*EQ;GM4O|K3K%RJp#l-GHk4QA9)pgf%;8k5bk_`rQqr`qu5&1bsw2a{K
zP**WX8-#K8&8{;!yhMC1Y}PI33lT1Yo+oAC#;_=zY1iglt6{jh62$dKXUBh53F6$h
zBzD)vcoX6o`)@!qPC2|0Xlt)OznrMiCh!%-BxKx4M=&iMN<?9F8IFB8#1?vbiLGsI
zCD7UdG%ds$9BfWww`wYTde(rBS2`qct&4u`Y0*Jbw&h5c+&XD}sT$ZKr-zWgLH}|0
zF?Rjc1-yv1)J9)I^qBOrGBub7LL%Vq{A3uP)?t-5qD^_)-+i}+{^3LA05X1PKIJ~&
zw*##w9=E+5#3CrGG<BzOJJ@Ghw9VQsf?duW$i*BzDRZ6ZpN}b^d-zZW(D4BPcrNzB
z=d%DbH}8mIXNmT@Y5;c%1^}Gy@VA$m-31XkuAtCjgEq*A*g9r?2wq$dCg1OZm}&+<
z2>_d!_oeyAxh~Nf85^5JHig}~Y(Vrb&Sj3c8ZsI4v5FFbgFkP?IhgKcjX(1d28HwQ
z>+tA+y9&=qujsbuHKPgG0E^#7ER`G`zcE&>z;`|<c{!p2a&$QmxQE8alNT07y{^#F
zgb(#WJ1ye(0%KWKs)1S4+SN4(T;&zHXh<}H)txh5?RMO^G<*Yspzcq%=r8m_aidtT
z{)$X24?9IH7%*=l>yrZ;Ft~CcxR$kfKpDW71lh|HJdNE}&3<V6P5Sid8a4@=I0PF5
zTeEE<<DbL7bMB_QZoKtHz6^}{=^SS*29GgIWU<q_&X~vkm_mm7!LQw6i0PDMWj_Oa
zk5g#Zloc9^|4^e?9O!@Gj6tI+8=Mjz{>-lua!3!-K?VVoy6w)2!5h(0Qi=c+g7H%d
zj+PJ>G#q*3;zC(6Mr(Q#;#-Me**HG+{3_*Kt#U;mT7HG>2{aEfATmK8m1aYSl$TxL
zsZw8P#k-80oL|7mWY?>^H1y(IU`$aaoDY98ejXY*m_dEoEEmIdcX0~aT1p|_f}WjS
z8Aj>R!H}gPZi9XR0U+$ig7XQ3*Ce1936@qyD6V2p1f0PR)F(_Ki$z<9c+s`n>C4|o
z`DfD^2IHCYuelxY-@Tit<Kh7)Zx@)KNaPx9UyseX055()9wn@B`oI8kTI4z%iqi#@
zTF6`%?t0<ikdx<w)2xQ^xvrayg!Kxi8Cg;j61_mqe!b-mK-Sqb68v1KN%lIZC%<qZ
z3!-|lG=-bpffmaUEM_=dDodRA6NLi;XRZ%%!9d(WIPPf8%Of|G(xf+aqhWu!0^Nbl
z`p@?Scf@Elcw6L^i^2Drbj9TYWUT8+ivS5y)mukZ$7_`UU{`?1611xWAndR;Bk@qZ
zkgDO~$GnP5f&dHrg7fma*W-s!zxSWR`S^WwG)Q3p)&N|x3V=Z3URTb`4%nm?Q5tHy
z)gV8SO11dAz~HPH`WKInH$k$=R0Z=o+sts`$SUIFL8tI0vRm}(-(@1Ykrno4-A>sx
zr`_4AOdi0J;<20aLtuSm2miStD)7$@K!l+I!O<ItqRsivpkieXU@c@NyRFNhYuH5t
zku#t~Udcxw-M^@NeiZSUjGrW&gIyUK-g@2Es|iDEyv{wxY}+29QIhvt{f58-68Bph
z&$*HC-2&<~tsNb?o6avMM%4PdE$AUbhoQz<HWg}8l{%}Al?cVM7oM<j5wPAmdO@}~
zEF>fuzPPk$3-LG~&>)D2=6I>7q7n`W*t+hOLqKHU<}#t*Z8e@pF}|v|@&1qf)sb}G
zwKB?tcF5S4fGs)zCX(`3YnmR2?|})5hGd8apw#+D#c;X&U`kwJP!ZGNSsP=ebg=oT
z-DN^R?`e5?HG$%NM(Z@+6+Z}lvPgs{R%(0q*B6Vn7aYqViHbQq@_u<Y>D`_Pz*l5$
zE6lJLr68QkX1?*)vL!)Ign$gt`?nhsfAR!u6g;WmM*W`jc+V`z^Vn(D0a~_)eRL(|
z5HPt#So#~oRV*s~-@1kRX}RJ1`XIST1tdHZQVLjH1V1TpSXKdBvJh-kY&oV8Tnx|n
zYY8REQm>PP0}v{geg;t@+KvFQGC)Bgi1I1j{rjRIv3)~iH95IpPzB)9et0s0aYf~(
zrh=o+)Bcc?b%VuGU6u?f5&VIH67qY5W$-rT54QsLjM_J~rWU+J9_`$zqf@5FiLje{
zw)b8hcLwp;OvQeU=v!<(&z-FQ_6}kP$%E&KT4D!QmgmbAn(Ryt)<}>M_Fg+m9Y4|}
zf^O|=%dqr2Vo6TkXGfnYS7yJgK5#h5|6^t=qriG-ieR(}oByp~Y${nBJ{1*J7QY*z
zU@HLMs8so`+RwBstBt6{Y>a<pR?Bfnt`0&GBL`xd0KK=>F`f|)*WHMN2D1It?ecuv
zb6;w)4ok32MseN8mii#{xS-FhKuHOOGQYXxuaAa$Nca~T_@B`4f9vw@6^%VYp(Y(3
z_fB4^;Z*OPy_<uHY6zYVChf_4qNM)`iPt_nsDCskrGdZd3PAM^N2IhU@^5wj5Ak6B
a`HgbIh&{1b!FVK~k`R@Bk|+H9_5T3jyGU;U

literal 73627
zcmd?RbySq^zcxCEf}&u7L5h@sfYJ>HEg{{YbazN23JMaE(k(HBbTcR*(l9U#-3<fM
zFu-u`hwtyV-@W%b`>b=$pYO|BtZ~L?p68BFT-Wva+<aD2kRm1`CxXFX#4^$^m0>V^
zX&CH+=4Ar#iKgW5UGPHiR!-_Acm+R4FY0{3%N2WREhiZ4mNWE=H)%@a4ud^_$-I1~
z>YlVd@lIFG1=qH9&~P`xyZ^>(QYRUmmTRTm-@D6OldNT18u}S4P^P0llHz&Id`!NZ
zs<4|&rkRguCZM{EvkQasOzu3W39<6?Ytr-a)vbR&O}48~(zzxOzEgeRg;?_0T*@at
z=oXmTe0l}eXU$I;R!47gPmd6cKW1e40yN09^al^{z^nMB3%8*6s{gOvek{8pea4hm
zfOIXEVz~gjxAAwTd3$HasZK&dg8Avw*Qq|(JW7(Feg@BpU^3RhJQb$yM84i-uwwf2
zPJzW`(zS|_cG;B(X^*)A)(bH1%L@w&w#NsXQ^jAtTr)8?Hb4A3qg`#E{)$mWL`~x?
zI5;>uF%f=p;*}|fXK7qD7E|J0@*S}o+1S&idIm4%t<7S2EP7{e+2!+ZvBR{tMn<9R
zg#6iD#L`zHFDP2BJ<JQpqB2vv2}|pjk(WpHrN405>E@%*BkFssRj5AkZVVQGu#Oy<
zsI!z>v){8F`@LM6|EOwY54HG8&-2J6A+=E&t?yab&w$4xIN-;<Pl|DqN$?0Xx&R~3
zot=Gq>EgxeeH-4y{f!B@KR%&lb0BeK)}^a=ZbZNW>IJ=Pe_fwn*-I?acp)0@>7bfb
zwAQEKdrwqX%-F1RYHDgZ($BKTm?SbatBw_;A!@|^+OqFV?*gpHZ$x*m?PKfP*RNiE
z!5#Jb){b&=ajn%ohLIr`d(pSI971R1G!N9-T(lCUy|F7fl5rOXmtS2#syX&Ni=?m1
z5>s8ZVp_^}h54DL6%=&s)q5ZB6R2O{w(~sL)M)$rB(pQOBJ7#}Xq{VezX^;-Fv^d+
z5xK)Hv&Mjd55_vHli|V0KTg!St+g|yI%C4`J^j^W3F8hdlWL2Xx4Z*4$Z*JCqs)Q{
zEZwJyd)Q99Vi1Yb@8AECWyU&+2a^|9Qc{ZPTvyPaDYBbtyhsldk5WSWIXBB38>kDj
zqdxnQ(ylz05lfQLgZUA(n3lDhmL->-`4dt{gezivvb^?3%sduiH1f+C7#J-6PXCNd
zf;(h?|Gt_TE~@PAUL8nG@m<s#TOGFND%Ph;#m0)>b)hJC@8rPpAjGMjzu-PRDe|a3
zU?AGt5t=h%u<(6dB6w!iNWaoOhFuG@?Wz`^Q$zdhif1p$(Z{F@PqE~QE1S3CF3W_c
z>RG`2e$O5t&KgpDaqk$Im{RwL$H#R=&Q5k}r{7<-Tu*&!6!@#TIVK??ABj{37l3Uv
zxN_yn9`ZW3L{K=DN^eb`ppx#xvi2;JblQ$nHUkc?qhwr6jZ9)&Sf7ADo{Q*ywVO|m
zdd&p8Ov1hmBi@xzZN0?@A7xiHED+m;Ettqf)lq~Z(;Y_`jN7rbwbhZIvZ&j(id9U^
zAY6F+)7m$gC`QyOW4?fZKyX-CQAu%HnmAmecnG+mNr@)uy?eR+nNqB5Y=!z*|L?1c
ztY?|{tqfSB-m5O=v-~y<ZqX!Socg!fSMuoEWO=qv!+VxB+XZrF7wouDoJ|I9uABLr
z`-@=B97KGdH#x1|im(;1@XgLDl^siE-Mb{S>f)LWj3YQI?ua-36%jEpDw<g(znlia
z6e)x_*$zP>k=$N~_Mucl>NMWFU#&{__c3E+R8(vT*|@(gEiI3PgmiclyVp)*1Sm;P
zv1`Re7~qrJdwYegsdxH*kjBKV1+44F%QMQad4^tKVGpu-Q?h(kv`J@Brgb*fsB+1~
z#6@twdc7H8VfHNj_ugu{CwrrGWchkr680?3FKBS#R^4YNOyYv5;RKt7UNRJ7$kpK}
z46ePSqeE3$IRluE`L7RGSROs<+vyc62A*LtnESe;vr{FAget$D@zEpcNL6j^)jeEO
zlek7*^Gsi|+oaaQaBl9Ss?_V|pTbhRYihlPDt@RoG_|!^t4}!zEjK3_ROI~4g(G;|
z0xK<<Wn@ohm}HQ8OQ|_d%}H{hil?Rx8yHkGJoW0|*wK_{!O??tVMgI<6HEMHEkFPI
z_52h#c@c6bGw7J(rH_w~cB5B~$NoBVRD0HsA5TR7HsO0LCA;aG*(>O;&&+50HB1<c
z*i=m{j(Y4ar_a7+r+7Z+bYf6FqPp{ZI#w3rBA^~=O%kN9hj=~1rk(2vkKH~fF&B?>
zs>n=J5|C2xZD;Gdf#PjC!DXG=%kZ4S5z5HX;3c0(_PFxJE?ImSH))Q}qE4i3?aFs9
z!DeK>8mmkX7ys9<U%@iLkp{ZDkHN_N{9wfg8m7F7duRj>RgwDjYdVO0gGW1fk|9Kn
zDfdsU+t_2qFx!Q@Q>7?dbX$!Oy-felpW9CpExilqO*T4JUbmjnd}!pBi_{I_sXv`l
zOc-BswDaL^Vzu-n$lETy9GAOjPcJhi&5~d4+Z#`M31+To&6^0upf}cD>x#)|kc&-n
zO>$WqQmd&cpvR0=DZY90hUoHTC2Q*fM@Pr}F2d0gqoCQ@S!2%l`bprS&l($RD>`dR
zg(y8ve63Wqw>A;uw&?j@gmQfM_qDfUtOXwplEY%MMko}<QQ70RZXfAnJX|IuM{4dr
zCQl!|m~~HPf)#^s%*SUKq3<pz7t>V^;mW-V>+_56Hm$Io&^|pr(6F@3_c_^;x}<R;
zIg%X0UMsDorG>>EZ!XyLJMS*PZopyJ$VI)XUol33NP>cEaDc4$XU6^!@bEITUdWLd
z+Sg9SRd|*b8-xGarK36dR+B7>S>O01Xv=%*arh?YDT#@()ff6Ehl`k2tYB!WI^J?J
zpY`5$UN0fk)#)Aj%JPVHAg~@^(%X3Y3u^IJonGYXwp9faD=$o(4pnL*SLNi?6OBM1
zOuG}Z!JaM0o29%qHqP3pU1OiHZ~lB7!NkPG%D^D;`SWK>RDR6wFe<q1L@i23tdF$f
z+Fx|<oTSw%dadFx$Dn*+`c3PHi)%$Yb*ufIw@D{6B$C7R8jNz*NrEKzs?$c_lgzQN
zq;J7-IqxWZmAn$y*7nLSn1!refu-Gd;HPwx>3C8b_WKbX-3#!r1VI<em}4<K;%CQ0
zg)!a-H=snd)TmV)Mu*FNYV`Y_9+8r~-bTPh7|0yvXn*{-mDi{5gA4=nV@#_KyrfE=
z|1ePxhr>z6;SGNB1DEyvAx7*dcvsuh^_vU^chB^-jk&FCcWParB)*Mz1%RNho~t!W
zySr5syAzY%l5w-@pHcaQVtQ~<zFoDYo@9U~a$nonkzR$Kn6Fc1+o6eT+@Ru7-k-HY
zjUqn2GRc;VTPQs+FX6?ArR{%RGGG3C(_=sNjJuY8I{GpIi69?7fBt^xtXiVZ6@&cQ
zi(^{z)wr*Ryz=a42SVhvhW@+k3l}bYDI%~BIUn!`C`#7K-PPhtHgaZW|K9iHNUe72
zeqY2>LNHz!j0Io|@RN1{8vg(8Zw<fMAx_mn;**Zyg`@I;tv+IAmYHh{M<?)Dtya*$
z`s$DF>4DXF$HLA&6vEM%>xvmwv-p2V(=hW%eqcZLRvY6r)B=mfTxLD0y(uE_axzj<
z^SxCR<eHD~gj~SK=keGz&(6+9$%(;W5QyOcd8;Cnf=@vmULmWqfcf*4u68NKbCnzS
zJCwqNGoIW1?=%YtLUtO?&gCa3C%QIIpFZvA>51(&-SGhT(0xk?%Vm7<KpY_6B5wl&
zgQ~WJHFWWi5si@RtAQLvbb(e`xfQyI#rxg6cdH;P##E-IJ&;khu*gFn<KyG&VO;AP
zWd91e|4q}bvejGby#g~&0v6_Hx!9Y^Q3n78xqxHVoyVFF%4!xvNHQ`qoV$6@gC6UA
zlkP9(eut`vi;IVzii)cFatly~d0>jI!G~)5tf}unj+8&2A%L$=A_n#DC6b}!3tt_o
zY$x&o^4{B+sQZok^G6yCT+M=&ojtbQSXo@$4}Hwa75@smm}sYaEEoH@9}IczBpweI
zdKH-F^bjG9*K|oq;~CDg&;X$yKtLQhU;m|h#@@Uy?O6iAQ2?vQL`Bghb~A>wK9u>-
zk=h`>-Pn<ePR6us1T(@26CV&4zW_@bT1xdTa+`eDhIvZ>S1%OA@V^MW)*nIpLic#A
z(q@3QxIW`PmudB=$KGl-;1b05IbUA7bSXi^vyxFE>5&Z4_3H<g|7@k7PRJcLWo6|^
zL6_AmFgZ_W>aY>(BHk)y0Rb&$W@gz}uT;SkxLwyZL2@aa#Nxra6IXVFV6X(Dpk{}-
ztS`0oizhq1=-ripwQ^-Wm1vOrT3TMpCkv`UAgzA^W;@b&q&(zy_r6|W2s8I9pQsd`
z+G}a*)z@#qnhL3eTyp_320;oY-T<+#aan0;r8jTV0T_oVaX?X{icvlxt8TMV$Hc@0
zpnI4fx<ncue6_d{E87{zq4DzN2kbru9?PaP<UTO`ZfEhn5)}-VA^{evdT*<ZhE-Hl
zKeF6zN)#Am?{qK>_Lmy4yr#kmD>OM^7_hxS0Vp=Xl>9-^$*qWtyhSDK-nYyNE+hvR
zPM23&k7RuKaPblW0ShN55;fW2>1JYTO0IeT-%qLVJg^1$G!Iy;cBQpe4B0vXF?rVE
zp}WmkWp@)19_%mW9Say3&A3x9Gsq=$Qen~bT95B&Oaa6=m??E#K7m_iWo1P&<}%E$
z&7J@<8DzFxOmwt!2kvJ8q18f1%%0{=7;N$dG^vHTIXbZR9X@3vkKrioa6nO}MS#)z
z=|EPHB;cellA4lIgojTsXe0rHxnFkzVbWGUMOdq(q(r;SH1zkE`@q(|cXhq8nW$B>
zu&~fBdHd1P$tjfmHVhViz6-&hJ}IfI#{%Qerm(Aj9Z5<<!y&VZ%nsqT8O;i#5+c{6
z1+#~SlccVy8m(TaoekUrGhU+tb_t+yb>R5@03J`jhP<ip<HwK8j~~Ax33`FRVd1Wu
zlZ7BS-PgMVgHbm>B!>A7BnY_`tyzC<o>4;hoI+H&wWTH3_YB9v#>P|q90n5(0GngK
z)^IR6a9dG09Uy<1NM7sVw6l{XG03h<%gV^e$@BA6a>#WVz{q|wxB|<;ZH4=qcPH?y
zPBzrhTeL?qfLIc+QL{uXgRFnIGY9|&;4Ktdj9?f)??Vfi$Y(e7ZYg<b^xX5m+?yG_
z4sbod?V*o|!P3w{2D&<2k{KKl!XR_Jk5RF;EdtPxQN#Dm8(R1X9xUP7xotVj{e~2P
zGu#0Jh4A$~Iyzb@?PQOlrxI{{d0p(}$432*LIC_CtcmN}c?1S1yrZjY5K}e9UN%y2
zFy(8|*w`4EbRZS{`Ex&ZZ;iMJ=n}ot*MQx!g|ucI?JU8wvR(pGSi2Cbqw5VkQ^nP_
z^5MgW*9vZc3u(bPEn5K}<4u$&yiV<00j#s#dQl1>jg{#f^8Y_J%_RxKma>;ueWlHq
zso_UpRwN<ZFxa>USkCCk$c(RaQnHRcGJ>`p(afv;nL&`#?(RBg%7RH6{gY-^klC`b
zvVtaO(?0_$92Yt~3NFK7o6uOtx3{*y-t(IvMyz8|mW6Hq%S&lr1e*6G3FNFRDk`G=
z@bFhl9>QStL6FN<F)F4M3=eA-YL!VE{r>W`+f>EW^m`?U)aB*eg_<SLYiepF<F3H`
zq#&GNEAjH>BUV<qOvzAWDBt8Gnc(p7IC_^Aq(maGjdEKU74d&8odPQhi!6wG%051g
zt{dZdrKOw_Qc|pff}_hw5EUMsZ}Ck!$v>7AEFvO$5YoPSm8%wDulhg<a_}fJGBS()
z3<;LUkGX3}!KY!6@mk?dj*1e0j$&LhJMmym@ah&4;O7wW(r@svSnP`T0CEhf%FX~-
zi<%I-h{%NBDtOxC^QW=tRD!4F1CNMIiW>*_Pu_9|xzS@Tl&}2Vj;VyC<SIx$Ltudb
zz^9i{hpXk~<zbgI!%3;Bi{3N`<|(F%7NI@sAee9yS`F{NKMAQR;Pv&o*M<N3LA#Zc
z>Z1+p1baMn#x5o%rU6U*pBX-Y4p6$)7II#WAfXbZ2PquF{Lh{}bMZI@?t-ec(b-!c
zQHdlaCC!4=Ak1jR>J$P&L_}2Xv1iGqSF3s_M8e<SpIpSF{0_TrE{Lzi2YfuOtwL%_
z-%OnEKM@akZ$M`5d{dkb3gk&cwJvMX(a~A!>+38S4Lz3_kh7d8-z}jmli2lYmt!>a
z9WW4h*hHK)5?s6Yp#6SYMrJ0w%K<phYv3AyETNCt?u5W6Y^$JfT69X|bC}-a-)Ao5
zTA%3>e>;>NpJ3T}`C;ptU}h(boZ)=cfnK+Ul;cnX`O&oNTRNCnC-^!@u?b?njbKFI
zQ+;tYL$oW^bsM#E03cb8RTi>0>?+Ij9LZ1#I_El$Zz(DNi$LcioX*CkaOx&v4<BYI
zq;9>!jMT{nT`v|}b0DESMB<}h;ycgJbJica@JBp6Y5;-2k<@qZ=8V}P^qdyE`Ue!n
z3R^?&qAkWjP6x@s92k_xYF^GR7@|z1msdW{97rT;`w%4tL2|~U(dX2o0lNy%5gv+i
z#b_59l;YGlHZQ?)gU^F$CkJ9Mh{ngg0|)rTGN5I}JJpO)GSIH@T7Q!aqs%jE3uETw
zR0f8o7WQAkPGWcMPLGh@v1^^4R5*3v&XP5ecI6k1UTR;i+G~UUXV>ep^mM+k<C~uY
z|EeCbGq0GK7&W0C$0XpU&d$z-zdbe^4$6Jcd`?#seH|-CFs|{h7>}@67G7R;py;SA
z;$wzOUN|^7aK^vV-<0Y;In{yd)!|9T`3wHQV3pNxM)+~P;jM+i#G82?W?zb&{B17P
zE<|C0;+o>UuQlPh*-$~gw%y5rW=i$S01BR*BRlem+lDSNCBG6dzL}YsRTMmqlaq7a
zo<BM@wRoh=%nB&3k@I^#FxYv-S_1@BRz{|OwlyT?+qc}S9Q8_Iaj<)6g9`KhhakDj
zhl^r!tw{PEe2+Q}@<UPI<taq=22~P7y=$R_bpL)HNW#lQI(9BpQAtS!(b1#~jEwne
zaCLGa*8*Ta>UO6vzs1o1!~k<Ut=_+Xud1)FzjVOD$_n4HA}<BI4Zdc%I+*7%|5d6U
zt3Tz#m;Q_m*;k3_uEmYU<cQrM4#XsKn0UK=?Ki3Z_PlwXZ$%4OvE@fDC`<z3_zOE8
z%+CW#*Dybg7@_r15AMxf!i@HkZ?9^f<W>Mec;%c!>AR#ANQ<DmIctFrjkg&{Vd*LK
zB$NN)S?&_?6Zw4mUQ1s~Br8@Je9I2zAb1Pil60(){Gjug`u<wjpR4Co&-~<<_bCtJ
z_(5i;IF&W4K!#ZrP^hoHIKS1jn!l+*sO>=KGN%&Z6zRHUC0@I=GZ^CUOpw+<#^SEM
z1_e`Kru>B<^gW`beFj!O_{$gP3b)Nkz4-We0(AnIpV@gR!%Q~l06SS7t$;uvW7$zl
zc-XT50%m~wfU9!-fZ)SF2_waGLm)dV%Y6RN*GT%1*4foWPLp6rI0ATr83f$iTL7!w
z>wz+ocCB+>WMrhmh?x6d!$RGfQfi=|HF=%`QZFwrU^5CMcD#uu720e6WF*L?to}sM
zCIA}6d^|8Pz+>3t_fLv@LjvK=_9&n_@Tyx474(05YGhnC5*;6(SEy43AdWn%dVvyv
z68$C@VAD6xq1?@zH-Sor%zd5d3}U=&<m}uWgN$CSbN_ew#K96HDSTFtSe>9CJOK6=
z=u&XVhLeTd@(T)H_4oG!48~i_JzYSm$xxRl;2Ij!x)_n#F98cBJ!d4JZqgZ7XvYAU
z2vjwI&hG2onts|KCiFcuoVUS`o3S@vOY$xrxAOT|Ks^KC+IZ96l){JzAVlEN%g|t|
z+HaS9|1-4v>!x4XIV|Pz#d(>SnOPyTWFXDjn70hZ^2KdO!-V>ZG|Zm!JfH&J0+QVx
zvMZiz59F(S0R6Xj;G%W3=h@CLke2y(hV4({r8vU7iI0`T`qf~ekTx+r1U43o7s669
zJ#rgh#@C97%ryhc%ghZaLV(c^`*4Mf@!^|6CS$)Aa9#9CEs#7`iNTkZtib#Py$*kP
zd3ytwWCpn^Qx5E9CV5Y5DUdR(fVNS)nhz%h_z@u6e7!nXP{kR5R{rnwv&>GwHU0rG
zYk<`WZr1M%78{7MX_rfb2(JVfDX;B#F7RxZ_2JhL;{~|k3p*LiuT1hhn0ImYF1Ow_
z7(6|gGRQt1wJJuvxGs`qG4Oo_C{6u<H4*3k8?lXd{Q1*YjKFzx;0USR#lh_%D4uZf
zeC2?~esq|GZ3Cr{{`40?5DdyOA|i^-)AP*iw10Xj=$_sMC9Cu8s+w?3gaEmYT|>oh
zoL%FQB;t5tbG3#Aw(*T4IBH750njaO+4F0Ag?GL{)K7}WTQz|2x;j`3pq-bS^-_Bu
z&IEbv_Dk)Jn5Ev=gCO!VXhLKZ6mVh+Ui0ZDf9#UjnN~K(lP6#mENW+iS+C`j0Py(`
z3|awX3k3kbfC(!rD1^z%$)Os(J@=OTUjRv0GL8_UIhmS3NC+3%*Q{6qSqF%^DYlJA
zrV~EL7=X-}D8bAr!OYVVz}iAI1*91i$XEq{pT)$+)=t1+=0p&LQj!PR8OkqcA49`w
z3xhd({(3EdmiOKRQ^++PEw=z1b`VqoT-FM!ERVJqN;Mx##$5t55(*xu!^6W|fQ=%v
zz@nyXQ-gTP&C~S^qK+VJ*QCTnM=t}oN}!Gp^K*l~kW*pOd53|X{t=iqP=Sj;xvP$q
zi>pk*2o%xSbgPx<B){~5yukx94V3X#41O5;gFwRVbK(ww>i5yLhLP+4z$aK0M1-oV
z`CWie(;3SaHR3fZ)t2@w0FbR>kX=w7^*|Dn0MK;c<Y<S-Zc>;^)Jt>VyS&q0@3*Hf
zK<Wh;59oyXvFv*JfSm(T5~x4t9Od7r93XILLDlY9eFKqh(&yL)P|ZAbBd~(20C}Q;
za0A$6<{dU|^WEis?K;;&-~r#?zrRqZ-=N(bbUh9f-V6=XBteP?m|A8V5^v7!MlOyW
zpduvmF$mTLF{~OxmIW_@NSMK*0Liha65LSQaq^QZph!?@#u0z?+qZilICPqpy=$O>
z!PHCe&w1?JKDCr-nHG?d)Qu`^#?*lrk&oP;I>Tw4V)s<QyXd&M8~`v?SFRHgMa~N!
z&tq^M2cD&4nA+_QR`_~bOUvM@6R1FPoCj@kPUBAiNIP}lz{H8~JxS}Ci;am{>E@dn
z1gxFs&993Zb>WJrsflFDp85#`Enw|>V**P_ncLgjD%4kD=H&k{mF@ZV;LuPM5Z1U&
zyY7ux^8&ta2}l;O;{j001!Y*BoTs41BiEfQq;XyJFpa(O@U4a?(2c>3@+o!xdsSe6
z%&u;qi<5JACWvNe1sX5Sxl@1obQ#cPsOVC=^_zl0W@~FJG7I2YbFj&PW*QZ?$pO{_
zB0e|B8kXtw|3ITsAkefva9tlx2et!1Q}F=^QGoCMM<&<>h|ywav0EGA)3Cojl37sj
z7`R^M&>SVOA0U~axq5|F*eGS765_ukB2a=_SX#0IR|D10y<r#J5fEbWLSWQ@jxh7{
zYl4L2;&ZAfcAAxu@sOBWxLVslTRQ>hwbB1~$YxQnXY3r?oTqRet0D8@A8&xLB^9#|
z(Se}?tqT1KOg_8_NuQ9w3j0?t!1hAt1ohm~QV!tH7UR_pfnXd7eC&|1RJ>^>(DU9y
ztxnWcO?qyM*i6=IP5GWRSk-JO0Ilh&0&raQq;toek5U9VwQ#iit>^xF9uT9lb8@U^
zTRuaDqig5Ub_OESg&d8C8NUTyv_atNzbY(Hu!1;N>6B^NN|IAg-hU#bTXSt|Kvjxs
zp2;0rvghYE=^(u$3?6<TiW;h1!0iE)m4T?EzVE3ERFbYe-l#)m%iRT^FU-$F&|VC8
zm<jCE$i5ZuXaXolbwgQUd)pD+SX^8TDE4>2#%f1PfC`ZC<js3fb<jy=0N8*GIG;Gr
z#v+K0ebN7)Rd6LMpcn`O(G(Kbrg^}<nk;NaN+m$mmgw3wRbamG2G4^HphR&00P>eG
zq{7=<9c)YhxkCkAq*q}*!WrMq<l}seoLd^$>S(1+qCXiJmVl__T@a6l|Bu60V38m$
zm0R&|6a|`K^v_kXA&7Nk%f(rO@SUUEZ;Q%T+tVcj%2XQ*q)aXLB*T49u@)fA0b5lA
zb+J$iP#?iFgSdIa-(Ocp>)(~J*DgN<y&kZpwSVXcz#mZU9ndNO=FhXSOEWQ_IY=G<
z=%>@Sgn)*{QUCXSQ6tu}J4rw#PS^3XcDecMIR_y9?dd{_r4*FK0gBcKDyNc+jErr#
z5NZ_+iJS2}u!DsBEf7?Hk2etpw6wH<d`ZS#1j+JTLOK20d>sgg-OL<?_^>qQH(vn;
z*8zAA66XK2U<-5^2*Ci6{2#&A;WHKRH;^V+fbj#Nvvx7TiuCqvy0RMVCVH_u(M;S0
zIKrKCqzBZln23l6GN39wodV!3W9YdR0s%oX4hBTxb1;1a2o>Z{-@dGH{U1sp4nRaz
z>ZUn6l=>$v_wAiH1D!z=UP46@1k{?C$jFC@-RgzDKnn$OI8><w!96A^>9Mb`FWSUs
za)S`YZFlZ%VrM61tfXo_8&@Eu6jTK0K{%|IHF);SU%gnr5bSZs+Y$&4WJ-k({rY$n
zjD8xTkPtSvWa$;!-lfa}3XBB^fDocLIK!e@y#I&V2%`1v*0R|p7`N#8W;iYNpg<bv
z#eR!Im8-hz>!%1hFDqrU{kOvWge@A>dzMFgdwSGgzI>T1N(W2Zf^@a~JYha_5Jw@E
z1#}UGw%s8mT?S5?=OivjMw6N^mOx1&)v&KoY1|&ka@^KT3iLq;zU5y5mE(8sYOJQv
zXf$RE80pTRV@XL#77mVLS-`bQK*bhB{`yI%Jb?r_b7AxhMBQDWH+jrL4M9)NksFXH
zfnFjP2}uc1J=|R;i5_r?l?1Dztv|=VfB$A-V#?|8N&`iJM*uG}MuB2)3`l6nzPO{J
zQy{*;LB|eV>TOt`{W(^<M#0;)=e0H4O3q`c+E8!zztY!#t#D{^oWfw|1`DZNdsv_L
z_IAMIU%!2up<Q8_U1HP*l`r>({*A+4IB$$=fz3sYo}Qd=+m7>t(pJ%L^t=C1%fxqf
z#(XSxEc51bo%QYV<|HIFE~fcI=-h8{A>^8cY3u!$Q8Fy7F}C+;<=xddoyrW;YT)l=
zfK=)C-F!(AS^iXc<mr;S_CtYn<f`+^sYYlKi8wx#Vqtw=CkJfp#<9lmQCqesQFnRl
z>5IVVOWdR&5~EJT&$KohV>YZmUE`@to1-Lg*-6b5BAiH%OG(V=Q`$s9;uuU`&Y~px
z0AB8vSk56enn;J~bP?T{Vi%y~L2O9tcVoo4z5*B}ht*BT_g!Z8j(mK?T_C3r9v!{X
z?S{76<I%8~i<?rf`&pF(Et>e8N;C&%O#P-$rP_z|;L~PLD<65KZ-boLf~AW2)i)i<
zFtr|4rfL43>&1#IO-)x#>auhCvj-nHDVj7F^%eb6<T*vfd(8H?)o=3xGlum|^HY*w
zrVs`|)&WI8R}pI84Swg`#k-O(WtI$RCEQ|1lXSL*+mG&1(vnd|{x+D`ZRpWo;S?&9
zAm4g*>cyYtkF`g!jObIiV)2R9k((FT-dPENU;%RvYo5_<>Xrsj7*v$=Z~3u^Q3F#S
zwF34DJSFV~#v|#nXi-~Or0<VoK4KdNk+YIrq_hRPw}<!4a7B?r?h2nL?nOpBYiCKj
z&pc_I4KHVQYg?AMSY{`GLj;U${;s=?JFsO~ch><$mhYL4$GStiT-@Kf=$IG_0B(5#
z(?C-cK-21;p&)Vhh#crbiZY<#Mccq}rGQTKe9<><`3oZQw{uj{{fFKOj4`CL4rp1C
zVzI>V&!z>3WsX#x<>Y#SQv4mEMQO41Cu3CS5BxB}&gQA%VHa;)BCUmxcA)-dsido0
z0C<5~&BcwwIUq)w6u)Uy1n|5^XtSPWX6bMT5-8tpA?NLb^gOVp%$^OZ`is^&$6>qO
zgvc+OSwkh0?+m#4CC-?my7A3?I^2eyq!#9p^J#Plw7z9*uaZOSJtFqd64ZX!aQ1C}
z(=)?&8IJV#T4_K?KVwU>$yT=*aYIT`WQ#~;Ro+C}Y$2Q6+uN$^ZbHs!P7p>NqwlFU
zTCZg}qU<qEn<p1HiojJ(`3Nm!+e1KXiw8hRK<$cM@8p8OW;yUZ^p&LU+`-60W~&A#
zt<AZDg~%7kb*ZeJ$-J`!!%s$Q1uRpjFTX;WhAAWWNPSKBgMV0*p}f9y>sK_Sy(Hh`
zSgCrrbU&T<4KgeM*`<aICPq>iUDJ^W2MI&`c3ho&0xlTn)6PFXB2N$GXqhISx6DKk
z5oOg2Tll8zY#r{L9-n3Y`t^yt%@&LWySt)dblGyG)I`LqR&ew%&lo-<_)w=N`lw{L
z%3s;=6X}hImoCcrr}wn^Mab&xxdt^?VZVt}UEJMLn~wR76XImcTkzjJ)grnoT*r5O
zEYxGqJ$-8h=c9bpz|1dUJ7pI^lWZ2dQG2pJId|=R$ucNOEcPrdW-hc_#t@7%>bW(P
zJ1u>NsS4acJ)x3jj2f)12=_R^=~cj>ir6K>RMXcV0)Myl9;I8>-i@4K^gc`FB6p1#
zDTtUh{1_O>LQnq^q-?VjVF1@n%N*kF{=66aB+5(Z-pPUfSQTID!pPKAaCVn&d#{wo
zp=rDF)^AbMmHx4;4+_feeIpHRT2MsmyEtNvERV|$of2B2sIX=(#42H@PgmQtA1kj7
zDX%?Z&PNUj$_5Kdt={k~EJC=Nocdn2`?Se%y2&QPq<3L{jY$vZX)%5+uezz`n!8N8
zs`AP^hc^;q{kPjFohsdUX<`<JUqc-RfPC>3HEG=a`SYh4Re6}5sHoS35Cx$e$shVC
z?Mq3Z<Ab;jqXG}gQd3O1EEws$CI+WgIohrjkxB}`lB`g_XqX<&J=gN|Cmym?N?6B@
zZOZ_NhVn}4v0RigJwH^H9!ONhWN6F1wn_4Tlpo4<8BB&J@%Ej#3C2C@2i|qFE!nP(
zY)J>w<j0Xarlrr`xKO*Un)r|pFSdAHshcPQ={*0(4<#TT0Vu+sxdc>j5WR{V(nCf;
zW7S{jVbJmylWIVxd117f+Tip@<7An7@7V^!Y2zA=apcdnIlhaoZJ!0C%pAWI&Q3Wk
zNKjHTb=#GW7?VQ&Ew5B%i;SSR*vmmSdiTh!2Zv;llprh6GB<e?nJuxe7*|TFvQZYu
zQMc?&siaP?Dclby{jWjax=o>-=H6XzbLOW6ogSiVKV0Vj+_af$)Kg!J=a9iHEiHXb
zBfE9$ApkN)cUE)p>za$IhZ=U<_tnVHwkEXq<pOaI-P{s{Ra~{)MkrGfWN;gNQpMu0
z^NM(DLr4aT&y^BUbd7khjIwbq(5IB+)oPDuCUa|tAnlX;8Do_58f+z2C);K$>+Dnm
zi)CGza03H0aQy7-ZLpFN@&)n6x`8V`Yb{WqO0vnAlNq2DP#6OK-mMi1I=?_ip@;gj
zwZk0|&;3Dn8u%y>WT>xb1Iq=~p>5e=&crT{M>vk!>;W5(dEeW+Tg5c1a1kr+qmCt>
z-S@9K6B9b~&{2b6;O%XyxMK?{78m}u8q0}az1eTPlXk)jTvsX-6?|PgI?o6E2KZ4;
znG=QJEA6`9_hW|&oowBRV+nAR%Pg-AK%O``Jzdz`EdKN7PZ@PFz#C)<J&eu}HU<#W
zOgCu6+O>Pkk&gzx-MG?rXPe`)@(+Vg&rhC{u2U2|K-{==M9}^=zqZ~P{mIs2Gc3Hq
zaEG$t@1DWl@+x7)W2?{lwHs>56hkZ4XZI*l6BZxnxux&uOIPfL(;}DA*_ToUUMROo
zn;AyR1g||<*f=n6cY$zyTKr|ANbFvm?K;!xcF6A3nRJ}N8)vl&v#K#3GyP!Gu4In3
zDg#*`Jmr>=<kwr#B{Q_Ll=e}$loR)6-N|t|&~h^fIu$D{QL6=Qpej=IPA7**wc?5n
z$dbx$-ppvBZY@*f8cyA~bo^0On1Ho~<K?qP+S--zvE7Nr;@yP}j@`O)_fW*)Gm&?#
zQwiwttc1n8Bdh$1?1Im=w}0#|E+JRhkVVbPsT;`ZM&xy&>=hc1qDesrJN7jl628oH
zS+lCgQonwUL}{QmhNsh<8noqX01TLTY8{QMs?z+}Y$kiX@~d{zg14sda#KH@$6OgV
zTU~wlkndI==-2tep4Sf;=4$})Kx<&H$7A4-KzFF}x18F3!NRYZmboCMrm3ErY4VYU
zMJ)dpr`cG4X{MlNy#XTDE^6ZSPMj}%B0oN!!|Klw9g4YrM+-eM-3{8?+8+CsFYdQq
zSGT)L96`OjInEM76lGhTyRrA~GJq6*1dEz7Df^peUCW%hUS40WKcw;c>o3~fuJUHv
zlh_HV`jR4M3ioOI!t>oGqxKGWr<8s&9xKqD<HD>5iVg6+O;EJrVDpjqm@Va^HO4v-
zwh-}Rq0*76EOz;C$qTIY(pkCfR8{F{C)x7tT#DWXEsE>LmIMC8W@=7@SSRN>6q>iu
zGlb{aKfsCYkLLIHJ_o#A_Nt%v0$*bfedc_2hz|gdaGnx5SPMDxN*}M+9a&dDI=+~+
z?~+|<G+ON&z)N$ezp(Fk>F{E%+wMYbn%ecFZ{Fn=X9s3#aW6=y^u3g>_tq=+B6i;g
zhCeaI`}*FSkoet{Abdr#m-;0HD2XR=FM_gCLZ?f4Xv9X=e@`GTZJaLrlrlN{nRPaa
zbE)`v&r(@QJxzeZYz4Eq1piT>bQ8zsVo?0KwyrIebb0NF3r-k(U}e7jpu*;E;aV|b
za&8XY(Ql;u1MzfIPX|H9oRNL7q1N{b%?M=rPz8K3Pv_@Gd1dbe9Opm9a_AHx^!CHG
zLCfi=$FfwzF6Js{e3!>;X*<?+KD>4_GKdVAFHfE5z=!a&T%9=m;vBT__Xa7qRF}}N
z{)KzbH}ev<HH<d!cXJ3gqc%H21MA)NE*73Vd*ORVSRC35!S00k*|#gn*^vl@Q@mZ8
zz;LNah)3lLgF%^Hp}~L<mPj5k<1*pA0sB~a*{K4NARw>Oh)V>uMM_8W9skX}{qU>1
zn?j=M;eSGU^Lq|?8!8rdXnHNbaP5U7MBvwpmkYf~@=k_Qh&Nv$XEKR2OBM9C*AwhF
z`HLH%nuMR}ubGI(LtRisV0j!dTK*NZgn3z>d6&!_+aS2V!*`BS?@DWs<d2Fy))rKM
zET6=*SF`i-xXS^s2MhgJ(#hKWx%B=Y+J!DeQf<T=`Yxh<l~biW3u_1st&yT}efGrW
z@5iF|Gkb%I#NdM6jAh1s-pX_jAC_;y?e^xU&hihawF@F=Zy6EgJ^R^+Z+V|8+u>jm
zo}-)+k(!!{o`jSiaju9ud5;}=W~Q<Gp9m&siK`1ggB{U6E;43y6E)y1^hhhwN<J;y
zi1L_Pt&dY)*sl#kzZ%B40=+ccI)xI>zFB6}Mq)EwAroHYqvRSS&s!(Y8#Hk=USTI>
zW1zn!ii$Q|C#fq9JITOrJtn(9T#Kc7O9wz=I@%x}`hDZf90<|az7J$~WaaM2!8L?>
zAL!lUAf{liu&tk(Qs}mgn@nW<SRT!eP2NR$A}1-A8YdV>teMmiEX>hhGPZTSBO1Gj
zWvA$FWhKqlEp>}_v8<*WE8`*AG>8aM9r{Zjq)F;_zI7yaA8G0<7oZJ%8~O=@J9Jb^
zF@|9pZB?V4C>z@^vK^q^heGl!Lmh8PI+B-aI1NwEE`okye1*56|4hDk=sg!C`lMZP
z<j*>oj8d=Xh>4Y*nL(;~XMMY8v8ITOrAdDGYsOwQ#nOpCtEtIgw_Qq+9LwN)<{yB(
zYruck^2ZnpkuY<bsy_ArevlMC=b41PZh@j)QXiSJg&z&mAnZjIj^`}S*5w#NX~b~l
zZSNA>jU!v%3Sbu3)>z9%bm2z5u3X#RP%(;OV6s2z?wQ`5GKo7;PF#tPfruk}av*NQ
z@dq`YsO%z9CtCXLaqZ3&{%*5>s5bVPY;|IvNbA)dkhrF}S)H`^#O&fp$p$5D-j8)y
za%j{v=$b}piBXRiOjRHhB6s&vDoTacO9iY;M+^5(6)#jkF(d5?cH!>IL^jjD@3Fl0
z*3bB`Sh{vkxFu>)t_+Oy<Rm$+C7z8eapcH6Z&<8cL$F?IbzE-ydZdr;oL9F5Ldapf
zR;`EYg~kM}_h}<3HM8Eh#%Sr#yULZon}EIT9Ht_^`Wz{&x>b=PWj&S{mPWFl-DZfY
zQKpJzaVd9SV{Q;NN|^EgHE6DW3a0_R7(~}U^?h6#q)aeai^~8pBIho`TR)Z2J2UnA
zSr!e6^<Z{?+*<y(DH+Gf;^l^dY|A$m>5iT?R@9Lr#<4Pt!2gNkS=67FgOy@gbi=+y
z=M?$vzTS}C`&^@*iaD_p@p_xY{11|gV@Hx{NE`TKgi9QI1eAYeE32wX6g|=KYU7hL
z9r#?CQ2FQ2WH4f$M*k=S)Qqk6Byg)bKE*tI!i+98S&R7VycXsZW`6H1LU&#bTy>L~
zDvH>MXOVDorRJkY-jAS`cj@8qaHt;K4DYFI5I}x1OAg^yrM5eP_iVZ1<|#Mx#ggt7
z6Ew;h^Z}K>ciN6vvX-9Z8QOIjjuOIU?j1J{6m-Qg$fy*1qFi2W*a?j4Rjb^-O^tp!
z_Vm=M*%^R<<<e!=-9r;a4XvL&i$C)0*5$STHe1Wr&feF;sC(BOUlnoN3%Qs@05vj+
z-v*5I_C_4R=`-zUMiWz6xM;M80V4xR-WX*T==)izwJP4xFG-0=W~O3w6h&~2AGv!u
zd{i9BN>|kI?T!h^8ortxw;B7kN9cH~Z@Ix4bB*C~y||GW<@fqo@7lu72QHz@i>tlE
z*h>19{USFO-hxbA#hcn+Ixio&Gzsy9p@p!H^7=c~!JQ!$#pcCuVqyxu;7h;)L&9Pi
zaatanTEU?)ji-_xC$BAD*Y5a8%m2==VM-b6fEP(0{CvEWLBs^$>F1s7dR>cCDKjx@
z@&;F03=gBNZmX;dvP|PF72dn}!=mRqnsoHxD?J@om8&5LY^rC7E7^vUEv}NhZB6-R
zQE8YpDzJXjSwTGO?Q=o*vw-OWsB_bg;1{SF6jt3?{dgA{6T^_~+i2Ey{hBkucYMO6
zov-TNk6bp^g~#R<%!kpkvI*Zk&$#t*IO6r+Un5mrUtxD`mW0z<2^psDFku<>*%xdh
z9UF^}w(G<0%4l4VZT$1M%iVkzH$11`J+@oRZsQc6abc`_ly74yZ8kV&<?R_tHgX3i
z6<Ka<l|Ws$o@Z$U04g(MRscj!C=TrU*pdlr7ayh1oFg?DsIINW>J({Mj7SnwE9d^Y
zkk=6B)32+|;$1oVu54Ir+yy%q5*!9(y%gm8BRkL(Zby3O&SLVfp25Z5Br|8*dulvg
z(nz+sEUnvF;Y-Paed9v%+GV>fTIpp}*FzIp*H1jw&GN8rh`&L%&B($DGv^j<J>-2?
zlGp3o*JNy0xRpA~rcgu8I=hxrH7IQVUr2y5hkaK}Wr=zhuyLgFb=7FA<ZI@ttJhLb
ztM=Bjq7*O341G_wRw0Sxyh9w)GpAbr@9+K|Otl1IKFw`PS~ejM)46o5M-@~9k*NbF
zPqoT1lRzBF+nj~(&i#}h^_uqFv(^*gHkiO^cTX<sF4GnvACj0?e*=A(;`o1ubevpv
z*!FIR^$dodcI<A{ZIW+KT8Zjz=t>m5wok&{;AXLM^=1T|>gPm|x=@+3Tez4^<RsR?
zC$zX>If-AU#{`iXF!ruW$^9av<(tE{w@Fr$oW2zO5)l)VDebhZ=xEFFK|C+t9F!0W
zpj+J&>HXBqe)}l*bqFygx8Z<H-J*`R(zl{=E>Q1@Y%_bVhb-@s;=7_HoA(*;dfsud
z*=VEmH&7=ijQHgXQBUWjku*D!Ug6W|_lV=Fr*`^;-|aM3Bjq?ill5Q|Fh!+)(8C~3
z2lb5uT=(P63sl|%4K~h{I9ZBm>^w<O`NoTnIiNQ%2RE>6sT8o;c&q>~!<jkX&soSl
zr1{be6&6ifL8l{-?xAyR0FLum4?o#k8E9KBRj}F~K!#y>4nh+j==s#__Tel0G;%w|
zlHPL6IGglLVnsLX*E;2G&NF3;)WO60Gd=<}rl*zSA^CKSH7IUrAGZg2%(v&nZ1uG|
z9v*fW53UqM&2RqsO0V~Bt~=qlpU7o%a$-7IDQc!U$m(a;N5?<?X>p^I`#P1Q5^eYN
zB+t7nxLLVOyOaP`P*zb<)7H+PpEo`_K4zt*{ke=BFCC*#EL?1ygR6|K^UxRxtyJ)Y
zs_8y1JXRRRM(wUz<skSXZ@v!*;PG7KZg@Ip0>nIWy|56_r3U)?yNKz3q?>};gUZrQ
zuiB`ho#pURW+;1(B!Jf<Wj>J`p3d+4v$>(Z3-PkoN=j+q5R(<?a|M(ZRMZARIIggt
z`9(kz5$1{6%Qp+^`Hh&2cps45D<_vd&y;m~oZaA={f*${s&!QR;1fH|5CS`&QLhQd
zW;X10)7{$ddwbglSDw_;%OGpLr}EXB1hDqK2p3Et+H!EhMK_3s8q$BcDgIqeNa`$+
ziG2I^ZFF)n8#qiLmp#A;TCPEHspu3O$Ld6*tMu0jKQrBR%=P-!{5CesxBiDxZv$ap
z#7odW3hyc|_d0U)uG<%iiHW_h7fEli!e>G5g*MFYO30C8pCM?GgqjNyLhI{=U6%XO
zWp$L4BIxtV%T+;dRuRy{(m@Y}x&SW=OUJQ9%5U#EpxwON>0OVk-Z~I$D;-O=y7uiA
zGPhqD5H0y{;o;#M)pH~h4sBH6tkL#1TqTqpF9!1e?vD0@O$UC325r>^4kaKOL(NC%
z)qJ+PhK*tYl$7wh$3dBI*!w__$7hg_%y}Qu(Z$5YX=H<53Qf>Tf``Yn_)xP@ePMBt
zML=K#6$_3efZkC{qs7I=eECFPHT_Nbjl~B?YPt_K)K~IpgI=(3xQU9rBle$jil8K6
z7LxN=+(fROMJ{&jcum0iG=Gc}!O})Pg@&quUKZ<^So^Vmou$jKwk-r5v~l_#dA8Lv
z1Y~#X5L02@rUNv%zl3`BKQCcHLwCLQu>*KoGP?Wd5EaKXT3VWf_h)LVM$J|DN&GA2
zxT2n=wy<Q63SpQz2~;vNr4ts`g=%V&l7WD4fL0_P&wXoR8qsl=FF;~0s@o8R+U%j?
z_1G3T?otf8adJUZ9CWC|^(ZpQDMB*5ptb?WVK&vxCq#3X!;$(RPAbdnY{R<F>c_Ez
zNX;QQ)o6Ndv5}y358C!Y0etl{Q%WCDtWlsL?aXYmaV|3-*%t2mJ<b?;IsM~@i#L>h
z2Br<)Sg*UIs(OQBmAzEm8I_8G*JZHQ6$O{_4h28>6MTtGh{X`EB<%t+Qt1VD*U@Za
zGrYZ_OZ=8H<CTsbpN7Sl`JQB_UeaCRour+lEGMWDADlXPx11gX8ndl_B-e6kGlKT%
zfBr5&t)TbyY|tm=@)GTTSW9bTkl8{Z{)Wc=FYP_v$D)Nr`9uWQ{FwQgnR7N8FyrHg
zz1vinA#d6b!%OYt^t2yt9FxIO9a%C_-?%N5Gnb4_OdQTm54bFl^tN>vUQm25YZLH2
zJD$qR&7B5O{`-#~KSb)THJx7P-mE|Is@@3v-g_iPw>3@|UQwX<{PLD3{e9@J8F--i
zm<*h=m_cb6EP$$;bgdl-55oJyM!T!1kTKgjHU=~+N3m7;Xwr@6pTg+yB5ypu+z*Rt
zZ!rpQRbc@Kkc^ZC{wCTr-j{g&jn^ijF<hDD!*gb<e7zRd**qDFRU+D!&CrX|&o9%y
zqyvSs_M@fL-i{bnt$O#TFwg}u1CHRnY58>X3k4s$37TBY=NhZV##qd)q2d61l4ZEY
z$?fN#rs$41z|+^IRMn`icCD@ZoWuhAe_}r^zV0gI;^H#Duu$T(q!7zt&;v@!Dz3@E
zgzSG^k#amD<GDr|T?+c>cTC+^#p1i-ci2Hc)AITT7){+*n9OC`eY#r@Dth$XCR%|e
zy@%jHoH&95;z!X2OQ2%Bln?){qSuhz^>{#ZOKfGc5q<$QK~91*8XQk72XhrRzTS9V
zUhF-CXaLg_{B%*eaBj~7=!0qBzY`>O{|1dNAYw|@=@n;WyiZ6-aJ51(SXx>xkCZWi
zCKS#Jh07%d5u8)tyDj=zIXPS|tE%7#Ko>YfGs?q%VR^KI6_kN&&{mgeKbU{>BBKXJ
zaI^hEMfXsl&dNwxj<NEG8?$b&IM3IX3mkBRm(@K0Tbe8rZwIy(-1b8*)A@cu?*^lw
z>7IYQD&q(41^xexJ$mt5=l^*A$>Y|8&#H+8^;7NMhotTTGn;7t8u^onZ{&u`M8~KI
zMgOJIBl7cWGvc$ls-$^t8?MZYYD(j*6Y(lOBN8p!qJR1DxDp?^J%Vmt5+30ne<&hY
zStbMq1-|E#XdFC&%enr9Fc+T}|BRp(<Emgc{Vwg+Fo`h9Ja6BS()8^Iu#=0r#MR_A
zw(6%9m3U-L1L8P+tYAZ9>SvvyvrCup>28O`!)TtQ<2EO9hASlaracT@PI)E?!_24a
zhlIsKuke#;D*J|}`t6L`hob$Z{H{&&%x=1|V40yYY}GO0A#o;`e_Z)SI1j`Bq`0Mx
z&%fF$P>oBvB<_|L;cow<2ItoOw8?S$^#iJVcuA7N(xZABY2}R*57U_bYMqwQ%I_Z7
z%m$onTxsjc-`tV8(uRd8Oz&V3DgrhbTf-;zs@?{;s|61P#H}zp178S>@Ol`i4fU^f
z5r)|<@N|jmpPnw^50UPvK#M=?*8Af?rDgL2-Z1R(1x6;ys@uJ=<V%k)4&mFu_%`<s
z(S8>U-Mw{8S^_?L-tH;Et7wXA%Kb2Rb*=cRYZb4-1(G}j4}Do<gBWb{R&V(Se?I?b
zdwhO!pUrAYKNYiSKNn7;&;Ug~@eg`h@e^mZU*g;eFtAAU<8C|!m<Ni+4zrWlSlo8w
zx>CTEHZ+XpMz4PVFRS`XNmqogjFVLB^ama$E33IY-r|sfR<9qX6;OWmC?C;1iE)q6
z^LKPO#h?hN#p2u}vP_(}*Ul73ABtYyjI{5P5N;AE^yf9eXQxd$-Qu(NrP>mop`XQ5
zrrq+Jx1Am6D{o6Cu7AEm#8E!DLtK7LWPfw(%PYSjF|lst$N;)egk{>W;RJLazAoX*
zvA)1VD$aY?ifIkP7+$OGkr-N<9lQYU)=o<hFw@qv(bD+xLe{C2!)b3MW6z4lnbQmW
zgP5VSu1i?-S!1F9jt}EIOU+Ge>Sn3ckD`8-mh6GUrW1zg-X>DRdt29rSFx2O7u9d|
z#-5~qZL30I3BoLj74NzdbL$+R7^1BbA0_-f9I6itaTBU4+zfyWxy9$%&;yKEIgjL5
z#-9G6{71y&W{U1!D*k&LCM6Y4;rVuATc!`G4cDzR8V)|0h3Ph}w4fPUKM~5d-A>&%
z!)vt22(2_<p8B*;*<5~l^fLvob3@xHbYf~rb@yL(p$KNwHiS*B(<tD&nM&?P@*^9k
z*-JoOT&T#*)QG$6YQk`fs&Z^<VNZ;Zw=SpMN1+-g(2|q+%qzKDpsBwlX9iVKdfKou
z)$|loW1WNPEf!FA2n}1#IUm1W%0=npaJ7~k#pK4Zt(lgx%blMH=|rc)iG^{Fv#nQt
zwV=-qKPlSC%^UXeurD-eyT(i_?D82-=wqp9Q`oR#vumm4J8`f!yePlB2S`bsFJ`@a
z<adV-+7-uA4V<!O$*KAkp-}T_O882|KF!V9XFXHHF$*gM!cz4Y#(S_%>;%*g^ee~2
zy)1`ZTCNg|f9o4+-Ls+6@J$JDPeb;R$2xt!t6d#^Xx8&e_)m4BJFJw?X_W@?-o5gw
zEAetTlPGa<!!>&9B&|;e-jAOIx*LACSlh7*Q`OC+`<JIDG6GLu;F%_!zB0S^<qrP9
zFxjDa{fB*rmVoyUwj4(YZx|4kJ5}u{&^OXrTt0m7{;~d;yn2J9CFEt2Wc|D|+15`F
z-&s#)O!q7{)sCJrWY9px3eoWDA?}3tR&PQmJ~mQUGrjP?Eo?%x2N?)`22miYe>zQy
zKAH?QjyR8gATFn@2Y|5q*6v!1T#EtT7e@V+(m?l;Xg9y3mYlTUsr0fQK~@G#Zhmn7
z?}G_mayM@=ijY>=lY7_hU9(9Y&d_5b!&rFBt=pNE<K=nlf5tHEvu$Zw6=jO(*Tuj3
z?&_^LHafMxtRiK#x?@r{;*P063)adZuHJfZ1>_IeJ6k+KZJVz3gj*R5MxXC`3eafw
zmFV53diI)h(rc%ix8mmKm-clN;PIUcfbug`Y`WHcZhbp!&5Ilu&=Hyb()xCAa1a&|
z6*aKvZ)IY_U|JSmP*9*=^;%D_@F0gRqaSnv^4trR1SdswPRCW%TIaTIkOV!Ae`RcJ
zY{-cj{RL(+yXW!r-Zf)W)3>~Q!m6rjG2##WrZ&;OEVA3Y!HiL#!C&w&tQw=R&a62t
z%*&&_6Y@1VIeAdSPwCNz-X*#_A#XV2r4<$1-j<YOJ-RU}E-pL-0qNbFRysU;M}o)e
zWkR(=2#S9u{EB>w?ZWo$FwBESHgS&Gv5Vqfwfk!=tn`iJU#Uo`GTHEZr0JhOxY1<I
zyLpq+!e-2xcifgQIAi+Y*r7((^V)~R=B1rCyooY$a?O?%8>e{kpCmIqrbf3<Fs?#h
zr7r@4^ZhYUENSWJ0>CUX9`Psl+^I@`M5JGQS53`9Fu7W{NCazKqWMGfv3rC5i4TIB
zUgF_R@al7wAi!f^azN=x1voMTdJ%O_eGa`uUoqZhkAwdd?|fSl5gQu_<_p#f42GW)
z+}X8r{bIngE>kX&pv;MYlJAd#-?tlE?wF~r%UpnPTxh(4qdj+^JFg4{874)~BgH+9
zegJc^&w26D*M4()F0R|fXt_tE8I_f>ySuxRy>`;v<P0}18*-D6MVIuMk3{_DeAJzN
z>2AMydH2A2k^<9D!w4{|30u3@p8n#nR1Tctz>GZ@*U)keZ(^w76KOp?DzM)#GhaXa
zDh<aLJR$<Un51N674A2R9}$7CM74ito-yAQ4)~eot6#6r;V<rI&p&QEl4bbhaztdW
zmX7)6=ZjK#=K$qX;4X}j@x+^_J$oO|(!cC;vDAepH>bq0DA&#8;UykEe&oGt1iI{Z
z?}S`R?afidmy?sTQhzXg&{j1gr?&WjJoG|2OH_=sXMOBf%c}uD{RUQAmUeDlzL;>$
zk-J3mT6o(F`dj@5bAhD-Gxp!|^I0zV!~De!!`GE`J~Y1phd4pImTmvq%AN<FG`RDS
zr)7!eAJE&DbBaQt-c*exo*phLkzSH`4ffIoXI-wwPf1|zRQd14C%zx`*Ym@nWFK)J
zL!*N;;sc`9M?ptpaD4jF*9lu-t<ziXmp?xDU`)&7C6%BhB?4xwykFQD`^<9ffo7*T
zrcpGfV%@clx&{0(FUEj2!zZ75mtLr;J-*XPt+EQnN*x{qI4X1!)a={ni$|nLE}3p(
zcQT2erezh6{x7=TJ09!)eIG}mVU)d92qAmV$PU?i?=5>%NGOt!T`1XmW$#V0_a@tA
zZ@*Lb`}6(p=W+K(j~iXr>v}%Vah%6-oYxWJ6i*gDUlt;2GOYCmH*HiDWdFzPTIA1E
zRUbv(GApj^iIXm*y?q^tI800$x3}+Oe<ES8@~L{{D~l)oEiW%mwIQ<Age@{U`aARG
zk-Gq#HMsSp7#bxTicV}#dhkC;z`dpKdGP&BjQFl+4IEdUry~?A7S<o|#QQT+IPrLf
z2M6!ny7?D$-kql-pG6*sXlfDz*Fet9xun!WA}tb*otkd8V3Bm8#(TKV{T3VhIUh_S
z#$y8$A*YhCG0q)r^W_rGbJMW$Sgn$zzJ=*(QBe^*@#l~3L)E+=Y8nq0mWQm>s4(58
z;Ahs8ehUs7<)0X6Hzd+nO^K=BnMcAE`u38Jz5wDDTsbRiYehWN>>(LfSJ!}s$gHf2
zDZ4if4?xvEG%{4eoj=kAR87*P#thZ58{;bd_gqb*z>5n47QB1-S?!u+8Ltm~{9ZSq
zL*V}J;ZgI?9c9U{2Oal7FgPvw)yhVY3m;}XXq_{H;i<#x+&R1WnU$q**IG`syw&2F
zr1o773%IvgS&va&ZNCfZ+ulj)TS@!z<E;V9&)DtW_I3;B(`c?BrlAWftn!H-;7XHJ
z_rc-)AW=B6Nl$uiJw_ee^o)b#dtRQr-&DLTgS8zC5nl~YKvou<-S-m4okD4k;;OT*
zAL2YR+nS%kR5cBvDjk#0xR3})fSYXogTtv7-xEtSm9w&pO0|@=Iv7+@NjO3VHk3;B
zz@%&L=<IfIlZv`}TGkVZw8F7`KV4m2nbUqPxZhu1JpSdjokw{Wj_>VF75Uh}#ZqZF
zj_~<X<5;>%d*tR+HqHGV{9yPnnNwA@bl^tm4134FjMLs3ID_t@#oR2!Ob|wcD!Ivf
z(Qd#p>%i0A(b2H#yhiLVMV)F+q4e<j#c1(o#+wm-kU)6WcrHt-s)E@RNs@dy|EA!&
z4aiPZR8)|vhUF}y562m96>2b=n423HkG;eT8XwoH*%39yDV98I-2e5G?&j}h`kd5P
z|D}L-^V@Y1-&U{m1+W-1>vYGHlW@*rnrGWron1#~Sqt4Egf<$m{8iqbv@5R_N-G^V
z4%IcP%Keu0<l9TBnFKl_jK?)taC`c{8XaYtl$Mr~D}DaP3noRGd3mi1h`$5u6II@2
z1c8>QPrZdwE_v;w;06+emAY?d9zGaY=X+aSbp5FmAswq<Y(Jf4Je)+o{gu|iwsD5L
z#Hp@XNK|`~@`D5fczWF%gJfDue6~p+xC9yC4&RVU=y}^?W{nV0$~K72YH)CRGS*_m
z>McYG*(HAj&^#((;S>uF89Ka2cz0G`hqo9t*p<_oK7;Q7c+Au5`VEZ08~79zB?g3d
zxV-TBW-NMq?}P)^-rM^aw3uSWk3dh)OrMXBJm!<~(}&3m4=eZ}(>L`(>bgHneMlOD
zE<Nrmwf9>2bp&qjDoS<K&Jkbw{FE=mm9@3Cmw4jLNt2Op8ot(k)W0hPy|`-?xwzsj
zw17a-hdy-=av6)u)G{+ywY5;IRDAhoK)L6@^L}D;Sk{%+%oDyHo~Wl<lLg0GA1iB<
z5m!Q?R<5VPPt2`SPHi}Po-7Fo31N_6LHp2@HO#4QYs&;eNpfG^w0<cwv&YwIsV{h9
zQ~n8ho@vBk*@}u}|1@`;-bgLqRT2SIVfpd51_!U~P=ruq{9{8D8nSO<$?;sh_-fo`
z7=o37Syn11pk1XPr=6E?8bQ*j1wKy@^MG0l8T|%4{;VvB15gq`3K$$3S`FKGOYx0a
zTXhM!7_TwXk(R#A@48XHez4V#4WXBZY?#|=%I9~o#+E=~s%|Q<DB0NNOVywlIuIoo
z8ynlOR#E*9Qbs>}{&+Pdp+d5XGRmm3BNcOaho{R0n}V3H{Am9{G9NNg(U9(rl`<<C
zUNA+H@(^oGywb%@5>E&R#USZqdPxoE5vdQPRr3!jszXPzq0b+k${B-KR|0`^x+hXr
zX@+U3=(m?q*56b6e+uD0B?v2gwIBg>=|3ETw&Ss7`X5v20!6>2Z-{DXkpOz4qNX<Q
zSn0q+wlP{r^}UZt)u@mN=8brSgl;l#n{BwFqN9WU6(eh2p1euD^7%UGRYj6YHF+2R
zV1!h9{QCZS#d`dDsju(cU+&YeBpmNoy`()*aDKp)ZCI$mN5f;a>G`qQ;I^v|E@jAb
zI5u@OyJr{hbRl~E>46&<u+;plI0jisRkaUq(^nZU2)NSRRq<0>;x}CNo6=1ON^-&)
z4C+_*t|4jk2zZ=7vt&FxcuBczpnS`49vD{%lSs1}F9`uhF!`~YL-0F;9+kY;;UE!L
z9g@fXRMJSs_J1$f$lj>F=rmze!Ez=mp~nma(|o1{U$PD!grJfcjQpa46MPfp)K?tB
z49(Cqq?bskz^tq+0$cfUrK~KdU^VM}1p$5@38(~VlL1$PHeRC3V{t{WF9HEm#lVwN
zOJa$rj^}iEqKbGrPt`LHXvwf}!&(Z)Y?3H`Zest3xbQ(f@&++YcM?BgL|YT6Gb=!G
z`00l`eP|=N1-S#b2ex+e^I^=(qs9?K+gG5`@c_9r9V6pn*@WFdF#|5Ge6w9SJ2*w8
z!+li6zXtNI)t1F*+T$fGP@Y;gM8+^#MjSjq3wV}dlwta_#Dy<p>}xaU^AO^+=J!hT
za$S7LIr??Ax&dj=$CnGQ)wmIxcdae79+6=wD=CR8Dk=u&m-pY%qn59{dlO}Fc=%It
z@-)5jq=~0)+owmE5Zs>AUYjy#lvvyccbr=w)j$EAQFn^4@M_}fKnA?)FPzo{;1I&H
zx6)_c5{fV4ecD%K+VzFk=`lcuRCfuV4pKtc2Qw8D;K568S`tSszf{GCi`{QAC~M{R
zR^-9gSZ=#~B(O)rL0=r#6N3^EG!p8coOsAQJ)Ab(a`(&Bm=q?`z|}%&X0kZVVxS2^
zTq^*1jUcl1b&3myViodW+8J&s`sN<iUsNebF8j=A)q450=E{;GP}SG}z#qT2Df_RI
zPK|<^Hre;o^2dv~H&91fl{&6|kKJu0ttv}&teRrftNwBu_pt$6oC9Byj_=7*cd{Vq
zJM6zDFM>Ku#*JA#00k~`Uj3sM>oUwnO-xLjq{f?_c9`1#rd36#&w=L!2nIUGt)D^+
zgPi{3hs2C=CKdWYCRF-k)gDg42f7JHmgN&sKL$YP0rd|{LJwS4tSKAyB1v+EAEVIA
z-Ae+=+8a=Pf{D|<F+l4^U?_4r19uiQk5rZop1!_9;662Sw9dQ#TZ~xPR{&H{%OEn7
z`}MZp!UXip@R}1saA~>5Qo!&b4%7i#uLP*5(gue0>hm@fz0`57BXDM52($eg<JMR6
zDkz5dIq8(}f<U_dmF^D6C#+rgp$LG#{;qP{#yhlA=(oVDn~CVRuyL?whO-a8CAAa^
z363Ml;<zO^2jjnZ`&s{Y?xn@WQQ<cjgL_SPJm+0F`_i;Y&Bus^1N*S9uC8l#rDr0S
zI}{u|N%CL_8iS{@`IMWRFs;A6H;KQatIK-7T0D$kvtQoM&Mp{x2iHaA@Rm}_v#iX_
zh6O}=dhQfFh}8k|(oPSpvV8NtWOo4yNLJ%Z;~@}%v$8@#aLe`C#$@xy2b(7hEtGd~
zI7uB2H$tmy(988X<I>(@+8?;U55O>Ee`Ab2s7B#OjrzE+GG>3)BzPS8{L+QOr~IP+
zO@l#|0u#R1r+={OU+Ql!-+?TU1spQLTH|$ARy7DZCqb(!u-i^`>ThxMUc$Nv1#Gz%
zz6DdvgJ#8YKQT%{C=R%R>BgX1#-NCh%%>RX=X>IHGZXV;;~Wb-e|sl>OsGelx2J15
zj;L@#{+`i&7iA$zPfrhoxldu4Gu1k`ZF<oU5>$L^d9=ef0yd}~CW)T+T52Q>i?*~F
z(r&&vUnrw>uaJETW|+1vlv&x?@cX2RM^tX#M1dZ1$jVmjN2^&e^trF9rXV6i7ymSt
z;k)%dP953DRTuLAyAu{mS$!Y|@Y(U33nJVFb58)ep-pz)9FGQxXnLFF(4-bl0e}V9
z5I>0buW|2y6%1mwKrp<aCQpF-E|GRWrB9NB#CCV$KxMss<9_oE>RdW8Z4g73PO;hd
zTYLuC{$ph1GdR{HWWV{7<RN*2BG}SVY<vne;92d(q4sd}^#_M(GhGz)Ql-J5eEx4z
zaPC(IOR--AOWh6o-7kX8ya~Fo_ye{p-_%VzakEo;yh%(<BoLf=4sq#A&vN=5A6sZ<
z5GUTAu0JcGUw~=}4hTGMC`O=T|BSB$om#UI>xQdW2EOE&{(n%pGmh=AmtqMotpaku
z(i4oDKC0;E=H$GGH&?>*k{1Sl72p?F0p29q`xV{Y-L4lus#(e71m7IT$EoB9-+_jo
zrsQzZwF(Mt>2<G?zEAxY&8yC-bax<fd)YGJgs_-j3J_FZVU1CtEo=;skNi#}8l)l=
zyMrPjX~%W(L#NrLe5TPULU`=Q0SBCnk$@<b-#>;y30i*al#9>O&$W@8oh_?Lza#Ph
z={J4c#6?nI+Q0bR(c5bV!%aqw0)MC%IY>;Wt3BRR2x?IpfK>O@s9`WG2(jU2<mH8%
zCwshw*j94vi5BphA+O?%d5TY&c*Fo&np$==^j@~MwpI2#<HTcBQa$^>mTyvL;$Wab
z#r`k$<mlDJ!w2|MRW)d0({2|5ZQfPQ2^yh}3$ZMwDn^n}kS}41M4vHMOY{sAO+)14
zY<>G*Iu0n@GY41WiMkVDAEK9{?F5}Ap1AZlPX%Y9ytxv+L;lFb4-{g!ToxRmJ!b1A
zYmsN{|9Y`^^3G?Q<+!(Qeqk|0iPd!_2KdN2zlLiyk$<Zq;MXs{r`0>6c6Ar7dB4`4
zfQ8QQtaW@s!g8x;m4|Yb1FKM{D{5<J`!!8b@EOgdJ<q)J5jx?rGA?{wtFq2cnG3h#
zu^0a^l3BpPo<G7AdJKis-vI22*-lwam6~uWimH%C_6S)^k-?8!lGqQ~(zypHywc?W
z!0Uv*a^iDl*z-p>QIMg=kqKqA_+Bsf7`MUTUirk-n#k|%qZ@TKn7)3@>11yKCcTKy
z%5p<e=C7hW%VO{432>Y7;a=6at|_}pv~sVHB^qm1)*31rSfBDyi`==1@~;6`Phh(Z
z$J#50*g71{&(3}Q(vW+m_t!@h`uOa?x~CrnGb^jeWK}8I{ZjMKTMChs!Za3szFMUE
zwYHU45_L^PZkMRGR&5p>K=2xzXG|klY%^ko9$5Joa&}hM=`U*AfTfrhQy4*_a7S#o
zNCBT3x&Y+bjtUDp3gNO-(K@jrw0~nICg?2Ts12jPhSHs0H*vb(I^qE*F?H^ui!98p
zL6$8J^6JZD(8iS<kHgVBHERm-GNaK#Y?ce0cKfj$Y<GOkR5y0ntQ=90BtDKUJl{^v
zp#K})$0`|N;ldBc2`nSPC_@Wx)GC@dN#0y0fXc#KSIOx#BP+`!;t=^?&v!4i<-J5&
z^1*Mk*I=pGqq@cCxfv;;tvnQi)UoMW?ry6FqQlz!=mvua+}yDc<_S~*a#qVPY&!?n
zzw@(9;=o_umAzMJMypK9er_Hbu#oU9=7pT^4b!%FZOU9-UT9Cnq|(9o5yVTAIR<Ue
zsfRoV;_4<zAml6H-`DCZaga={AC!pFfsAXI!sXGw<9e&#N^7_gAW;mo137l>v)t-3
zRx?Sn2&5Vr1MB?EtS9_#TQ>d1C9CQo)q3{X8DpEh>j#c!Gh5#}ang=kS<_wm<fN-I
zjf%91cm32I*Qc)9im?@FxyuhO;GA8*9{30soy%U*JqJwFgss%_RLjXyoB$da-@5Bg
z(bLm2Af>pO8J*f|urgih-y~!Ff8~AC96)GF&>$)pT7~p>Qb4PzqG~(9eF#Ajst*D|
zz+9-KqoK>RfnY5+)W-q8Q(D=s-m%vKZRXXO5Fwf6hAtcL?D8uew=<@j0S#QG_;-Tn
zn3!z$oihGGq5%ybcU_$1`Y(Mz@&WBhS_>)4YyaAZ#T)Q)7KaG0V^sn%2HJIP_Cycv
zJIqT7Fg>BosfFic9`W^HU|lN;0;<0dtG0@wV&ta{MHZsP_yl|Eqfh<Lf{XrK<s;K<
zXUkD;2<b`ZHe1`LSLC;D!b_N<Y6t8Zupbpw8bLupnls(fKrR|phG>0w!+u*xHM?IS
z_xSw36ZFUw;2LN#PJfX=V`po-sfT`<6fO9s3|$I>b?zM-IA9dF-a-fEIF*Z!6@1dg
z#2x8!`D&3|=x5KS%V)bLEKzc;?m<5}YbR}?ln8KJ6O>X?o#~Nd`#&y#YTr0ux?HOC
zHg|K~17a-Y&fNfn;WxF)wD|IGR30eB3!smyD;HM+rnutI?ChpZS6UzIWIAZ}$I`-+
z^#)^*=<)7Af!?3BzPz~jH|{;SL6&=1C8g5o_)RdSC-J_wlO&(~QN}0MFxFPwJ3IGX
z-MqY#3Cb;)9v{q<_81%ml?p;#41?!&QsJ`;U(s*(z`BBY=qX?9f3LxU5K?w*mpJs=
z*Z<@8)`92It<NQYT9a67ewHGCf@=^H!$q>s*<w`qSF;RI%aL{4SEM0OLvT;MB3;JC
z{&7|qT$Y*J%VF@xgHZ>88YX;|(_=d{JglUqCh;GVXTm<J0zlXGjDUq`zpwVt<6e@&
zzd~HJgdwa&R85WGY3yAnDOreK-1*GM?sXF}3!qKO-*E_t0oyUa1qLp}>T6ZwgkN0h
zj06k$N0ZXd?)skLLPc`6E?fcq^|Sm>R|;dVy5dW%Cnl5C1oNoE*f?T`tmHb0N=w><
zErgYpwXesq<S_OyEFJ^=0>U-`g$)KTDJxth4&h@@?YDrEQ!-8Q+B?&4Z$=5`_jIAD
z=u#9B8LFEi)mLtp7eeRgM_*@4(zK^~r;lA#ik73@#&*ZbkKKlme($||{XCj(?8fPD
z+hYSyGQpdW*ma_MI%lt{K3%!3_y}C|NnV$>%3gNm34m!qoFAVoR|y4dU}XlHS#g=e
z76UuK|JWA~`91@xY*&s8$@lESbBPTAp{BWdt5R1D{gMaYE0L7|DFJZ80`ayWK2Ns_
zy8HbDjYz5Vq3CrL)uq~Cimyd1NvXx*_Nt^_#TVJCjJ1dqq<xY<u~B*CVk5OREb;TN
zKjMVGthKvKO6b$Tt!&Qr_meQoZ0f+*4>!EMT+vQqlUsA$W;*UF2v01p6xvR3K_wLz
zmsBU6*SdbNrg{Lm0^Z1fXivsWNbke{Xb5~k1k?~)y3L9y7y(%X2#6~ocPe9&)9pf$
zhs+EFC$?G)CU6dH-nj~@szL|Og!Ktt@;6CJGpDl1T4Mz(mz4MlB>Q9^s>>u!GT$u(
z*{g;SX*Wr#I}9s+t*y*{t-2|SYvbj#)_UFvlqnub&v1Lc&QzuqnwxVBp6c12Iip73
zjoJAv=x4pB+;0PB5&DE9W1`<a)^f=tn8ij$zHVE12;Hx?wk-588^MFp|Ftd8Gu}Oa
z^siv10d(jeL;xTcplw1a3MnxodYa#tZ%P4B0KHThpLIHGLXW4KNR=KVK^i#A0I-rK
zZYy}Yh%p@Ry`SQsmT)?FkCS9+$T$yawrIt!ZEaGOfdB7pmZP_Ax6Z`UO{;hg|3uOA
zq%%wG<@wt;sn_kagl#MS+Hqn#BXnraja_K(u^TQ;`^J);(YJu)g>~h-JNCJD8<IVE
zCxY;EAaiKLsPIdzp{>!}tz3+Xk#!m(jJ0P<+FUFIx&L>FTDGFsA<{jK4FfA2xco5j
zf#gAl*9d@IgVGc3lSys=bj+*?5;A1z0G~}y2$G43WPpAq=q{giteEX4+4JAX=gsf&
zv3f^9rYpE|^}sM@q&%O=Z0Smfs>+=3sn@)xARUTG&`?;Nd5^hpfzyI3(xV0z394$B
zUOZ1Sk9z@6`Ppk<*-|ODP0~zg?B$Q?xUH(b7eX|n^;u>^gKKF?5AAZ2nv!;cRxHe(
zW>>&Ql02>d9n1^^(!-}mUHxeD%dThIe2J-aFehLnxC>bGst|k&+=w}cdm+#FAJE^s
z1(S(=gAN$uAx2}ywfzC-)Yg4(==K3sN+L}$#i)XtUElNd_lc2LvzUZsR!It@>np|9
z8TS!rEEBn3Hn_SiUAyJ^V>tL}w*79`v#KXHyk+)l+;TMPlQJH2x;s|rdN)IXShnGt
zj~nODM)U|~gMm=@mdXWY{vhFL&xdNkAM^?Q6iip;d#ioO?5D2o%v|lEiZ-wWPjS4~
zYomzy{Z?kI@%ypR#p{eKggsQk!}BXcIPk!sSB3c?zjF;~v+S?nCc|zsrzmmh5=>(s
zAxePmAZnJJ6hDpsU%+&(fZ~6n#xaRN<hloLP=X3FfCf@1x(qlHwODgO7z>s&l<7u+
zF)ap~n38{4IZ`IH#BG-+EH@&EWllo%|8niu6QD{rtnm@(dJ~FCCU?#*$TzF2hs`Jl
z&ul)N-782lq3V|kyn)U8KGeSRR?wS3v4>AwH$-51{Oixj`Pk9#U7bRU&tHWvh4jvM
zI4c~ZMo3Mad*-t%!%mi>w&&41t^X>0X;XK*LN6ZZeXX7-mmevn$!#pubQj&R$uD`U
z;0SOtL8-N$8!s3Mu4k|;MDe9qopTN_tVFmAKq3cFEg0STjNyNU)whzG2!J<>(Y>`E
z07R;&dflt=kUa(*o%r!PbP}zXrd$aLWDZc<s;F)jJoAu~!vrs{(*vm1gk177pLt41
zpaCrt?oBYv2EnF~*kzpF<VDJxH0uFtufZ>!`T4aGZce<UGrEn)!Q|3*ebXwo6K{sL
z9#Ca4RMvPJm+~K*^vYrg?caK=<C05Z;afM)py#nAomaJb)t_OZzHG4+n(w~!J&x?v
zqN2^{4ny>J{(Yp?j0m-&#U*THIwGAHv5f+6BLtE4b7RXXzVv)&&oe(c?>jz;yFy$T
z7ZVFQQb*-i|7sc;hnWvp7C#4nz+jyTm_ewn$9o9^THRnEb?1`}*Tgg>-T3qW@~1N$
zlh@#Q&~8Wr&PIG5Ay%*=o;a*D`JItZGhNc)=X8gI!%g?46J2N{=<GZC`bP4zH-l9)
zG^8MXT(_8A$)p%)&<z||`HdIqkK|Jwy?f?`8Ez!=dFkjXHYX?VGvnze#SC4Me*b{m
zs5b&9k-hodl*o4>FYS0*Ar}xaR_`x5tEh^rs;cE*zOJs`@-DG3JogQYe&5X)<Beb?
z46l1XCrRGb`eEJUvXXe8H(gjkpBK|VDvvkHSCU-nMab?n)m2zsOd-+Z;*dDptIQ+T
zHp7)z`xvM0eo3cg%HG-euBRDbB1`YpbMzSMq5nY3Dz*1}XBW>OMGA5S+hXOu3xSzK
zKu)igaB1Dg|44A;(9F@`>q?<A?ZEg2I1j0Gp1;%SLX$182hDIU^c3a2-UFO<RMZJI
zFkBZvIre8|wM!_&T436j4s&#fnu+om$*2jO=3Mes0o&?V&5hVIE5H3-*d3!}$&5+r
z<NnM^bM9v=sydp@P?%Zx%o@vP(avJyimsM$BVj6h$gIuao$w5A--|SDJ%zm54T73X
z-h9C>E8p4ea>3L(`Pcpt30t-C=hHzL!fP5xp<!o{2?-hKl77CIY8b7@*47*eqrZN<
z&v@!}%Rl*s=Yqrdu^-i2#0VAE(&ULfyT_=ZiH2<=`!*(*_myD!_g5n`m%ZSqt8|6N
z@!;Mq@WYk>tKX^(LA;RXFtsH5(3+m*!UTx@^&5m~<5jhDYisA(-Tz%PLnw)|vdXmM
zOd+Bu6b_($?6*+q8>GH`u@M#X4S9dI!-AvC6L;jt^(2ul^peD41TeG;ou`0e?$f`~
z%Z~*ciQU;WY-N^9lRu}+Y#WK%EF5_4nNMbfg$QLwrA{KUSK}wbF!i6V9OL%3l5y<}
z{OP;(76pN!MCYqK<X*NGn^ODD{Ajw{eGORzDEx)HZTUPF^$&B-5u{j@Y#0Fz^}1IT
zxBZ{;R^E<PrJt&}kHLb-OxT~jHJe5|^XSL(4&jP-RF@8Tq<UwGVxIepsh2*t=@lK5
zuzEcXFjEg2QqA`W&=!{+5aNW2IvVv$ey6w@$(~wO3a|G(Yiy>g<G{=Ga~ZI5cH9M!
zpe^S@orTI6`xDdw1xK9TsV@I37J}cIKbQ}~=W`7Gcnu2s=Vm>Da!86A@MORpXr@Y5
z-2kcQ1IC07u&U{GyYp370bu`+YZFyO7@ArDiQLCQS4HLd^iIQQ(V@;{-CkAcP3tu}
zkM+B3RnDd>r%IlL*s~dZ!V59llo>z&iU=wXe#xc{pG~eiC#_;i4$98UHd%_$XGJ7-
zR39%7?62y+pW1>M%5M8iis4-FU5%<!b2JRTcM0~#xJXpt<RIazZXqsD#IE0qqNG~9
zYveNOFlPI6+oz4z{36m)KhI#)$>}F%Q2VP<-prr==<)(F0gg@@1+F*l`cBoa#OgE?
z=hyaHvnS7{U<<UT&?+o9zFHWkX-WfR5!%LRFv|R%BL`Fri0Z4(Khp0cjoH8$QWpEu
z_+HoBi@5)qj=%S`4q;_V$-0O-Ob<vdTBi+I?uE4g$KeAO3rw`RCmn7yy!_~QzCbDB
zEe9^)T$1E2c-%AI>=4p1h05q5q5J7PxcB=k62|Z{*x^VLWWLn;mpcq6v}BRa)4wVy
z=Pxq@fA#RUclVl3?5b9<6`jm2MDRGbzw$nxo^<~~x2kyHM_g}IYFyq@e9zvGgD{~-
zzOM20XS)fh7>0!EL^~aaHJWO~mVnjNP{kj1DW@M!%k(R(3qgK!dmhNgBWVgZ9Qt&X
ze?FP$AXV0Xh=aj?sUP*lr%O4}#F1^Y#Ms?yO5*b1pd7K{K0y74kw9ovrSptbhKGea
zxW^9L{^D1%M_qH*TT)Ckw3VgZ{rxugjeg0Bzm0b*4bEEt;!kVI_i_R*7R(_2f(>B8
z#GheuO~x3&h(IjRM{P;-mqtp4>3_HOVZ8P@KmPU3tH*Mu?3KDzzsv1~p~C3xVy;R2
zTp22nCTAA(UV<u4Q`31hrYFR&CKP+i4@Wy9S><A%efz0ab7K&q5%qD@?W(3{hR={;
zWg1f(xtmUu4{&e>Zahls|7o6$Q>pXx$&)Ahb;0DmHz}Fi@5}R`xOH%emQu*fT-`|%
zmigJ*^wC(}U@BIh{qIM^^Ya6x9}2osZ>^c}w^SfoKWAB7SxKo~4tz*^9|Jw^tUoM0
z!<=QlU;&wKLo;w-es=vpn0+apAK8H~5kq)G!>ooQY9XBRJ9_F$qSHhB7YVY|L}vTZ
z<~v<})5x=VY2k}=`E^TqeKD`hu8h>w)uPl=6nTpz-Fp2l$hS1A@%*KOu^$dtEz~YV
z{NTcy{}Q{ryO;8G!Owd5)wcDXX0^rrmK*4=+G>t~%@X@K#n>8Vq%fb>#y__zxhMui
zuUN*j{PRue|BZ}9ORAu9hGT74UhefU5F|(8>v?rC{K;eyGq{UsBkH>i=)%;fVm+A#
z=)beg6^896x@oU`UkBu!;A*yrBlsWPpC<CU%{!*4qOQ7dbf#JC$`nPr{UjNNi>t%A
zq~P;MagA^S3R_ZAlK$ak=|zN!s$AzI-XwX?qb)he*D#H_yf|B!Du^zMsQqR$$M%SV
z^6}8SjMUJaM4gzLLlU#7pxkv|Q-c{XkJBKpgEEzRI~Rj;{(`Z@QCln<S6bI>n0Y@-
zd4s%Y;_}@qR&O-NwKd5bHKNv2xF(z|xBS&5J>lN!2&IDv;=#q}*GqI{ht5^=wKgnD
zzQnQOWU74ir-Httja_~`ACe=4yAo4I97fE$+_xK@jt!KF@oSI2PLpE|7@x8udB5sF
zB^yg4m0@hv%&$SM$6msB7d8APxNCCG$w?B@YV`SXpNjSaMNfek_R=SY{KYTE5ptGl
zj$k|oJ!4UE0u~!7&IgHu4SjI%s`8npk+Ks0ANDupVsiL*QCY&NqbF^3oGQwZrgc18
zTeI+dLJd{w20GT_8D;It*X+Ded@^MFh$@tX7Ex1in(HGCf~(o)@|ltCe2ucePwM7{
zKI>|r+Utdy5V_*&@~jfv!4LLIkTSp7r~w`Vp9GZV=f1+1p(JHK1&RCJjK466L_0dS
zc}@~<pI11uUW8)*Fq$Ap_WY}xv%FaIU4>JE2j|f}D2`5UzePC^UML^iqcr}FET&oR
z!HIq(`2G?#pg=onCPTxAHJCZpua~t+5@lyN^~8C7Jm6@?C(NlW9OlS-%~XiYcdzeY
z_vv_}n0M<U)qkM$X{ABOo^NiPo~+v^-Tec%T99mK@NT*$<ME^W;T!j{$$BKHWXo(?
zgqSYl<%GO%F?6{V9nTjc>6~l@Cr;<Tqw?CHJ&s-3{Zn_@F?hDMkX(21Ax-kV#qNbo
z2>3Hy6P()HN6$JAN=j|VeI++*Zs)?;7yCbyCU>$7=150Y*9&*31-n{j=fs<hc6ONd
zR+IGWnaU>{!IiZVR$~FZqX+9FOvf4~E1gbG8-jTcf-*9Xw<K#*5FD)H4Z4*3f{13^
z?k}tX4lbO-iDXE>!g*pQ=Ol}lGp`If*QF!&o|>NUwjsc2dF%dzXS>dI%{JFg(G{kh
z`ihV4Om}KEMh~S|ve9e)O6<;gS}nX%hY$=)Wkn+(Z(LxW3YjA5!x^)c-#CbSRMnF?
z<LR<4mZFFL*GJ09tiWPSAf_Ulw_@86Dbc;z(!1i^wwHpFAfQ1cDT5%Yzj=fq#^Ly3
zlf!6|ca0;xx<W;Gbaw6B6l%U=sU@y<nVs3(=z3i3yqMj&0c)6SInqbdV#Ee9EpmZ#
z7uU4F6(g4W8fy63P=Ern5M^6Q+rcH-?u=$&8fzZU;)7%X7d)g9pB?%-wGji(xKS1D
zwm|Ob8}(Lf2wO|yrado*h0+l~alSaLeoy3gwFyH>c>CX!0p{j5t{!&f6LK6REprae
z8V_-9e^UBC{46`Xn_|jqM(}-&eAlC9A|X#CU;QsMU9r0$5W=~2vjTiP6UZL`77wh|
zlNztRX&QBCw*F41p=bM@&}2RVGraj=ukHEv)vFMFUYxm%;EYJ#^P7$m$v;xc&qVlH
zMc%j@zJqQxGC5fuHZ7gN?`?NAjZ9JZXEl3ZQhtB`jQ7>x$@%GmqZI}h+otB^)k+)@
z#=Se}w*%LU#pe8zIsb&KR=Mr82KPC&aYif@d`D(E_BX7q+voHh-w~|!59SCC6T>3v
z{6i@;w@w-BF!-=-PNwUtg0FrriHFd(?=jj+5-)}SuQi!B(b1pnXIiUS)!F~9dhvXl
zMZ((F6KQy_eXyeGxfzZ0xpG0duQ!=8MYQ~Qe($8RNJh9%kCM)GG$&C_=O?9`79&O+
z?_Gc6T`$Mg*=H~KY;Fa+^)2ib<(_+;$UL!6?xwM?Z9~uRooiW1UdVCkT|TAfE~lqc
zSB(aI7se8YrzZ!%T~mAu5n4^Zn9~*e6JP4B&TY*zc@!WcfoOn<<O2@yp%Z%O>76F8
zXPvWqUaP`9l<x>dh@6IAsi+-S?CblrT5vc{*p{zeys1H#if4j_4GEQY6ThNizJmlN
zo3cRWBIdNhoBkw+)ZF+`O>qCg7;SPg9@)37(*=$ybg2}dL*j*}WNnrii4sS4Q*msj
zce-jdTS>R;!Y4KjH;=jyLqmxEMs<xUGnz;de`M5PmaC{ZMjf^U=34z`1yZ#F)0M~G
z$lPjS(Y`g<i(eknvY)mWpDYwE@cIo#S6?>xMV-Ft(Bz(VNZ*^UQb7KgqAtZju(V(r
zc1uz;oqyn&DlyxvgX<RK%BI^gvP0_Hd|08Ig9Ku3@~CBEzvDe27hPtyXh@60m!AE;
zjz+aBf#ABGEWN~s4(l#R@IC5y=B8^6inGJ(4#AC!!|2pkmu55(OjqH~*{E0g`AZ9k
z71p`0&#J2oaWCIBhGqL2_P@nRymEh!YCT+7m=%9X{8R``KHg%IcfvNC`3%+U_yZyC
z>uDeKhIDvV`-wnZWFfLae6Cjc^M)GAmslji|6yYNXh2d1FPh%1R-tUFVo9J)JkDzV
z6wai(&+ZdJ(mhWooUVOjr;&>|;03}%K*R4J_T%8=-6Kw@`yA*<z-D=_YW&<sNKNhx
zT~}T(ILEMU8}Vg|@WT4;aauo}B-yK*u&!nP-9yPg+;zey9yPqS&8uXSZnw(cSAD2{
zZ|B^%*@7H|Zv2jz=+&$RMTO(zG_({zmEJ0U9EioReq8u6u(NYBXQCw3-dCsl0(O}?
zn2bnNNMiB5W}4tGOmi&u&C!na4kvOZtNVQ3tS;;m-e53WfZRm4q1MbFMrcqv%0evX
zz0;x{Xk|@zc7ndE%d5k&L)R^V%T4ZqmJ*Y<gzbY~+@sy~E|y|(O)P@;E8PoX0%p0(
zgEI2Ln*oqdE?v+2E%MK`*SQ=WM%+1njnhXRECesQERACeErqwpR(iL;#xtp48Lrb0
zcHQnAH0|o}Sd^Zg&}giaOsSV^!fLx3Ss(SBO8JR(tm(KKRP8*x%BWowh;Co^A*?M~
z0VWdJ)9V%-ogkjO72M=CNq8q{=foS46%6=te2p~5Cu$KZ!GFc=JGJAbSbZWsXFB<e
zD8RYb+R(Ac+Cf{GFx3l_>sHuG@W~Erqh<Ld74h`k=TKppTN~$yPSYnE*-u;7gsMuQ
zTvZ%RdmZ@SGyp^87s&!15wC%Th7Y7F%Om8^K(Q5l2*oj=)Vleo2BlhnNZ-6bJhNfB
zx17+&^>LaeN}u4WKJ|!T^#ZW^++<IezjMJ5Lyy5d^zT7i@xC=M%wlwek`tqPshupE
zT2NrMSHgaIgTZfEpDS>mMe6oGA+9|>WrsOkSl!8I2cB-f%cGl@UsD|D+DB%5U(9UX
z;I;Z4Vl!QNTmN@Jk4=W8yimX$tffG)-r0@&xD^hIDHZ3RNr{}gzgJ(b-jyDyp2t>p
zU&>vT_Tk<B%+TTNVz{gwv9}qY&$E7fSEKfnVK_E#X)|I)QC<19u{N*%`<j%6J?utP
zm)VHs;wSQ9ACgyfkIlSp-$<?v!@Y~Xv-8(oF$g_|NfK55%K=H-;?d37;k36XJH4mw
z!<)G-q?*oiL>p$fr3Z2GnO(Jp%SXb6pdm6DxA9!IbW!E>h!Faam1Pqz5NSK2+mqge
zzNdR4uQEHftzKA`_4Z9$is|{AAf6txGF7$tT`yv%!?E`#O~ln)ZckxgA--MpZ`S&i
zF4UJ1Cj@Bp&<Ad673R_Kv4<Ua^`C5~0{Dv{Ee0LMTMS}R7|&`J7(Dy))dGqLvRBRb
zZ6+%UHEnHy7zmrRiVPZ1E?gT>38|CVeNLT$<_Olm^I&0a28(BMGzwAP`4!iICA3*@
zVh3zg`nlbcHl?ni(GKfUqQ;d~OB-t3l;kFe%F+n3Z%7A>L=c|W`$q;NT7~tNVE2I`
ztTPtz?!@Ep^VCg+Dfdp3+l(^D7LNS!%*1EE^ZUech=#x5iv65}h}!x%IgV;ApkLdr
z(lnw9W+&I|@W>&Tq4bGAC!n?#{UI3k%_sPk#dZ@9%ay?&o7Kcu%_nr<SnyRRG^$q7
zV{%X5xyAGt4x>lxT_q>I!<phd7WQqo%o*Dkj4tH|Qdkv!R6_p;$`S8e%HoN{MBi@I
z_d;6&$vz9LdRNX|=!BStVq;odd0GwHE7L7myC1%Mcx*>5M4(UIy<(cZpn#2UPu;$|
znec^Yk!YIliqF-33dv;b&%q;%sXvK{FF)q%?^TZw{^6&pRruZBc#McLuM6)nnUAzr
zbDTj)IvLnu$(Gxs*}FcTv6*Nvyz&W}6I>vzJd}#aKVT)n@K!9tZ!1|nKZxp!d$mR<
zv3S&c1VO=}=Xon7$43M@mdQbS*X@xZL<a-2)1TY41qPSrW;NUKz@7uzEks^%MTe?|
zK31TtO)(6Y29DWG)4nO4dE&8jtvMt31@-zh4etxPITrMX@_$6m8qg^hVW*OGmdfyT
z=KT#`h!Wdw|9@;gbbc6o!UzGpfYGnwo%#f0p$!D83#nDDE|B~IIj=&yo&}#eP1pMH
zQ<O(&GkeC2-@&fC)REWqV9LsJ++*Ta*3-i(@xD+@g*8Xy0sT__$G5-asIj4AHtMdJ
zT?d;uHdOyoH{F!;2b?gsy1*3)h~vnmL?_3aQUAG9A8Ej{b8)(@QL!O+{*02i?1eo4
zdBl5SE&}`do(9xlr}Ksk$HN`OdrIDrZ27pxQorhiy5zG58I-<4_Do6N%k5c+zPOL$
z>3Og8r>}<#A26S92gE%$u6nby5Frq7K&tj?Erbv;q>xhVeGt3f)sZo&F6&Ys&}3xX
z-)i4cz)S2lds|><iFER0EZgMr?9Q;Nz4_$pwtHEAK5nOiJlzCpsx~*insstk*SZRh
zU<%^-J6#@eAtg*v)YhJ-9?PQUWc{SZ`@lxc96PbbF2_(7>(u}H2Cwh1wGPi;BmxDZ
zs`Vury~&4JS(B_rL=D>c3!P$E^96Rgnx(~`u(_jmJNvAUyHX=oxfklvHM?^ri$36b
zx~xQ6VzU~Jviy*%NdZ{V%m)oXNr2ISpbe~jpr=fYjD&=jZU_FlF1hTRVb8>Yk1dHK
zTzUuCn8}fytM7La4V#hnAafX%PZ0CFGG-(?hSr9ED=GO*A?W$n+8hZi3V`BUqepW3
zTMXOu{1~ynHI)GCm$);1fP{E0!58%60`{#<5p!tj#(jfQH7|RN^}UmdlVxiUkKt@U
zi(8Y0W4fMJy%7=*LU2BFieW#0qlkw^Z?+iNFW_-NrTph?`v159lP)8=0w7(pz4H+!
z53G-?2@jKN43^3n^=lJ>U};dO;c#`F>w6XI6fTzQCzt!-1^eJB0xWVr#>PqxYPbES
zVxFC_Uq&deMvx6%5rtMAE=;&|Ay9qj`fv5s9x0p~GGBg-aNy|v%<4x=qI!$L&~q8D
zU^XeX<3(SCRUxioV<6Q-S2f$EI8^N4G-;&wY2K#>77)ES-$V%aXImrRrxB&bd9=8<
zv#y|DxY)pq2)^h(Jx0Aen=3vyoUS38PNR&i*<%T)A}BOE<5N*#l)&v`_PbZy&>W!8
zRBu#LVB~dM-+n?$BxIIF`|sKd^YK_j&Wg0po#Q<(%3=$#+p^3xAC2|myB>NVEVBYq
zd^G)ZYkx6Pa9k7=qxd_mG0fDsFK6bAQeeBP{w`LY8b~Dj!{jz%PK{$YdKbZrbLl*4
zkVwF8eze`yU*ox>>w#MCdx}bW<%nPIbIwayeI=CQvwt6*%$el~wr%TW{Dd0}p&w|o
zYadtBavxfM)cUVqN)5W6t_5JH^?%_Q%4P+jn8<_r!ZwPo#f~U|Fkd9`MMGG37&DDU
zeyOJ?4;G+*7D{j$zrDPbG0;(9*aYE5K5e9<r>Dh&qX_H>U}8WbxDCHxA}tBz0AQD<
z<UdTdshS9EZ9R`A8Va9{XP*1-Uu~&U@$5Fgo^)9adlBb&S1g#K<L##KOI6~<=7%i^
z^tcxo0=_KUwMS2XmtiQK7giC>+3Ow?xK<_UtL6Uyql^yP;Pan_L=6T79nVUJu{V9D
z%1rT3rqEuk^vHfW-Cj7hngX=V`>l?-_m1#(Hm}M^^Z-*Dzel)^Og7hU_XIWz&(*yR
z0)DmP5BJ3-_Yfm{yjSgE1qtM%e~SJJ&dZqjUg8l(RAMKzydAP>VzWUhujLD(r`1=`
zlb}!w!>?-TwJo+Y?1>q|*kTW?OZ2$m$GeBN73YF(?_1*^@axFf<>aw%_37>9i=se6
zHJ^6>#N?kpb22A8i7&8Ln#ZnCzGg6lX)t!qR@-a|m=R(7Rwjl`y9bMt%AV}c?|ImZ
z-4@%O?^|E^)RU8+X(}SmB_F2A-0rX+=CnFruy-}azy0@T=Yg%hGJ6!?YRyMc(Ys84
zR79SKKDhPn(F3|ChmMS+{Z04prjmXC`SXDkdgNpBRT^heX{m=bvC@y@I`%IvSXu6g
z-IhTdCHdq=mo`?-Tx_{*6AzEOC9hPbbb3;??)AD+ZVHLK$&XMbBc`nmwCZ$5={vge
zu)0tqNR(-rQI)$4Q$Tho_!&4vtQqL#xpUrXOUA&|=Cy{AS}jm`8-+EQLM2fy&HQMq
z#4@tpcp-Uldb%dx;uj#e08*by9+H8;JyqnC^c}q`EWL!Ukq+2s1sjVf^W3iu;xWj0
zcH-CRqGe#wh#1f$BBiKdT_OS&<sra~lI!RFQzzRdeGh~OEI7LE!t?lO1q2dd_YK!E
zvShrzoSa-#*m$lQgIYFzkmxHh(k2RC$1fCuTLx$0tJO!oPDMp7ll|GKj^7g-6kLse
z&!Q9Y7(B@MRLrW4TCqqxed6E-ZIOQ+?Z>Ad3TT9{IFvF!0~y=6d`wEG0J+YKtZaI0
z;Ie7%e)vp2zF|c{#9cHj*D8EUZok9LtGL75E%dJg{JBJedk<TExN)LI#lyZ5bbBv)
z)sh@y%BE<cy7Dtm@3H02aKDb-@(7X+5Pmqx|EA2)miQ0i;g*88|2J|<BdOrv_WbTN
zT@!Szw`Gr>{^YgC^o|v;T26a4Fq`Ep{v>j@^=tX@ZaF<>tu-!X((|r%uJ^>}C_PQM
z7H11#j65|-jsfAWPLJ;i?eQvg?rFsQirAWCB8vH)%vKYw;OV+F^6f^V;76|HuB8oj
z<5d!p7j_@Qx8}Z)F*BLh7pcBjn~pA4)gr*!!mdBfA=`K(Vp<b9cHzZ;RK!UeA;KNZ
zQfWYl^^s^+A%%uP)8R?<ZQW0|4k||7vMM9E^&eBx98OeLFhu8&3w>GPqyBW*t<fvQ
z9+P$I8G~7wB4|wU;{ksH`QGx5NL?)%aCbsebV1w<<OlT&Z+lL)?4dChwZ;60cc{Tm
z^8WqEbrYx{ssWP}0Qm(BSMb4Y1Q^$vpu+%|(*%YD-j^pUewMcS`oU9(pkxNvV$?ex
zcUQ;ariWlZF(DzL=ix?2NA`Yz2OP2!*p*0nF7;jMvkVPESWtXCi8J36XPPlLgXjLZ
zbswVJ2^qtS?#<`VGV|l&kJNdZTM&YpC6)tkVb5ZL^@v^f(t-EIXiJ~SYF4XVxzpAp
zFYteqGWVIsSg|RE$P!?m;M-KuY7P>xsktH2ub{77xq?lNa~pqx`^|%$wd{O7*ChBy
zSGL*W-uwF_fgbl|rht?TS@*~3?R$T5VMW-Z%mV}XfwMnu>`sdc&L3)`n9RBO`eJ35
zsz>+qy=zRZ0<5ctu6%^_ucU88?j7qY6bX3HyR>SpXvB`%sSH_4Gs`%Iqk5wL;VolW
z$I7jACyz^9Ijd{$*uaZX(30Ejrg|<({q@xWwZ3`-x|{Ix;k^O+tbtET$+BAFLd-M$
z;TX1@xxQZ%y2kcyySk<QT|zuh<}Ulmp3qMs=NwlaxWZqu^r06^ks=q-`1lsag$vV`
zjHFNB7cM5GFK)XSNc?+uc+Q)ps6{fTAKpsVD38<i2wJQ=#ye1BC<2^xovrxxsae^=
zSly+7HU+uNw%Pqex@?hp!`&U}KK{6(t9!OL&NX*uFVgV|iq9L6y^mYh8Fk$Yf2{YZ
zsH*a|9BoS6yJ=-jo{~Gtc$iQ1|BgxX#6$wDTP)FHI+}8$gq#5(fz4`=t~XgwRz~KU
zKRBYZ0^7Z=j}m4pJop{h*%KND3&EkfB6<E9a+uVM7u8#YUtwW&f@xQb?vZ{hkSGzm
zyP{8HL39R6$RA~8heh)4p#1~M_lw6>TP4-j52Iy%TadlQw)X4$u6<R8L^c^Cp&I#m
zMlBEJD|qGO5#SlI5oJEfx+#*M+hSLYOG?_cF;;x#s5`j78vZ#iRgUzf`&nvk)+0W?
zc-R~iAG%3;pV5rVetM$-r+f^2hO`3qxa}t2sXP%+>*qORfi>h5&+b_e`(eLQ)RpQP
zZ6QseDWW<tzqj#st<&T3+cpHulG6vFw|#`aR+r(`Xm0Bq<&9F{sS`$N#%z(@J-qYH
z_&LtHLXJx~y~&;AniX*~J4+3*qv3kD6Yrr6%;v$Un$}{6GZc&;VKcE++Ag$p94SsM
zD?5mN)5Aay#)~5>BT*v{^E0D(zq31L3HWJ4FLKA4)jK5PTz76o(3XiFweD5_E@a|&
zB)^D#uSUwhcURM$?*p;V)?D`V6n#Jc)@p22W`z&y4S&yU%#)KoS)}B#f)7aC>VuAz
z#ePb$9<n<xpJ*mUw)EXNo8Qhdle9x-^7he4j%g5M7~w@74*F4wGpiv@Jnk9%T6+8E
zEw-$z!n-I_jiN;TMZDT<#XP$kxHfvfKI6;A>LI)oVfJrox`ZDw^JL#s5J;xcGBkJ6
z>t54+gncXUb%_ENXwW=IuB}<gCku4IHe|XdPr#R~$&%KG33k-W0cEBwf>QF{P2f-;
zL&hP3Jr#RP5^4RvVb>H0uiicY@#P>3Sbf2=HJ>T37%6H20Qc49u&Z(I1fhJ8SSVrV
zihE&EIxcbl`&~N^?xsrl$FauccNdq}UV3@VetY$<Jqo?ZV{43%BqSLWT}2+lR!E=J
zw^uGZ^6!88QctLyahW?P9P(VWcbBHQn)9|luha7NzxP8vDO4-B*62>U=(n$aev)e%
zVSyChm(pA~A%U>I^g*Q5@yLQK(e#y=2^((&_UWfSJ-&L<wAlLuAM2-a8IG4`@TI76
zwvH8KCqLW0m^zmZIGEYuj_eyzuyeRT2<jP3PxBHnTjEV*i`y5;6t`!#yM_my1P(kY
z;nY^<;&0Qzd-{I;0I!(acVd>$9@8f{Q<kisH=(DqC`q<ZwxLTbV0;KaT1i^;<%6xK
zGFFePoBb;+-c0E#cA8+=Q8aBB(|S0xXmRNGG+LGK%}wdLXX#oa^{S}w9@e}fwad%1
zDAlrZ=Cg6-SB{F{kt6HeN5c|C&>@*6e3<7GEHz`lFiySx%x4{%s)Qehwhh}Xe&{7X
z$(yO^^}6RzMPn*Oz4K>z)4S~Io^TRRP+fNXMNkV;b$=z*-3Z=CdFztc=KT@EI-Dr5
z_3ed=%K@Ab8$c+aBaVBqoMVoAV*L(9`rjd0w4YsC5{ET%tG~ZBsc!ya4wm_D1R?%f
zhY1UU5TFto$WWlX<|7(RL;R2h-qp2A(7x`1OF{4e5B$*mEbCymXGK+&zSlhd%O}wm
z93%n)0uXpWNDs<4prj-iZQ7P<We)Q(R;lOqemnb$)oaXWTT|?tVY(g@C>gldL7Vb1
zI$FFKtA~~N1)dom?w`lW$F*+C%F27?)@kW^6cugd&z6^aL2MPGl-b<W6mWX#39l(Y
z<%wy_2?am6_5)K0usy_|o1gB@gway!`zw80&ZcjnZgw?4q@_tW-!`9%cl+yew5O44
zwo*z_xh=-Fi)~aGf9zqz3Wtis7_*Aq^`m<9dLrQ}uDFee)1}nZJee}R%KObj>f-V~
ztAeZYZNCV5Ymz%TE1q@D5ff&{d^6|xUlojm<(CrE3ySJ_3GheXi{y)_ZRs_%D@Ya`
z)<3z^b@%V&cU-TdP1EhyH!w)AFwnl=naO15)m3cSBQ3|hni9bAI}Y**YH_cuc)9<_
z9ZQ%s6!-g5Z)w{_ZBW6i0nlq*JOxQUpmB3^Z*5K`9Jy^#DQov!7Dgc5ThB$V->#eC
z@4K=cy@9mR?y}*wxQ5y6X?<KF$(^wilFUhX@OgD#qkt^#d8pFkw4fC1JDKkVbeWpe
z^A=v4X7lQ(bnX!JjKxMp$UI#ck$k%|CP}GRiUziGzvJdz84Aa2peKJADYdh+b1zNO
z54XQX{@<ZVzYVo)fpo8PY!NKQv+li(fw7odKK*Nr?7LD&ai35Z{PEHqN#VAiHf_sy
zsrfNGKmR5y?5<fcG5C747`=+;nf#l~0(HaW-ooD@S+SZ18R?+7-xk4pntEo$G$*Q)
z(5ff$yM^x;ua<1jg!KLJecNx4zBijkoiF=1Pr>b@)Vr<m@J+6X862stt*yZ&UJ_(l
z-}28<ec|<~`PWe1vq#bQ`&Wf_KXJ2yoR1nOq*z_m-9(at<N(v}{1TOIti^nB3BURE
z0t1h`5q_T_S4!5JHx@N>y7D?Y^2^E!+^->vu?O2qN^vc?*xoIZrTLMH$-Rr1l!%L;
zmA$2U8;B<g-IUTt$H{ai;dpIl69V-MPLfJ0y3gM|zUlwPpGUM>bS$YXstVT=32IwY
zlaaPBdu`YPBS+%(_n8t{j&*;WpYrNQq=Y}X#W+^;ZgJn_Fn#k|uZC<*Wiq{T#_yV{
zv}buLPnH@Uin%+q-ve2@hkav0^Ap|nrwryJqxT2n8NDm4nT4$I;XhtZKE3JcfU!36
zeq!5ahH2)tun`oBtuVABbGcJq^R!X%i~*Jjo8}XXTNTZ6c{6`zZaKJo?>3kn9EdQ)
zy?M&0cUvT){blW@7xR-!jG@n~0yC_U{BOwDhqQCg9iK!?PgU)<mg$b(Z_nxPXMY@O
zL1{mJW+5fy;Qv&&O7F1h;Vl%{Me<x+eCiLKJa8G_Jeb|ux{R>WIP=42eM9}PANq3>
zl-(B7)x4k#r=_O;Cx7%gb5(hQ4?TH0h#UPE#O-EkeL%9gcW8G^@akbIxB@Pn1MSu5
zS@uxs$hy6h{&(gdzHj>7KPA_^(WrYEYj=T*{4<-2W$&VtQM3BG^1M9A_nRZ#Qb;kU
zoc)fvy|diuM<qp_z9~zUwSMw$;3T%~$7Qr1s8_22V!Rm?$YcE$w6V8|j_)mpvi15L
z3Lzo{!Xz~~@@k(oPYdE@w#Xj^(>{u(zjqUvvQFwz>L;ShQO0T?s>}iI996|E!O^lK
z$<~O(fX<erS%3FCU448J6PsK*LnNOR@{LTcIwqqyVpi>cpcEkX`7WiZh*%w+Uaic9
zQ^ggp7?FH`{%9`OcbnDpVi9e`=X>9J_v9|7+L`#AaK*yC!;~pW%N}v#kH;Tx3aC=p
z@D~P|n+bOG`E@shMu(2yL3wjy^p|b6GTB8Oh}DX{hpZm^TwosUZ#?VMk3kHc_ACw$
z;#_Q!d1v^@%4njV-3^2ERDC?@WWf=SX6?=A!R8o;k@;LywI*^So`f&xX5p&N(*@s|
z#)zY%$7xyd;wwqpPrvYdCdSAiJ4A{bx0<Z|@>=em##MVjH~DY$=<`$&yM^uT^C{Jb
zZ5E|}gnp;kQUxQP8!-?h^N5EJNAewcu7ulY)|?g{c_Kf%*WT0P5+6Nj1H}rou!MxM
z3FDtn{BA2g2zht?#qk=9H4(5DbqD@909}R+mRv&$(I~aTj{+hZK(MADz5VOVcTJap
zjmCCIG4}L?k7(`$HEISc>+37P%!e-tt9Z<NQ~YT_i_u(_eO;4m+Sl4?(cN#3?+?_>
z6tw#4QuNxZwyV2<LWx8Kz#%8iINFLi*$b+4W7724>U#`>Pbfg5pRMNxE9#(|0?}dp
zqJq%3lJvk~+-b2#k+ajOn8+7zKkhA2i&jla4k1hy9SP6VCJb|)P-Tf@g9G>MNyvn|
ziz6G(*R~qwGz1oE{ZtcnQtv~NA}1_G4fjy9`_goHS$lm-p)c~9ZCF$DN7#HiG(>M`
zJE|wl*^ICJMBjD9uGigM_{oqUjH7!Yu3r6ZlhSNtY*S@W_l2%C@gI`X$!M{$j$T%U
zQo?|7Y}2{vlU~-4=Z&N{rp?yY`Vx&N6s{T~bDF$=og1^svi;1>|6Y;$&ODj8Z=NPB
zVGx~8NGQ3lF21!vEcowOIpyPeG=Z^N%pv>Q_Kw{<<7Jjj`<j@_$ZlJc9bS~)^=EFb
zjXTR>0z-M6N+F+L*m4G_=NQWTe@uO4R94;A_KkvwbVzqMDBaykhcwdNol1y+w19MX
zcXu~PN_VGpe~afl=e*w-I)*>+2Yc@|*PPe9qO{SiO63LQt91!Bz3OwDONB=8^t9&H
zWcJX>)MeTP*?H?$uBH+@eLBSWRz^Z*;H0ig%kYCmV0q_*1D2w;4f$MeB1^8E_JsMe
zb!~KF;=W^MlR6*=?d|VJ_8a=pIjLrmWC)`DGb`82f=yx%0D_tBW;DP}h)Msc4Pdy|
zog>ynaT*?t_4TXk>+2t`z+xVJAHe7#$>-3T_Dz6qRTVvVl}AE2WlmIT=K5?b>9=#d
zkH+Gd$4w86kkY2VZ$?{Bc#RE2NSqDVQPFPM!TQ$YQUS!+eibE6S36$1j|V$RW{{uC
zT8zG1?Tp?$ZJ9Nb5Q=fUAiEacigq)!zcG596p=+ZnZh5Yl_Oa8seF1@@_^lU=w+eX
zPSCEYXJ`K?fAEwsMTrs9_U-C$R9`>e2Qee0eGcydbOWFCI^N25LQ>&*yZeJzEjqm4
zzAow-%r+k@WYY<g#6-sXU_)$1iTT|PaUfQY==;2C03vZeuD#JW=)}$BxNYvN`lzdi
z#VCFu?ugr;gN`@0$Mj~f2K6w4cvWu;YDx2+J+(C9hqk^*K}0&sjv!;1{giai4+4in
z<P$l7L@mX0Mt+>Hd3PCaaznx3xANp2F8V^L^|4|6J~zcR2|E-kGKCO=F*#kn<A{q_
zVgHH66NQPYI<szM)zhSE?j-(_;qc`{|D-BQh7m6qt;0(-KB_t)I>2(9-yvYqEfeub
zfyb2gv)}Y-Lb#`7u`7!Y_WID0a(ugcGTU=}-<nW!OLCl!o+&pYV=dvz9LP`m`_1g1
z%Ju;>n0r5rmyfTzArtTWe@B{<Xy8^?IE|GzWiFe<X$SA(=uw|keFc;GDpcH0JX0uJ
zkt<Ugc)bAjGAIbkSs1*aA*r*IbNFjWdCUE>H$3|Uz;W!%-C(W!YS^QC>td6LyLX@>
zMB(msZEgIdGWci-t~J)~x~sR`<?2X$hR2pTzHc>r>@k{Vq&%TY83U^Q7o9ZCt=TK6
zAZ_NYEn*R2Z3RMUEVTw?ufVjO8rK{DwWGIR(q#wJlh0m<%4U>Y>a$X1{2?$*neY9*
zS$f*n$ztntAx^2LSFZDe-L%Q~Q_ZHp529O!cb}#5ep*i^UhvwF-rYAGqC>aKYPnKv
z9bf-K8b@-e$-8=ywC(b}?q$f-K)-LuvWPB+t~<p-Z?K3=lEM4@A8JP)$!7blwqe;D
z$;8FfF)mom-H$(-e;raH#JUI5zNxRhYPOtf3K|<5v+%Z7>l_{q?d=IOnpyJlWOt&q
zulEHdSo~~j<S8Htmq3?*SGwcs?E8Z=V-S?L=iC|NEo%sLT188WAcq>5YBFL{neKYS
z^jy42EZ&cBg}<yixe#JA;^v=d*$iB#0%+-AAcEe+cM?zUQEf*!#2ir@@Uy6ysy4*B
zY9SAvS*EsU*!L>5&PjtpFOv8?YQ}Y#0)a~q71ee{?6prI{6iSyB*9|ew;|I1l>loH
z3Eu_0BmkHSR>W1-i?^6ZJD2th7B=--VU3ME{QUf`uC8hF)Ii9)bK(A61oU^0n?!GK
zZ){zM`7asfp|)qFbiBS`BD02}Y*MBY6Vn7G94v^(9vJk7!4Mk{)qUG8(g?C}gdd0<
z>ZHnH6INlv)MnFJBKd__eNMph_UR<H&J9Mf(18_kvF4Vg0ogsU>=#epDA$_Spg<WH
zmEyFW?(xj&sdJWRKtMG`=@^M|u(T4Fts{=m<M8Q(76#&1?|0b4)m0(&fp6#eNh()f
zYk1*2d7rB-5mjRdq2PYcg7aE380ZmT3kE|+E1}L(nS$j9j)Dt~5sVb$p4d`1httBz
z_`N)~X~lH9+^PnI<n(XxmB#N+GlP8U11Dp4oQ|8Hr}5@aFgIwjTfGSKZ%WYkM%k6_
z@c=W;%lhkZ+k-~^(0OU8B~G#VtDTjmo#BGy_RXu102G%%P9$50?7=3Arld)vheRg(
zf&~TGX%6lv0ATnoEatKCkn-)0h8c+ty1FxJwaX;BaL(L;Nh9a*Pk3Soy!%5pq4G=z
zFrbT$b`aS+WlWpc1$H^$sA_3xS<g0*B8h(ePakNX?0MRa#e35B5|B0GCnuG`EuWX4
zzY1nd&&xk1z$qiSwB>fs+2$;ET-m}VNc4Gj(9zKu!Dy1{j#4emZfL;uYqwnt`)ZWE
zS!FR1<~8Vi^2d)TqqHcqrsunhsAOh+e&UtA^9?O@Va&4kU2%K4uJ;Yha+nJ%_Qdw~
z=jka$4?PaTuc&pz61qQS8a>426P$u0X33jgW+tCa5%L?REFnq4nN3Ho?2y~rBXNbN
za{i^0#^T5zZRG4FB|9djyCQx&$BS9&otP^z(rH}Dio#*Hnk+RalQW%Wv^@^f#O)4e
zT!;#X)*yP3DH8$<UM=yc9=Dj2(P5Cv2;0Vcn9&Q*c$*q-+}VU;V^w^Ofp&*}-t!CM
z$3}13cnX{il?_FUNmK&7VbFXSZ<i#KQ&4EnnEY1m=;&a?i`1@geHYFD`0MKUVCa}E
zu!5ZoKH$T*)WU>}2JTl3x_)<)@dy1(Sq`QsmA9AXRpR3}=Z){1zplmgvSiOJdd7HS
zd`Xx|wPALd+(SgLJ}L|GA!badGU=(Hzo7Rn3G>!OO(D$7djGeuDlXJnI}{Atz!SYx
zy>Su)G;hGXCqS6wlBYrHJO7t|8hg?isa)A);M5Kq&VNgN-1_A0X8`;N>D4PXe0sV^
zEZGfe=N0mmXH;GSOcDGePni~cpI{mUU?VU+Gq2WX#*UEA2Eu3nUcU)g3yur{-VbZ&
zYx$F^za2ghq1{p1Id+n)-Cdco*8bHZEcgQPXTj-W3n$~f*B^f@ENuacqN<0T^;g?T
z$!r7xzX`Cad2pBIrNfTW2gl;dH9lPTk}Jj}&gpr}S6*7`-cfYwetP}INAiV7Fn$g!
z+msl-KvPug;VT7sSKmCzK<tnoJAGfh95|hv=?uONw@oY(=`K?bRQ-Xz<?K@XqzB)k
z!x|On-ZadD?Ab{K>m!^mucVPIlSUbp>+bdAkNJja4`#*^Jh6#Aj@<GMzJa&*FAvRg
zl2+QdLz7xN@@--d7MIG6wiiXdh=s$x!X<*6868toi18nSOH-MMJg?gsg8X%)J_+3S
zk>Lxt|BQFoW<J}2)5=op$bI_q$GG!jew%iF)=75msY>B)r6GliI2k|fTb3|RQoE8j
zT(3f6K$W)Fbw&G<sx)%aUgA~c;mXVymzDjIYb9^LVZ|SoI8HjBO|y=U{^#f=(b{2&
z%h-ZM=dqyWL3cuF4w$omkPPge?wMxbaYDs&Y5a0?w{<l{Hs=0^8ZiO(1u}?61V*k@
zK+OqkHk6eyf`WoZxRah>7=N+c6L_qGvr7CL7zJ#H_Z!-Cf9*GXkMSnpOoBG%S>XB#
z313oD7`t@Td6H<C)?9MYA)y^vU~)%;YUY60eO6@ScuwN7-V#&>*Zb4nXSXh3RU5(o
zfIBp*-h^TxMZ;hI=wH1ftrrru<1mVfJrNr^p_@JUOA24-9_PYIWAbPT^65}@Lz=CG
z(0lt>E=-iR+qkmKdrIQm<HW#`P0SGZ{J$Z`!a}Y%3p+XQOFcOYdO^%xfgbM$3wIur
zlT-2|0B@;#aVk9JP=f6Bi}|n-vOcC*0a=F4@kUCNtRT`py5O4eE8DdN=ZqyZ&7v>=
z<pRK`iOJLbEwJBZ&ZqdQ^R??$_c%RySW|))0_DqUaROh_wlIQ3J&=3mo7*e!5u6+A
zc%$=}TjTys`y9_33Z<6&s2IA~v&lLtI|l`{68m<|;w@7<xPWSR&q?zI414jCrM|GU
ze3unw14z=^&do<3=It)CqtM`${0mJQE+XwcRMei4rWZ#D@Sk_%D-lA&MJVD<w`#N@
z_uQ7mE)NwymK0D;wSKBCo>h<SI)dyRG)zqA#cd6G%<-SAnEzW(wE|$Cf`sI&zrX){
zbIiQ&uyLn;IdR;e%_oNecr#b5J=!8wI#3HL|NNLT2kUZfixjYL*qJY7Dt52hTv$bD
z@v^67@F2I>>$bLZ#%tyM`<@+anzbQ9f&i++g06U+-T!bnVah7GX>UmTR=Xu7z+(31
zdlYfR!yvC??&!+euE<dGOSt`&?~=RK*~)K?*Vc|g=LU&df7SdnB+67z)CoINd=$*F
zq-K24p&8w@ZNkcr3Jp!Qk`K4l^;K+S-Q()PKU@xaDFsV6$$*smw6`R_DyT6|eX!@9
z`--ya!zs<HOeD1!s!aE%#NF0Vll5^f{E>wZ)_-t{oFz1LDo<n$w~5HBH&~U#ALDit
zb5M_ngw(BQU`Hv)qNNq6e^SRV=53gR9@X_GCT@4U(r)t5m^8XS;XFg&Jbe`b=Mp6Z
zlR<y|xOJjjiK?RI+SmFlE-iL2zw%SW3r!+|fu#=p7JkPc1@2@{Hcf3Q_rmU}HfZoj
zen9SfH^IgF3oI6ebD#Rk{5h8n79v%4)yaoadcR{1$J=c{TLdDD(?&9zw0Si(I8pJ~
z!#`fSG=qCF{=;;e;f4lKp*eCBKnMv5GiW+LiepIrTc`W|;m{KdsZRiNX32MPfxpCP
z#Fng%xP|dtwSgE=o9N2R%d1SseSdi6H(kzI7$y(uPxO2Q9$z7<#TxH?9z7czcb^%0
zaBtq|FE%(0C2^tK*w}#B^=so>$8Sa2vQArE?`w~ogUP_WfA5W%k^m`SP;K%}<myKK
z(y<u5j^~|Y)S3%*5^g0ozG}s4F%go)(g+v4FDfSchCLA`ehAg<we|GZLpuJo_?a5+
zl=!(2cLPLUVtJiNF^yXy$D?~rbvM{#RC0k-V+=LJzlVQwvAOi#q(d=$M{M*P>gXiz
zIVYVINU19$fBz<2UPX7LS-U6G`E+CF&&A}F8XA!clL<@ByU*i=<lT-L88t}&M-gl@
zWHBgLKe;$FIycDNohc<+T3Y&e`=P1185Q~~*p0M4+(?t7necs7+6u#E3s%S!hZ<}4
zaB;ud39FGE1s-|v=c};ix;1mBeC6|}4l3*XPB~}yYR9IdCRYoa_g6U%^$0rU-vZIr
zJ~8*9a#>>SjCM70uithf7^1<oZ+|32^Yl8A=gg4!8pj&@PzZ!dq;Z3%HXRyZ*#~y2
zxdw|*ivG9G|2vxdh7&m-Lq9<O$KwRNYb|Z<r`A#-H0XA-5!*toDgbw~G#|a7Qj+C{
zi3X?BxRMKgl}aw<ASrjs97F-ZgLfxVteQD)dLm_tB-k?->d(hJXQ#`KXQt~TCVy5#
zy|4gxFr7F+c87N$f{rS&nx5<$$Te~mWGXaZGXW*}=UpnHLCeNQO)51V=SN#Z9*B8$
zH?`J+t%4ip%;{gGT<OHoz3O8Y!#-t8cDg$kFPUv;<kVCWd5!q~b-xx$-m&p(R>}-?
z52VrDtUY`#id@df`WRd<90E<`?kO@9o8#9<h^8ZZ3`DDYM>4Li5`WLaZK+vsbeVJw
z?YRhPkr((D+#EMD5j;#kMJrWr*mHrzW&qcKU1x&hNHrJ%fF;WM{(eYn>ywk1ILKXZ
z+dAZ#t+v1dbS%UEzq(MSixm{PN=gd4XuJI$t5Po(@DZ&F_d*?0YW+R68B))iqG!0B
zY)8-`Gfc*ii2JTg1g)Bcw<511F!1~4h8(C3-@kal$PM#IAT1*U@QUl}Yb`qf<pTHt
zv^M6|x3SrLkN@8icSD*l3|feoiYhRW$om&CQ#`3~c=TzvWCwQ(`01Ky;XcC(ai%G-
z#sfld&<3QW=p^wwy#bs6K!%P7_E^A@Nis{Z!J8M=AO^Ft{kugTS@p28O3<V}*=@_$
zgU4$tw5JOiHnUvjeJAQge6dp}6{ETv-46lL0~@c=s0RfC3e`}|k#%1fa4Y`u`nA&M
z{Mt8pGSfo7o|}{jD@E;7CC_ZV1*L+5_UUDqAPV#V?I_pGVv_|oYWiX1Amf{kys=Be
z4MM@3zr?i%tZ?u6@{4GQYp*#qyl6^FmY|_)x@6F0fv-5O_B&}cxuI7CF>Eo{u6N8Q
zk5C8%>_36}h*H2LTT#l=k{-Ao0AP0!g`)=kwXY!X64WS;ot>TCFF2NxmL}!rCoG)W
z`{2csoFE;Mh<C#s7~?i&F{)seXVRXpoI}9dvl}dp_VS*S`4ZLA8OpCaxKKE3#FBEQ
zb@`?IRS$e$w&Ei;*AM;>HG8=%vIf)Y`Jfy4SBPpEd3mqnKYWW*Ti>d}>WT!%qPttn
zv8XZNe_hQhu<J0LF2VuEiza}4tacc4?vtz;sdJ@PPn-a3DZ0hPpkZMR-pAvQL;2{s
ze@qQl`zeBNXgMp~_eE&LLB;#=m$}85aZWl?kl@i?n3|^rEsGr5!lmp1I~^O0I`vOP
zUDf8!b*?(=S5kh<(8Oix9+yPdp;;+%k#6?z3UGXRjH6;|+Wv91nf>OHt~@oRPd{px
zoo8ib$kBdEtNqB~9-yu4W8jr7ZPeK?9iaVT%y*NWOG7UsrmOq$^Dd5lW&<jL0~Uia
zVbS6}(F>6|c}4dGZ4YC;N_9a>mrnunR(%UzWm78Y#nXRFXC9pvM_E`|CwEDA04{%j
zZ$l1D;X$>d%ud!04h$rtRRat;t6~w17#xHzoe4bVTXN8ijjOq;N+x#a+|o3K#O~8t
z+<~u1;v}VL3ez%LQe1Uy!hYZd+&6s>opv4PBmJQ(n{9s*6zQ1I_<dW~bya-vJp={@
z2AtDvFRiqa5~PrBU&VY&#}6A@q?bbfLVfMmcXuG8HEg~)G_faun>axDG@KgKM{i7!
zbqN>a4UBz#Qlp1sF8JU1qS9++E!0hXL3rH1NGxK<uX1gY*ZMMMl<X)8Qfm&G7%3x(
z9h}w==H7~4;sG~x{>TbPiAiQAH#(B!TVi6pTHPX%vp1pH4ZK*X5v~yBBYy0?0tjch
z3rUa*pRD9L!G`W{WEue;oR~TI`g{FgeNc!h&XCIOI|F%nci>4odA?}L2BiD9MTpMT
z5{MV4N2R63?tJ7%$|+iLoJQhhOW(jG%&t49_Z0;$jUD<$VuFIx^#Z}jn5*1RoW8^n
zzkbbbnBZY!OL}@XvAD19?z$-*X4BEp<rfy-$!zakxNp}x3$3&^6F3>r)+bfQXk21q
zMD83*oDQzjc)RfmAl4!KR3`}7pXKM3nskprori~!0*(?rzZWLdPQ5Kmh-|K%%O`4<
zaD3Q0YbX~6P=N#lh&tUAV3F($-^WJ`Zf0;cAbv;vmyv`w1Q_<O5fGXd=j+VwCo&}P
zI4m<RnhDAp%}zo>^_N{jPEO9Dh1#+>R`!l12FuCtlH=-)U#9zgYfDvx)+Qs%<99}z
z(UZ`KF{yO5j~#_tStzM?bl<f2p<-luHgOW_IQdEi%Q)T9#?*f02wnIxp1{>LFvETd
zWr043rEzPb;lXeAkX@EgGoV%%0~ic#-h5XUog}dQ!ybAvcA1rBm6f-FU=*W{5tG_I
zyhx;NMMHv0V3sl#sG$uP5YQ{dRsSYhJzO+LP_WJT9myG-`0p**jdg*SQ{AR<`#)Y{
z>l_H{>yy&Z(5QKp05(<J+m~@Ieaohx?*Y~!b@lbA7#RK<#nyl8!SpJd%OQtFLE!tM
zj*K%xchned?MU3siw|ec8~~Z3DJ8M;$Q(!EW<5Vznom@@MEBZeO(`9vGLC;QyRKag
zLok}pS-BcC8iqCP*KMwO!A%bMqX}#lz!4m{WpgVXi}D<6hW{V1dAi=KJKZgd0{sRa
z9+=3ofUSEgB2VMZYu_*ZhQhQlkC_Wq1MqaaEW37wZ@9)?s6H34R5<-bAXns4c(`M8
z@2}dq_?gNjGmk3Asv)DR;Y+&{g-aHI*0?C^c5AmFVh~Zdcxtwk^0#RHTD+#ix*9+p
z2=z^?XRp8SO8%WzKgT^XMcYTmmT^;m1u>DcA}{Io0`CIcyQt{vwrcPQABbD8$w^iT
z1hXj%++2Kr98%MKB%|^WP*JI;o=n+@$s&Fo5m}9%2pvIF_UoJ&O8tX}p@^>N+|6ST
zGA(_L6sZ(xq?Rf^oUg4C%x@TgBRsoyy1S#HpxCOs1M0{8;$kt<NY3I)H38ONOCf6d
zHGeoC&aw5In)^yhRxhj?C20!vwC1N5X6TbRiELOI87Jcz@M#%8TZyw*hH+;EAhU<3
zFUW`2YHVW2aG)m+ivwQ)015(_TLFJj{Qtlwaw;kg`|?IXS65CTELsQT-p2061~#yI
zfEE(WR(vqfSyDEC+nvw#GshS;YDdz}!A@=bDl<;~={UYi;LI}9;l1jE>}=ZS&<u&#
z1l$jExEmGUMm((}-z8~Wpa&UHy;wn?31hj-jTp1Q;0okGi*yv2JBZ#Xl(C(FhDw7A
zPR}VoStbsgPZlK=%drJLfGxQjiHL-cgw(28#VA2nFI7}j6w86Y2@8MA_yz_(MJXqY
z?h~h{AKFQXW$*4JV%08h!oc4TVMH1~#^nP<OCaDicYkB0ezVzI80o6JXf*t{Qxh`c
z*f<zOE`*X-T<nwp3w}KSt^hLP?BS(_HzmiLWZ6L3+)hBHh=_;~e}!~$aRFRYz!S(<
zUOPIH?Nw#Kd7MA~WyS6;_6>dgurAOTmq``L<I7|)tv9b-u#5A7hz<LCfxNs0`+BbG
z1Rv?o$x{Q<9J{2m#}O|CjhX}oF1PFYf_&!5<;lD1KbC;;l98F2%%)3POZeZFpB9yx
zefgd4N@}bQtbAdi5wJKda(|+0A``*pUR)*qRpZ8ldUauq&sQ*^6d5ZuA1TMKL1(gQ
z?@Nd)@+UapwKGrD8B=}<39523?+?v1vcr+?K`a9+iq|i2JXGzsLn<TtFn*@+NISwx
zY~)Z|2}nD2x+kaQT_$lipoO|X(^h{Q`+j+_#2fR*3h7bkRN2&Fcl=n}+#;BUj!Q&J
zotDDLY!a{FH2j0`DuJZgfcXbZKIb6l#wln;n%~ZR^0L3?B~P^5bbbX{ZQEPo^p>U0
z1_`GNLW(fq0V!`pqXNVjf)*VD`i#k$8PGP2?O$|tcY`JafL*o9<WQ5qj2n36>gf@K
zIU4Xe!NMMFMgiX0*w}c+b8}ZC@Z7P2ch&?<X)TBQI}NMf=})=SJM&n-fzcNBGhN;P
zm=#BrOGnG4pb@4;n;qJks)P?k{_6T2+s$kA<D-tOMOf3Rf~k+KY9K;GM~84Yf$dIW
z^7?^^^j~@*Y9ysy`plR>chah&YKhjfu`?jWTi#i+Up2i8bGQrpeIf8<&B|eLaqQ1G
z%js@dGOZYwC^0sRw1ae*Te5;rh>9d*Vs$I^Q^y-|i;m5EBX#zx+9!AnT8?{t9>p>X
zm~n!mUQlg*wGgS53Kn6K*$1p|@^R7z@fALHrlBs&Yefi~o#MRcK+C5eOe22cM5~X=
z2zF%sTJC=vxT>TkW%dXs9ahp|I3De0PtdpV`VH0zve2KacSY(n)Tyb?f3`_~Zt15y
z5!ot`9rEFJli7U6#azJ<ic{3lS<+iDuU1r0_*z&<_wL<0fG-LQ3xoSDtEFYhEVpE4
ze|*=GWAgUS9t6Y#i!Ojn_8Go6K?Y9MwXeow+u8)bArm3BwF+E+`8ufj^2B``2{4Ee
z|GX95h!3XaUv!BGYP@7Q{@UVcx56x=G!ehBE{lZm7Rq);+eQ0w0WTDU_5g#@ygcf6
zQI2>hsdfMCx%-^{u7?XOMQ}YmTkEt5(78L(czb&v%vFaRcu7dAH;1lRyO73kVA!jQ
z@V(97y>^x~ck~`99fZTb#g-(mt2Pb2EF!xY`lKFQUsLndg5ZnXnRC*@37K7U?~9eU
zXfi70nLX5JecTC7_}_h$J*59KisSQnb>tNcectaG<&3Prjn)#@J1b#5=EVoMWugo9
zjrtv@gDcmKTj0CKx-<^c$U?q%lPY?>VcLgbO})&@L=5c}78C<i0`&)HWk*_HcN;k(
zLp`x>23aC&4uTJ}t>cpEA|)9u=-1ZQMI0R+{fyh*$0NA7xHL62ot&PM02;f$KlJ_m
zeNyvxYf9;CG|-0U=28K4LhUWevClawgpagyHd3>#zap4!#A<iD^Iqt|6+qHi^&y`U
z5@hFC4BVO>V5f}@$zmj}^D~p$_3BpnazzuqVgK7Tn!Siic>k=x;eEVbS1W1<FP8V=
ze1i94QmNM5gM&Z_g&8+=+vDQiVF>5(12CsHooHa|#Vp!-3~<0D7@aGNUwH&;Awm_E
zbS7$Y*QCzB^L*)S*#bUtU0x#nS~o1KBB?p95Bfi*-sZ7}Nk+|aU>jJyJ@O8#;t5jz
zCQ8|B%x>@e^ybLRnsIYG49r*G%XVaUEC&CyuF|b>BLH`9d97c<cBf{%_n8re3wFq1
zWEJ~lJG2$!m^q3GE=l)Q1j^f~;#)aT$cO`{sU$Wt1dxfQCt{*Y+B#Ne_sWt7B+XSW
zjAJIKKixBYns(DhFx>&i_Tc^k0RzDDautxESo!KO|3kh8<1nZrU{4Mj8?X)Cxwxe)
z9#;kf%rmZ0@YMsPiv2oY4a`zeE;qIH(OGPWej~!yxn^Ltc`WlYjta9exfw?6V1&s~
z#+eXOd3&+8FVZVn`rE!L3&HkNV6Dyaqj_B8D&U5JC<kzjUNbZwJ#qhcQD+9`S=_-J
z%4oq%{Zj$Z#-t`TXKM!ku>|bqN%ISVo}^W2@Ug=K+-RslW1|GC+iy`nedEM#PL^<L
zCn;%J6@}BW>TLN9k-(2R`@O-CcdBZuIB}BN_htp}!_lKA@xFBUB}O({;1;bDOeIu3
z44xv|gm!czJ{cd*)?O3*{e88zwkjfZ`S}Q&8#U-dSbBqhK?ZBp=5`(<ub;~h0kGK3
zL<(-FVvb6(63C$I7hKB_o^F8+HsKdr+siXUb5xnNA6<FB)hT|ZKi`U^xum`^`c#{a
zQj{-*0u4+(ag!a)K1M3$i=sdSRIOycGFCc!L^4`B8#v+u7g~^vQ=d+{>z+8NYFeGA
zLfgWZ&~Zk9CdEkm8LUesqzB5#R&o_lxA%tuv0R<skhtJXp;xGUyehRa!d`?A@Ql~C
ze@;B_H@_`PC;{(uA=(%;TwM3T(udaI|MrUsZNPpbnJE%zoL9F~eOmO*<-je%=YC!n
zGvBy-k3Ni3;xC@rJ~I<XP}<u00PkZThIJ^{Q#!vWyB+}l$=qpja|`uewDHxty&NKr
zj4}E<LtaSA>0#b%b!0@=(qznr{<}o^Iye2=gvgZI>&uY+zM$+>yJZK5a!rOw70fa3
zm3`&6*^0KtuxJ5M=VYy`gb80HZ1I;`u2vKy14~G;LWVz_?g~KQoW#s#4c|kzqG$;`
z5d<Zw`8CILJoz>~MTe4p7Ixx_y%5OwRo7JhS>%Qb^87e&??F~{$>Wz&U+HYHO#IA>
zt3ZtbxV#|ZTS`u@^>YuBFEOXZYJ#s8qXcPW87YI`LW0K88|2kLNI|R<u2K2LtAIg0
zsKoO9^pbR1Q{+xCu=!^E_xl0F3xD@IB^$ds)uj@5;$qbhEyky(hSoWO|Br#BkjW@c
z5X-1N0Crh6>(<r`0FD7OD)`CXcowZ7kw7F=0)j<7`t9(=D%Z0IafdS>-H@lY)lldF
z8CU4)K;(7Q{qMMSVysjg#}<&^s$OzK5m+>TT2l%Zl&-rSA%f}lxizG5+VIPwq|xiB
z7=99_yw>FeWyuTDCj@?DF@}nEN6ay9RPupc`qcd$+arPUu)Xp%9&k=K`I2Pl^2o0E
zXF+_SU8wBvH3mGkIG5&`9_94$zLw{r@~l-f9LT34C<h#Tw=V5%>%v_i)eYeF3bZjZ
z1UL0)jEsyQF(n|@=5Ku$We@%1uFuq6pkYr<2VHZ{8xmgzbo18Q;hpj?)zQ8k;<HNt
z(icu;t%2W@l^ll0K7l||VE7&f+*30t^nYtnV50uL%POoF8|Iq>GzSD0lUD7MxEyl6
zuH+9h3GVKPFANvA^EOTevr`0YA&1*kedFiAi(3g9#|--L>OD@wy<-tO-O=aaaaN{u
z-m~vY#L0h7_=3nDL#Hg$a+pI4)+&`_v&drTBHZuD!mR`3Dpe=dHk8yNq(3gv?AgnH
zcAAlQH)QzD?y`0{{ej#X7hyt2^qszNrkDC<vB<O&u_UfIi+{qKgSTSTYj}9_Uqp%)
z-gsDpE(@f=P)7ITC^RHlEa`TG<s>kcREFO^J`B~XMqSw(xL)QsYfnxNns(K8w%#7o
z>!~tpx6pKfDziAZ1thuCw!c#xmd8GEz&r>%5>QfWJ?Nfv|EJfWWVT-51ceubrMNtL
zGk`Q~AMF`V*ArrySo&44sx)V|uGRnwcc?^mE8)!X=M3Uk&YfT;{gB8W3Pc-A?)war
z=G6|Pv@yZ)EGF#V@c|zEcT!2N+<tR_QVauZ*R=KwFXc>tdZ<qY>ngF)3t6c7jy*xm
zTaQRbyMe<RTT?q=*iwjDnGM^;BDSr+TX)=ZvJv9iUA?~ID@QHG^F_V1`;$@-EhVao
zE$f^?>-zkq&d^l$&m&zwTF!)w&VAb;NkJTp09tbVvvx%4a)Uhjw#{msK2lAHUMam0
zb7I6R4vV_aLTwdJr~|clKD9-O-s00fTA6m~tXNQWDlrDMF>e6V7vv*!cL#ubACGnj
zS}3ioECQf_gE>tgrRnzSa4;;H=Tx4S+UJnVuq}$k#LaS8e6LtF%J*f^>GVpm{i>-H
z4&Cqfc`TZyIj8!EC+(A(#k{qr$f}4$U9kte3o+89PvU#4p;>u(eY<N^_74C3Dkaf5
zJuZusFtl&)LV?ZnNjH|RRt?G>$hQ{C94`X~0)Pq+{Jj{#1J24K2QsI-%ft0vBA;hN
zNMMA+<<SNpb)b{@0~Xp>3l8B9e1Mx#_g5QH2&HWPaHJ>cNsZHM3t`D_Th|ok>%e5C
zKi{Qerdzj<CD~tHA~sy62QP{x<R7Aq<|Mwu!5dBJNg+PWa|RxE$WG@HQH2Ypr(J=+
z990Y-*<Z#*7-JCK5_pZ^K+F}Rp=@m&um<Xou;EAad7EKJ2t+UmCA^&9l$6T$QWU^Q
zGUYHOa14pq;glQ52>6m`z@SxobiPz&sX=d4K=8>t<r>T(jiD>kwARAdHU}u1iL|zm
zP&=YbaU({vj-0hMBLK<4kOvOH!`*p$SsDZob&sSe$TYfGOjf)ghUSG*#d#fumFa8=
zf8y2JcK=Fu@bP^=JKY*v4@b*BH76z=RW%Q~b7lncgxQFgPQGUM#gn^hIsoE=xSGhw
z$Y&A29dzXW{}xUk^+4=N+{1(TJ7Cr@S7Rk3Cf3Q%!2K*?axT;Fegh5=fLflLHjp=z
z3f=}7Kxt!0W26D39{Q*#;J70X%CYoJiT0Cf@Ivv2QW;s<t>W5c(R}5<!!s<Bl9F{t
zY(tdjRy_hxf+`x^gr^L+aJ&X$Txe=88%Ky9^RL<Is!bHX<KLiNP=(9KC2tjp{7&Ke
zF=a`<c~{G<8b>N=R0kbE=5!{wNAP9Fi>xpH7GE(RBXM}E$#}7!Z$}Avt&tceXke<m
z_h&Qp0lVQiV)8h2C91_O>SX4UnWyf)xA42$X7Tmu=EBHDt~$FpZ6_|_WK3%zbAq<7
zFJV8`SmJ;kFTkF^0~wA<!(u>B1qcK;0vb@CI}P{d6yOQ8-71%z^J0vT_a|E?Y&5G2
zs=>6izPdR#giMoyxxP2K>hg-BZxJ!Mf2hCWP0iNExXmu@5fbudG~wcFTL%ij)J1ka
z6k&*C$?r|J>>K#J5H0t3$hS-l&^{)naO2?wF0vj{4y6B^x&^@VX7j*jGa7*bG&(?a
zTm`H+>+?0&)zXc)X-)rtfXMjxpkZSX#<+<RZ5|K@BL%<`qGX#NZczEk1J4Zq46zVw
zjuJc3H4Okz0l`>F#&xUjqV;wt$MQ;??$%f3G|gq`5&Xd%m3`Ig+uQe+JIp{Y9YYyJ
z0q#@c*REn>FXVrDLO+%8;SJL**1t?W3O$^yhIL+F%7>LC&qJG&Pxi(SpPl|yFi;l%
zL6r9fiair5pa#LqoBK8jf6qZtq2@rex(3p=%MOY;;28m!2h7@0!Dz2TUJ0dnPhI8d
zJ~o^_x|E16d3`bnYzQWWPK1%tGwei%hT5;`re4yXy{@-T@~b-LEb8BfgNJY2s=Q-a
zU0nqWah-DYd{NT+Bk>%M%Y*tOFD+!2yGvn{_mTO<lt6r;{h6eQkpFe9fLCC%3ZrBA
z>ue8E>gWa+BQtaE<Za4MMcDsx0T@W8H9L>)g?nnu`twxb&UNbW>%5Sh)j9y<6=d+{
zMV9t0OZ;~^`*FpXEtm~A)brtD1}s*iflsM;FED7t`}4jrnFIIPhqa;McfJx8CX<e5
zx5f3%4PevI(9w|*6R!lHJtymcq@vna23Vv@9C*M+^#=}%tIbyCbs)DpyivB=$y^lj
zY^5R3;7rL%E|&mEM1?!KGFj~qL;U@UX|sMi#X@hhc({+$^MECEONEd}{2*;SwAaOo
z?BoU+VjC)#W&1~K=qbGU4k^6g0}=lt--HJxNxHwFutWr{m@;&C@S9g+!IEm_X?b)<
z1^Wg24aW@vZx7z+izK|E)t3;$5W<ScsPYHqs*=fu?tI5Y(Ok^&5UISqv-%RP2r3iV
z`1rQ^?UvL#yPpa*-W1p$FubcFAB|zFthn_#w<*tckR#v1OC>xKLGVkTWOmckR8+*w
z&1pfDVvMxdpYsq&ZB;?NeT}T4%7_EAQ`k787t0Kt*H<ub+i$NA5Lrdgmz5$31k%kX
z5D`uhsbAj)S>8c~4nnoHc-`%Jw(PyPT@j>+h$M%*P+Au3Z3$%WdwToG_2Q2LnpcU6
z26-sv;{}(Y6W^Nm<~YOmYpo0Z@Y`S*Gnb?ZKc3e6*eQ>z2?mCkRLaoRY{bNV&r_$I
zi@lNJ=7#YB<BzVM0i(S(*_#-R3*gU({l#SUI0StUHyD*iQ?s>rOk$z)<?K%pjuir3
z*DR<|fHdPvJlWZ=L47%;rfW1;WhTK8J#M=8)M&j!bqRUy_ywMBkbwxRPa$0CE_w$<
zqY<N|Wn{bnH?B^p01Z@hw|+X1e{j-jWmQ*^4X#`Tj>VsA@!q^)O!0()c*n(IQ3?!4
zkcbGr2-Waopb3K7P8nQuSq>$aY61>_j<{SeAaYVPw^ddCdege9-ze*)&T}4_Nle&4
zV)94DDw$t%##C+_qLWfIk~|#g=D%NTC%lyu@R6Br+;20@+K7i`Ubc;^K$eR7RBKIu
zf@s{V!vv9i&XE(IijF41?s6T4hZ_~K?UYv5hW*0WlBICz0e8IQ@4c{$PIR`bHOVnm
zu6J-XD6bvyaVNBEdP^cxF+U?aQG@Tdhmn)b&#AXM%YRmSkQJjHsTZc&BnUKt#XKal
z(NTYeN2YG~pz;?BYeU*G!S=bH`Dmq^#h}f?HLz**6^cT9HAnuZue8M2YZdB#l{wC^
z+W4VsJSjPLAOeP{%=C_tlJUvc*q8-&@rcD4Ne2Xhg8B($i<I}8B&=bsm@r@fOed7>
z5iSSn?lj=J?)`8}V3!C?q>%`Du{pL*T_eczg&dFpxa|jvYz$O&fJ>1vVrAWOV{F~R
z)e_kb{A=>7%)>r@o7r^)Bvl~MhHEw}a^^ORjeH^4-EJP0z$P`jrMCptV4^D=uAsK=
z&@g*9q560FKAfrbM#HRismjEq;Rn%6!QEJ^Sr5vZX2!c^Ir0lWrq4}(bJ;$@cFFbB
zUgG&XtZVoh!NNV%L~d*t;v>PjWsV3bS};2bAb%Vbv0$WFJYx+rIRd6D0l~o{+9#lL
zf@Y{qPB*jF^F>tg^hN*s$M9`E*A*vo_+oFOi?pulgpJ=`D3Or~(;D|Gu|b}V*u#GC
zuUp-HIv&GBvsEPqX(b*Idw4$`_fBb+hR8{OIiD*m)t}nOxgJ4#Uxdj&IGj7XnG?`G
z6z2xR+w4tNcqPtKg~m+H;x|lTR8>uN`Zq=dd%T?Q)o^;-(|Tm5Dfhk#gr1v$B87#8
zGZ+M&0drYEdPrOz)-W=8k4g*&E)ikAq#ptLJ~x|KAk0e;p#3PxO2mWjLF@}JXxpmH
zr-H%d3W$p!uWKN56u9i}g1>%#(coGJd2ZpL*N`BSLi-PeY55qW0MYD^b!qtEA$(hZ
z+_+t7_Y)=kY}r5l>ebzuw3J13N}0BTEb7NO37$TjB1~>19Kl|am)CYjcdU5!G*VU-
zYc)&{PE^G`vL9yUZgM_WHL$#c*xuU^4Ci#4XwkMle$Ut@gjq?jC;?8z0<ZLd$nyTG
z^{Km-tHmtPY|QorUTdcPRG?&3SGyl<%<mo$ya6WQCt&mDY}0i+gf+14E8r8*uwNep
z-znY!t;Qtw6n=#^!g0R4pI0L7$wxLnLsGpD_X>TE!yfL#58177qsqPC=_0TjlF%1Z
z6SA_l(6$dEC;fC(p0#3kmutoSCtC~bK`}RXJKWm@yle&&cjp_6pVcDFgP-JA<NUHc
z%N#h8)gIay_!{4xEjZ+z-<Vx=XhA41b_VbgeX>C+AJu)*zDTwrxJpj%9<CoAGo44z
zZ#*$DFo2lTNj&!DW6F1k6bdxxp7Z$OyIZc;LKf_L2_zPcfw#ePeA$gRFr0iQ7lGV0
zS)9uG!}GnV5DJ-CX7_VbkaFh+aH{7>HeAo8DDrn^zsbQPo5<_M{ilNk;ka<I`nmvX
zYSvlaL2KF^jGLhEEb(l$k%Rf@;Vd_M1Wu7V0)7&+PJQVwr1)a=9zG)@?B8qK+Wi$D
z#@59%RVvLE<b_AfIg2dEPdhpCeo!WaC5vAm>#qMP4**?LBq)?|)gF8KxzBaXV!B9*
zvjiV0`tO_utu<FTh{CSJn+L$e@WA1p;Pp92<yYu60gb0V@3A^{mnCP_b^<r^ZnfSt
zXrP~Z5q8vTj)nQqufAJ->mV<8Z*q>;NMO~tbZyVEX3GhN3cL9c(EkD}F&K~oGnas%
z^psfEik>rFF85Qn_LvwWC}J_6Z%O`pVXWutFG87Hx_%?umk#ce>PWMd>bQJ~*{4@$
zcV}J50+Ejk_ysSa7)LC-gB50eTaEBQ9J%(3qV1QSzx85`>JD>0R+E8UM@5hk0MdKR
zt7+Axsx?A2ihGW)lz>{;-F+<(UicyC-xs$E<|0F2R^l`QTToK6RqFGw|L$ehoc>{h
zg~v%dY&?ezA#g*?kcy^!_STV2=4RV#0zdAtV37-Sz^u54AkhtE&@UfTp}eN1aK!~N
zp`C>`Wm!|^L83@}ad&3}y9tZDpLx^*;Fau)Vcr|XONxeB3_(zAxyR?sjSL{8k6zd;
z6|?R$<ib-bfFT}ubGAJJnjlcr;&i>aKzi&-+x|F=2Ow;;gT@rEHXIOaumNobz+xq=
z!{6c5Vo~5=6gY<4SP$WNOL;$lEr`_|zQ#FQ4f_~qk)C~^!VY5dLXF9liY-F?;XF^-
zZv)X2ZT(dl#GS@>x+)r@n$r|XvI+JysQ8-EXm|oZRMSF1L&IqPB(AH8O^2N~a+3~|
znK626Oic9s(BOD^{Dd9Mi+iI@-1jM|gNNIW+uueA%;J*#3%0nTzkY<3cD>B~*i=iV
zaT{m;{_pR%SY(6mxnhupxCyo<%9i9o$d{HaG!QeJEw}Jh(9duo|F&eW1oBAnVI>Tt
ze;WkWF`$s`MdVmWtFkTmSgd+B#4^h2+6s26;5^ZN*2<lDXEnzP3g&d$L-WP7rc;;z
z;M*he*EV!XAAtL*f}YMu@Sn683@rLx$A-+<Rggot?r;}#{P2jvZQ$8Mj`qe+BN+vS
zQ)b{Dn@w!lXhG8pdaF5#u;XjQGj%@B>)rPtSK{VoJ2fWt&!O{lsjkoDzO0;FZe1O=
zRMamJ776Jl<DSspMS?UH3Wy6_yF-T5e)Qh$Et`RHt^-#E0-XEw!5jpr?F!o3p=Wc(
zlC*5G*<ujr)~DA_%Y*60hvZ_MlBUD&tHQk_%0^Fux$**+qL@o?n176g6M=_TK~Kgu
zD15OuovE{b958ddoiF4U|F}szq`_cx^+DuC5jR1p4i~f<aUY(hx52I2vf}E+=Sg$s
zJT^;Sd11Q5fWo=j&ofm)RW<52`8yC1ET29?iMNm=LE<@#4Gi<Ss(iGeY3HitOi+sf
zmT`PFbK{M^CxJoz=K1z{31qC{z^?$L!2t{3ETCTOtN1{UMgmqopl7gObec&%n&8dW
zr*M^$k^*s`zy)l~yjnVK!<I8IFRyLW9`!Xe8vm+Joh=m3q0(2=>O^`fOi`6MmCsT#
zmJ9Ncy5AY-=squ$QxQ2NUc#-9@VWh(a~2H~-CML`-U66L7I@GfCmc6%RZDe3@saHH
zESVtp;C=7&eRBJ-cb}RG6-3_6VCh~5@=y%bjR$_x^hGk49Bi%RkzaT7n3nwdS?0Bq
zD3%8s2C+J<;Nh@;WX~Eo4HfYHLjyhI4J~MF>_|W{WkAHSGdHU2?ZaA@0v;`-<apsL
zj&8B=E1f@eFoRFcl6ysJbx(=MefhRynP#}`$zkTLsqh-vXERV<ZucE$rrHyajk@j*
z(tP9f8i~2$^KCAN?i<saR3IPgufqH(5hy(%K_JBWa4yJAvR8=L0X$pZF+sq=+RjO|
z7LFnZ%I6q4a|)>pU}H_M-SCqt#dVD}b#oVVg+KN3LFEC_gcO4tXE%Ak$E2w#3H-Xe
zgvem;LGNY*HzQjQfvCk`DMfL8ZVh`G^C^L9iBpw`x^t4W9KW<cPau~21(lov$t~sh
zMPLG}6hd9t7L(5nd=z=4=T+;^a(Jeu2w*jbZl|W?+4)CODqnI=1pZFnT6(W>K7u?5
ziGZ6r`)8#I6(btR{(FuR#xhjz;VNg@b0~e`RcsxQ4oO!zPRF%3xOY`~+m1I+YI75|
z0u<tvnFLUs>zNelfR6~B5rX`3tO+r95K#T(W?Mi$nQu+b&j}s}?our-e-)1_1J6A-
zW6$P<UMBo6XG|_DiCoD^^OWyQb|~nHxRQvderzrKjcx}%h*kOAAgXR`tipX`=Dpw@
z5&S~<G4rrRE*pq!Hl5I9iT-VzUIqxK*WEsW=<6;g*nMD--;B$3kJB}90&vTZ39KYE
zG;qK=Q9P3b8)5C#_1R;3ce0WKQ`7M<^{<}_Ih%~sY}~paOssJ}-1-#k0P~E3{&r!W
zU}_XL1co5y{`eaSXk#+N`RCI1ahUZW74%Av8;f@9E{jfB00kAg^f>joQs`K}-*4@5
z-Db%fVkhQ)GU@q&<3tEd9SK2q@Deh-8MKHdvP>_`X1bAnprG>hE2`NwRTT*uf0b$+
z{h;Vk%NJWGE=WK?07NogTOm|Ni(^+=qm0;&WTA)Kb_LRI`zuxzUasT;4tPdaRyLu1
zZ+YS$9o~+DA!1%B76aAU#gc)qszzy>l4hvIk-4xp)NpQ_PRi}l$eqI`*k#Kx=X^&z
ze!EZmNwx~et0$V=^wXG<+(`w|KttpIs^R>%%nh)gjX4JBAxOW$1BKID0Cd#3>ysm{
z8pW7^7r$st1v>2Kl;4f!ZY`coPou<wHym4(LBnrpWeUpJKO#gVWxxIP8?Dy}CxToH
z=H{N+j=r#zXr#8Yw5ysb(0DizJxCM{vxhJCEuAU^#ZE0c0oP{;7`aq@?tEpao^w7o
zeeH9ve50<{S!;Ty<9oPBK+Z5<T4k!1uiW%kT7P~&BVz|YYYx5#4MB89w83>K-Am3g
z1Ru^}gln5RI|XE!hKh^it*DVpfrF+zSP&7bB9!Ybxvf2)w$DEvrffco9TPLImp8V%
zmbxsl9j>zM!qXBz@rTom-aIB)F+48)a6ZELz1`jR@_WNo*rK_{RIXOb^*sur@yAV$
zyp!9y5(keFSqoq2g^Q`$Jh}F@vT|#u>Sei(U9uN|n*uFln$~#K5CYFwCm3(y*onwt
z3!1(P({WNz)P{i}A2yr$q#6Gpi>|slPG>1W&oHvCRJ3tmLv4M57xq4Nxr18$J4#LT
z9<p@i0u6#@(sUHXVZ%#ApZY{WBMOG!ia5NcDDar{D}r`*nInHpS>HdKK}ZF|2O9tt
zc_{rLXi^HESVDhWXez}fOLeK>)G}NCt`N4oYJ#*`)h>P`m(ucixEL69;@0kHffA-$
z7Grb8Fm;5kNJQzg_O({$iXtxu?*hDk31A)_6BC1hg%uqei<wNv_V>}$A;k7I(!tmB
zu6p&{sKHI{Bd@-Sp?RCuVV@_w<=7^*${Yms6R6ImyI)f<cj`%eKZ$z6GpdKB<7Z@)
zDp3uwFPD0oS`9eXNBBbqJZw6>o|2n*4y0+D_-~s~hiwCl;;l_mn{Fhc$j2L;EWkcO
z00l*s2)K0o({r1Sr3rZgL})0H1NpmgL`nS?ZqJV(?}BB}6`*SfpG#ERZ%Y4WNgOMe
z5MQXP9FR~9>4NktZHn7`w5-)q6ImCmMz@MePCc%VE*tFy7X8?)rek)=HqVB;@BQNA
z<0TF-FpJdFCJ&trsx{!O9|^bl;y!y_$`~a|R!*Brj0pM+6Vy>Acf%5Sev!`30Q2&5
z;>&`)OtGwEJ0zWrD8tLs?7YDuLG-$mPdUsLxH6JRp@E|VzDZyq&FZIIc~W>kuaU*L
z?Z7P+mBlL%dT}Cxd^+A6nDAj1(H);(fG783)-oRawflk7?1ej~&9c~N{jq(gI)SU>
z3WU;;XVQ6T^~z!V{%X8H1si!E!DQO(kE+JVEu6v5)3k7p+%H$HZFh2~Vh6_z;1WIF
zeZ=`4;U55owFAt5Zs6&C-3c9Oayz4@qsvT7`vN$8Aly{;3@7C#DmohOS{QD<#GFVx
zk>)cm#_RO)nSmnF{T;NL3!p!4@Fo~ZTBO&g`VJr_;8y<{`UQw#6)#4gsIk(7Bz*9O
zh-{m3&HKXzIrt-){NutER^vZ9FRpT#4tKqJD9~QIZ@V-}$c0f<j~@@9aqFgcb6?sM
zlV5q%h?YLhcjvuYhU}Tv(5rHrFk=Jjrsq6YV8;>}74@vuF?>&;alEsYAm#$@^4$8S
zl5*0>*Gx?(F97vl$v5*8TvF$@$)~`YgOa?B{qh-y(?+7Q?5ZEF=K_J?YKhWk9rHTk
zIViEEzN_57@N}7wB-Ck+WuP$Pd<lgh$2YpR?0kAa;pe&*m<lmne(W-tCzk>%6OcdR
zd=QEUm^KP3Do!0&cP6C&Ip&-~39M$q01=6~q39T|wwlM32ndlU=9m22@Dh9IoB!ed
z8#CWW$R76=D$GD39Cz~bP5?xYLx=bJ0hTIIwty^EaKb>)fwHpl3TUAU3OaTjL7;+d
z<01$)2I);8#ZmqB)!C>3Y#XmB_1+J9aw+}s3^DLt{}YV(4^AeJId*a52}G%2zT^Ve
z1h|@F7NhzSjQp?erB7;|z`fP_c)_&k1=>rH<{rPjRez-q5w;Kz=QGW+4{hj;quG~Z
z(5gd8mW>;m<3bBcQ_TPV^DsaX`|?=Se5(GDNs&*iB055Xt%U8O9GrHmi5y_+hpgju
zp`t+;E9SWDLj_@PCafkezpmBRx@xz7^csh`_c$EYYB9y(8K8@IvnE<1@TQ1k)R4B`
zOkZg)m?-D;pjKnl5uYnkGfTDX1y$e~MBL=6hk&nw9f6BbwJ-*>Psu+k8uXa+s(xTS
zA+e=Um!m)5Q|@)UIf<_W&?*TeDs5Lgq8-_N5T4yu3<71Ft_ECw)bHyX@nQt0O;6i}
zDE1=i_7MphNFwC0H{a4mk=q?NUjb<*uy6<NO5Cluw+-J4=@is6LtN)9+3}Hj-w~o_
zs}{zC2?|87q}}fFjGg2DpdSQO!=t7%*`tPi#sm&Ikk|(TS$F1ZnLvsr1rHD2(?*G;
z_DG1$+}NRTm5HvP?flhHJWC*0&()8t({c>*On0hpPv7ve&+c{mrzYcakOzd7T9(vE
zl6MrTSGul;a{PSPC={-OHJrxxZin|$l?UqDXdv{tlIn)zd=<}L3u9^_88$Suzbg4~
zXG!svpJ>L}viB#UHcrKg*Q492n<WE-L4VjHzHPnK?|rwM;E?!9KJI1H`@c*-4G7KJ
zr@#H)b*!@dgI38^?s+Z>vSfju0QeqM3ss;&hYBtE+@Z=NgJO@^5{VgB3`R4~4L&=A
zbUo~9aqZseAP&z+z1R6+Ru5scib_i3rmWz5o7`)-bdfB&OQauk^6svI`d*g=`W|qH
zgKZFYXy7zW@ZwLx$L&?!+GTGZU`$L7dN8*cN>;1J^*@y`fW=FW*1lN~##1`T)QS(t
ztxnxStVdmf^i5=Q)rVkx!*Bj2>o-osjQ4lP{_9MB<cqiQPApO6-NjIp5*<!W!fZ?4
z{(Tkd0}gr9_&1o7$5$*ye+}7ff?_dK;vz@p9>z<&VUHSMI+aFQ8f2&r-1$0}Lt^K+
zJo!@^9hqmEu2AtPm;3I>O13sbHC+$xobEY4vjRY(`Ywbz&;EC-k;9hNO$=V6v+?ft
z!riGRnwv1%$D2;MXUomrR^8sH!^XM>5l#L-&D(ijziAYUXDbpo$$|)P5ObYAu^Tq)
zx$b)N;GHyLT86FZ^1+n%zYyPdr%>52M7+mdqxF$6Jb0K+CzFUh6u1ESnLQB1)TB#v
zS^CJIzHuN)F2!w(@|G;{1*d=~!IvO(Dt1mXy~;m8p_&VXANvL}r>(qxZF3FFx&Z2Y
z=3;Z$$eK>n=GA(`E;8~uF2#jOJgnOT`$W*z!xFlFdAfRf)a|Gid%>Uewv_avTY=EF
zMsCnygb`QC)B{uNy)eM8K6_#|B@+>NKC+W8Hx)vQfs44@K!X1|?WkLgEb?_Y-z7Hq
z=v8{S*`#vPXmMFvc<!Z<8u%bhZWl%nOFFvTj%|aTHRSm5QVufiYLh-+<WF3e+}7Z9
zm(jC4u8*p_9?F-PCU!rXbJU*i^}yIL8H?s#W?^UI)U#ib2Wv4BYBa~@&e*IJTa@U{
zndlm@QW{ek009yX4h~3C1p-}A9qj~aKhbC9<mi`w`uUrqc0P2`y7b9G8<{^9Ic_k3
zT<TNtbTp>_L;QEHe?DyDBzDlnIvysz<-iQ;%TvCfLk{*QldSe79{q~;W~N9?F@KNX
zoVDv~s~pK%@Q4A9K#LttJctE+vC>ix`?We1|D6mfk56FdS8}QAC!NA}-<bgO%d4HG
zA(1HZ6nENpkYG{L`$HR_5aWJve2*m=@_F%Vjl#cg`)k@^NAYo8eu6=n6O`sklhIf5
zm^EU>JXMPWH<0uFnebxG+6rJ`$?Zo5GO$l9O47l!@H`S)aEm5>MzEZU8dmC2*l|4F
zSUJ1d)PvL3@kliEBm94*y#-KK?bj~;h={0&2#AQJ(h|}gN=UbKDD8uEHwpp@NQ1O=
zgLH=?Al+R8(j_hZUytwmJM+ys-^~A;a~wtoePr)_@B3bHt!rJ&QMM)Hb51OUMt<Ao
zV4~%X14xeu2Xm;V7HgZPku;8p1DByz5gOk5OTIqk?9r}->%kohfy}F0TchsFudv?z
zv9P3&FPv$l;nJ3N$UJZop{TZ7S6Q+(EK)ZeV;#2w2o&n2+34+nbwTDI78&`j88?AB
zeFzCOL~uW7ZyH(jypKP0<%d&Vi|e9(n~Ep;j}d{jCJKsbwe*PA8(w5xdM~e0=XL`b
z^udZYzS5dRhe$>9W~gR*u-=pdm+U}4EsxWe>3eC%nl%nEhsGBuIY#FS2Cy!-y!?BD
zx{Uesr58t!S{J1+`LvV?ls-X9KgHv_{b`iCfcQ%}O_Ey}&@q=eN63clJ|CQBzmtE^
z!782_3d#uYo{>7Ed;R9SK>U1rgGo5bh}Usb2Jmzg`C&&vG7M8qd?dPhy|^Rd^YZ?P
zUg2*1x7o*sQ?ATsCx<l5LS$aHTAWJ(0#^|H0>7RvT25d$)}6WR_T1=61;GF4-ilqq
zF18Q>0p15u!uXi4ORm?Q9REbV=2w3jHPylvtu1lEb=Ri#?o6?r3HAu%na{h1ji1n<
ziV@_irRLUrI^&?UxXrpb-y`X<XjDpnM_!NBAKqornVww`y)=o=A_#~<7A+b^iS`Bo
zLBO>zz@=UR2$`@IFsi;Gqig_^?En$@hwf2oZ>aRH<u;WA%80}IFBsin&?*nScR1P1
zp>}KDuWY8oV1iWBtu9nhE$MzkQH0ZO;jk&@)mv92HdFK=zlh_q6rBBfyAc$mi0XWU
zI~L|ncY93M%FPu9xSvEVq?Q}1W~=lyJ|w<L$U$_IB3h*CjdFr5-A+Y7XvdAVA3w++
zKAiVI0V!sEpu4*ruVhM$l~=xyg7k}y5i|NOM8Kr-r4jkCv-uB34cZ+*3s@A4Wo)p$
z`ir~oMMrBKnPK|RZWi(pGqcHQrS<w$1DO{J$P$wyy_GbR#CNjy`DlH_X|^~3WvAUD
zE26)~_!ZD)_J4gga!Z8jYcCa<wSy)egC--}@xk1Aq?tPJ9MN%IRY;^kVd`>$R^{&m
zsw~R)Vf{*fB_l%t*kbvtr#F0dD#WV$sUvcuc+~d9u!Z^J?F;VCl$n$<fto&v8eUDe
z%!~~Bul_K58wV5Sz$;{{JkuGFNbUPPvKG1N>&qJ(FzCn7!JQxnuL)Te3y4_wj_w5;
zUQY(<O<-m|haGq2=)`IE-QBTLYc}N!8Kb6FGraQLTuj;gA)IR%8umSx8ET9v^`>3!
zCUmcA7q(#0gpa+JGwlVibgbCu2J!G|vS31s3J@m!e=Uk@$?u--E%)q{-=<@R-7yV_
zX{GZMdyOLfcm3tZq0dO4L>W=b#-a;X*3xdOmz9}~d<5+Z-D@9@54V!ObcM!VMvc6D
zsir3jgkeH{r!Tm~`<7$zcX*uHIbxQ>V!tgLCG~$dMO7J(UVI-y|MoI$kGI>?m1O}6
zo?V^P<D=(uaq<7*0)(rxSx}lC-Yu%DeXHeaZ!rG`xBkbYpV<LdcM8jdLihHhPqp1b
zUR6wk3zE`i3nKG8j)6W=oI1{GT81C51=ZW@ku>3gM{hOMbl*0u8M)=#7XXk81kv_G
z-NK3|RJKJ7VXgD81~s)|hB<&>u*{H4)l4zHO$`_3PyJ`VD_S(1Ipj#+>d5hFayvd{
zJusi{Dy|_lS@YAQBe_y0e)*lydEi5?U>LD*M<eRUevQ*>z-4}U+ev|ni)v$P;}dgd
z3;({*s1U=q^OvRf<0vHoZ(j5N`eRoA-hGXzNc-~BM>9d^R?GP0<9}u+VeUq|E;b!A
zvl`I@1SF#JD?jgcs0<Oj7v;V<-3_IDkID!mA|k#3jgYsHuVvQiH|U}F<!f=Zd?b<9
zeNt+-ndni6_~X)vX?(DCx$%Qy(~*T9Mef1o((v@u{6W>9@m!}D!g|N`MgAqRX+NQ6
zq%14Gh*cM{ba8lt_4i8F?BxDJ1^UC=<&jaglB$mox^71cyqEiCoBZYvTvn{7o$cHf
z7~|g{m@P&-$4Gd8YuNAcf{p>OE!fS+BOt8qOBV5~JdB*!fSVhvc`cW(YVep42Fd?%
zt9KEgh5^;ez(k<Me_A+VP%<X^$Pxu=;K=c7M3F_jdsw~G{z$&Vx{IJ>a`!|1&hLgv
z@>~uiPn}V<8w?0*GoL7nwC7D=kPL&bJf&<y$*^Cr^3R=LljDQ{^}{H-!(OvnC(1}8
zU$w&v^_h)vJx)0%iE>YM$_XKcuuEh5MT4$-a+%6Zz&{X44q@mhUbo^lOkpcFv`0#7
zm=|^QRqV7qS5cwd8ZVP(VrDMfT?&;Doh~*C;D5lU#UmuVIg%2yb9i!+^zgFHjS>xy
z#jqPj!lYB_sq1k(ItShHOqX|lc49fpM9UQ~o%0z9A9TetofykSH;}SDEnl_UslH|q
z!7|giViVg&+jGw)8};9&9hX#i*xINMU-il83fg?i3!^;Rk!BWHECW5qHqh=v_U_)k
z{_Ynbx@}q&7LktYqZF?Ei9Y7vX1{1_XwYrVH0TxN<b*v8^(V&mTY%|oRkgL+A4MPR
z^G3~VMR&Bf6YbSoG;UcOR-c=nvf<tP3nk%nb0_dR%vD%&mgo;F5in|t&YnNjby=aZ
zv9*B$rPRP%P_2yum;^)&QusNeuJ9z{@)eFqgGjqvCfjWIIZPE(RJ;wXfcSUr3?2PZ
zp#7dG;M!|ihVH2oe3<X47&(P|=$S#D&2cj}i>yQad%+_J(lw$`UcTe{wHo3~%JAM4
zu^Z5B{yAcKh|}QMuG(>{yWglIqG-bPkh~3tv4+p;ka69y8Zr9Vh`}{CW$hqvi@<$@
z9x1gfNERt;92T?|<?+--_Aa;A4b3t`r&fo_8k^^n)?4cX-kybL*Sy%nLrE&+K1X)7
z``;3LsqW|Cwmqa~c<a8ypc$-+K0rSE?(|;E@vK!DGE9V6VR3QNFJDps`=Vl^3+l>a
zRkk!Ihg+G&#bGcQt=MdY35<<ZmG*x3gXZagVg`dEb-D4UCEqCifjQA<0|y|WMGese
z9Xvu7{qMIcYJPvaF~8^|>^6^pnHf+^|IyL$5NaOr@x%^W(_Qd6RMS$8;}$hEyF;7}
zjWsZy8<fFO-yCdCNk4gV3AMxo1WHhD1m0kc!$u0QSCV@Cnr0i<_QRs1qpeCNGGy<g
zc?cCV?85w(WYyej=YQ_ma1Vaobo6s9=8JDIp0VkAP=Rni-g`e`3~dnmg5BGiyIou|
z>%&)MyN)&+Z0XMi;&z`;>nh2LPmfh6Ku}@0Vm}h1|CshoABmn@2b6@;zt2BE*k5k6
zef1LN5VdkJpEgij9#c}v$O}TZlAoEz=hcB!@oI|g5GijVnM3NEulZVGF^2O^3mx}7
z=YweWT-K1{iW(Y{@JOJ38W9ym9l?O%{REu56>owR8`cq~Z79?O7Hno6YQWfm%M<5v
zz#$U<PxPmVs&X$U`+efJegMOzA(14pwvf>Qqy{E!L-!a2@BlnDZ8*&DyKEo1{W<dh
zqy9!%c!hRx-vz9y2}p5R@0#kv&>8ryb-{*;_z@S^-S9TM!MG8qle~TL*dNAK)6u<$
zC@+0_1h2z~#H_ww$8o9|l+@;kL_lga9IgYK;I~WNGxyW&V@4CM8=e!Aw79b)S}oy7
zR-0WDy=Nu76*4chUfn~OWO7`3u|Iu=g-rOG(Z;pY@5^|qV&|pX)gN(IXEi1UJ`kuL
z$}NyT>L|ti{&Tn%$&#Si)vS?U={Bom<W89@C!IafR60F1=DdbJUA<^Ct?|W65|0|l
zpdWljijsfD%Gm@_N#cT3wMNuh=Y!^0mOpNSgGdgC8lBQ@!FQrwl&E0g+r6fx-J*t!
zCXK*>6+=MaPeG`9eY#HT6)O&muBW4?*M}nG`e>2&+qY;&tS8z1FzUd?+yn;ka1)mn
zrn>Ki4l)(-g<ZKpFvw%^m=JWTX#`GiFyC$aHQ7l?W=zC8ekz@lMw@*+VL=)!mYM!a
z_OR;B7Q;(5F)Hi^vN&BoD8hHJJan?z=<&hs$bfjs^(3LGDR-WasNU&FjA1GUFIr=D
zh#|;UlRfpY_DY0v?JwRrtxV>L7!m1!82M?}V8yX9;29^ntkS9GslS=ovix1<!k_5E
zG0IJp+NW+pKnKMaEA9wmEQgx2QU}9RHr#)>VVEX9Fkud;EWu9mbRDdK`pwPh8(Fl%
z?BK|-3b0`?AyMVz*StNechSSh(IHh;)r^QqAU(kFU5zRmDkxj3jeks1;pjP8%$^^0
zpAdytc9q;=bi?S*q}_pro88&t0?nP<jGe~g<KYMi&SmF~b_$@bp<~}b-`bdo6Zl9^
zf2*U|PFh+lKa8;K^*h!K<=k!Gr>9FPH@MfzDq{9`Gb)dy$u`>;-alC?@jO<U{7RlJ
zetAo9*Hhblf5@;5nf_W$QSphGSThUrz7v}QQ4lp#0Di}&bDb{e_JMN+*{D3#++`I<
zqJJ4AJ_FqgHlayLdRrVukbGeKUG{6T)<jDghe1#B<Mjp$HaEM}*C(sO1DNaoP$oQo
zaL?&*JQ-11s%~l`pOPT(zMQ%1A^v-X#r;qudoOS%E&wu=F5mighr4un^Yie%Pi30Z
zXZC)q)R*p>lp9aBCbg2}YH89_kuoa_oe|v|ozMDKFk)@q>abpE&;Mxma1MyihrvR-
zJ1vZez?Jd$G$ZF&jgcD@Ulv}_AvXBz6b`gNpC9@_6O5wT?T0&W@{&qQ_BB3^65Da9
zld`=k83v?P>ov}6nNgS_5HX<7vhzZdK(Np`_rqr&e;rI%njGycT;n=TR`7?(=;1J?
z!wN!%?X$y~)-$(`C`OrB_V>@o+}SM>+hDj+)!D%W-T_Ln!NK9+)lCL4Auv!8HU>m6
zVjdpnl1UG3SrgVdq6d-o-it32n3{OYmCx&Eg|7t^`}ITNcB#J1zgsq}O{VygmHQI5
zkC8@9G(KCZCG)us-uLg?V@v@+T+k?Xk7gx^GAh^@#V%1585x;gY&~lRJc7!n4de2x
ze<#=k{gJh*rL`gUicc$?ha{vIg4GTK6HtDK!}xl-Hb!l!Q}PFblrG?x9WRJ|@@yU=
z^ClI;D6ufS!$^iq*`XB^SaH@HdfbBlmf8j=39(p#J`T;8-7xj>&z9#!2Q<?E+38@O
zFL#9*aj;)d>~YkLiNhJsq>HmHPhs}(VuuY)7`EC091qka$B5Q<O)nua)F?wfpFIDS
z%7E!Y9}yi5yb1@Dm4(2_ad=tAp_rz3q@r`vF87Q9SZPHz>-P96JlnqCB8a16oPWJN
z2j5LAkmow}sUQIeQ5xls)b4f26sIWWemd$plnPN<wtG2@jB#IiF<p7g>Y&Cus-)Bs
zMp8rAaJn1YI`*eD<Euz;MmFH;h~7*`OBjaHIdgu%1sN>F?6tN0V8~&@vhbecu9ZiF
z-mBxjPdw7kpm;DB%MP}JWo8+R{9im>7Va8_Z9$$W5eC!qj7Pt`v?l2WhXPM{2ZWMA
z-v?Tb!Ly{WQs_k)!6l$6qf9fw>au|TcT5i$8RL>)VN={z+zIswS|%pJhwXE6`0#;V
z<f$m%z~wYGJ@?3QHY2bB3PYVqG4?Cus*77xnR5!eKD(^_MXPqe<MU=U9$B--K%3ld
z=K3|nuBjkRkCy!SerDjpjG3qL9F3mRvK<}C*81@>uQX6Ge3Gu;`hEV9m^XuRU@_ep
za?`tB%m?_7t>;8J&Bwprx2%f{+1@=lLy3}MVE`MqVMhKjp=gv5OxCQbs%q6yy$a)&
zOvj3;puaJ5AbXuHiiJ1yYjm&EfA$JWMjSE{>7!}z4{k>`Fk_A1C=V((1dN*BBPR_F
z4MW1iMYxGkLoP~8<RQ!G05+{W6vtvK4M3L=YGV%Hu}v~j14LmuVC(d6N65bW;IA?A
z%?3y38)NQ~Z$PUq#3!j{nr<PVptV^woTR$Q%*61X>aR`3>K;iiA!6h01D$rXVixIV
z+Fezaj1M6BKG?a=j=V#Maru7cq{IwC7sAP8!l67Ws&A|7U1_Q2avyupT~5&vcUIFu
z{F;=2ZEPa$5P&tTX50^x`%FK&Fcw|W5J|>?&<UuJ;YmrJ+C4R@xm1jdLC{Rm4>cv|
z?}R(Hrm`mp5Q_=edv^T$=i%z|7VbX@$$O|8&mEpX4_&6^Onv(w2nJt4DQL1BWElND
z#$U6>EQ9bBOuyxytB1J@&<hP|8F3lOKClBc>Rjw6KQVNmdw-p0q6jyNfg0HScxwzT
z?t9|hj1y7+NS%b${vS!6=P}y`Eg@P1J`dOgNTMfXC5@;6b+Mv+iyj|2?sLC;mGE#h
zcigBbzcgO8+;mLyV$XOz-(jaY(G5_}E5fQ_LBJfbJxWPGWi=umj}`ts8O5kBXv2MZ
zT2sci>uGRC-%n{YdiOqt4(EF-G8c1Nz~X&H{V=;kSMg(AdRkyjhP+AtueKl$ckmGw
z6*Y>%K^&SAa&j#;>RbZBt*Pu@&<5PT#VWc)1+M1Qqd^_#fc@X>f%`J(Rr-hp5TLS8
zQaB%#NkjEt%{pm1QIu;S?((d1LXdX~17EH6plk$LA2iEbRPepD5O<G^h&X-O;Ro6s
z(8$G;-O4?5FHM5-X-^3++gv}z`gDn3j|Jh!`T7YuX2$gP716L4+rJ;tB?{i@Idf!P
z3wy$Sx@{CMQmk3>yCR|ekdc8vYD#zW!uaNF?4gxtRA@+Ll?hk3s~O9A6I!FkN>^Kz
zIqpOU{ixlSh-fa0WL}(~@5zR3^)99-tyG^4-y!)qm!}HBtJJgF&=I|!;e4&CqbMyo
z$=@#AseKYIF1&!&W<1YTm9oNLt48Ja*(#&2%PW7<9;p9xoXYzZ{M7ItmxM(wi3_0S
zeaqTV&?lMSJ%Z7#Y3b?KFebIqT!1vV*jLJZW#0wk7HKD3<P}ug5+)|03t`f2x69f9
zh=Gt5<6)vhZ0S^|3l<3E*+F(fHCJGfmr*b&81n2?i(S`6J6>zvgil?IK5^4{=?-2f
zf>d7Kci`iiHx3~=0Q42y#5M&3nafm%A2-o_tVN^mwCY@)#`)ytdX)jID}8PyvdvF>
zs6J@tQ)1VElm5nW<(ZvXZ2NE+Ydb$G1E@~tnBDK1Ae=d4*gUtn!>sLb`^xhnbvw<?
z)d9R9DY4;z-MIrrMLdSrpV*^-!I0iMd4zz-76-^9mw(Hi3nE}Vj-W;LN`~_EFLG6C
zlZuJ8&Qpj?!7vpwqyoZaQ#}q3h~a$+@|c;eXL@WEQr!EZ7=P)na^V2vgpBCAGFrd{
zTyn`UrT(ax7)c8Y3+-1!5Xk-ICEpDp=C(pL3G&PlF+=zp*F|s$fiTqeA|mjiNzJ8>
zR3i$K0F<^h#7>a33MOs&kKJsTxybyahU&ytR1CFYb9+ESN`8lcpxE=m18xo`AS%#`
zJ|;x1rmZnHF}E-fDFELG*Z1h;8iu!&_!p?H2Dgfk<0$CXK(NefWHj3`)n$@eaOGzn
zHC>k3M@yC5mDt#)BZCo@yCo}j5)u#!x@mn$Qf2NGLo>!+Zqic|o9JIdqD8l2V877P
z;0g9@NIcC!9<v8WxlcRG-3q(uvWMhA)sSWg)>ALy0!&|(^?uV|1o&8C-Dda4l{1Z!
zLIs|K2s-7_YWqh}{NgwIs+i|f0L^%ItAp=>eM5u(44wo)t@PX8Lr8(eZ-qs-1U{By
zhg<Eu|EFLdPYq0ACTKTb%E>KAfg%CC9F+S94oSp)0jQr9>ei_j8(sr~QATxjEWC^$
zh#fM&wuV!_2M!sFNq4-s0B01(+vTA0f7q&eIA`oSFm_^KrPq`_Nb~33AwBMUg}-6z
z6`&mlG^a#R0v8!TQq|ltMXv2421OXhq*#;PtQ#;s0VBOo{J!B6o_NiC<Mv-|5*9Q`
z@^zB;dZdaK=MF;Bgd$>WuDsogJ!Fcbp2-TF>d(D@p?QL*w@#;^=@`7(|9(u!QXeo~
zvxB=2F40TsFO05L8n-<|#6QP--Av#%#cxEMPI^oJ^f%tn8-#gyc;(7`kF#Z~g-Y1>
z`I%>I&tW7sGit!;@2gi@n_fH)iI102QtDUiI0UjPY4BBmEs$kU`f<sRgc@$&+J5_1
zl=`ieJ)D1%jCg8D6PBQ^Xne3f_UqTLkf^8@Xh-RKuBLORV#@oL;13{0067OA=wdHd
z2eaWlWh=G@+V}D5YUZo9*+8(h(85oM1#_a{^h;5%2Y=~S6lpHL%0e3K_u<3!^aNKW
zC0qyqZ;oeDN(NaFDI^7oaua6&*lc?JusKMtznmjZT52UkU#+D--*|Y7zp*k{$mIOX
zBLrbzwt%NzT~taFZ@*;YPH5B0_OaWy+cA_PrLSceCF3WGd1mm}Q=1-H+DI2}t}_HP
zb&qFDvTlThtCu+np;~r4v518SVll98Yb+)L1DM}PDmBrsMIm>nea$|vco^889Sch1
z@3a`AFu1hAV?f?rt6PW}A{VZF*XG9G^iv{g;p!XDgw0)U<(nkD5%NN3rPC9w)zi&y
z(?@U4zgHJo49<Y<3ue{DQcqR21#l9%xqD)K_n2H)#D--F(*rlGRje<#%>Dp!-2Mrx
zo%pk}nqJE&Q=iw_NU=?K%w2Q3EkM%7JQ@bXPNHW%%Lf>{c@Eq}5GJ}Lp7aR#H2~Ca
z;xYas)oi$rX-uvapD$@)F;y}9@hc9Mlgp+YM4L%G;lMmdDh>>7)~Dd%(V*iHAOIg&
zx-z_Cr_K=NVidFrQH~hc28#-=>ORrS9*(XOVWsJcSC71LZT~ueC)_Xf)16?PARrgK
z9d$lg(AL6fS`ha21^0^17*pMzk_|nHwSoDDB-n)YTnBms+{Y%%IEW0l9V#)7LhF%1
zYC^N_?bC|8yQ8u8i1q9-lhz2s@w!uthl1+ZjY7=)x9&OFb@BR^cj!0YY&v4#BCFFB
zqs2`gC;K^ban5z0%UmCX_-F*Up(4+2HXH~9^H3NR1sj<q_#PHUc3z%JY$otOUvp}u
zdOyUm4VB|k)TjB25L~0`U<FJCr0s-Wn{DBJ1$w)>x<5>k0igCV;YSxX(lt`2QfJZD
z4rEyh7omDgLoig8BDVdy*IMfyHO$woi!3B@B}~10Hj<awyOMI&VY#Dw@Ep9%%{Nwr
zb>dCsJ5rp)w6klcbpYJs1W7g5<}xi*6x+D;Xn9S2Ht9V*RM*4r`mjM;n6{g1dFgZC
zeW$%-ze<<UFX0aPEcDR3mu&XqZ38q_0mr51uDRvEI0Pj?Hg>=m0H|ly>Q98@;z<eK
zznMRcC?zgx$T7%V3dtJy&M#PTTXCn1MXIkk0Nt?7@Y`$7RwD}i{9*a>ZLnz@LC))m
zoQn5qa@_AHo;g3<tuQy495bPtud0xJurL<AjV_|yu#VTyNbta`FoOq$2`x`#-a_QI
zP6<mytq53Nnu^O{UF-j<eY$zF%z->je(Xy89AJEAi{xM7ZJ_K)Wp}gdfOxu^Z#ZP$
z>8^4?mnl0?a^S?%7*2my@cb7=O#TYb0#3)L=_OPwp!NgzA09jPCpFyY9_+}7=gqN~
zVTYlJ)b#YcR2%&?@(Jz0bLf0@zv6Meqoa`<Aa}h6n{GgnA0wHdo4<27*)nE3^D=4R
z3N_mZcD=;W7|&9?cgIGC)>1bRK?Lh#4_gb<7f|<|Moq)E<9nZw2Z;})GMck-7^Dbx
z2V}6sIfOCaJ+0JcdYT;4Z#54)prwCpAIL<f9v9a-WUlT1OHsrj$}25ZX>;RF=r(xG
z$S%%JtUB(Sge>ez=m0t0Wgt5M)OfkC>?5ZPk@Q6cK~epePc&DPf^!x#+$RR(xVQ3~
zBXU(3&i>3-y2r>++Z=8#B=T(;4hJx!Jq9o}hQsxMj*dhb6@+oFJGiE@vA~|Fn8=R^
zT-`tR!J2tbEr|=LBHg^%k8i!-wr|T4q40tyreBm5SE=F!VG8ox`N^ii@0zV$Bi%Qw
ziTdnc=7+BVB(=A+_<Qn#a~6t!T5K=zBk$aq3~@J4IB`38clWmwKMIpY$ug#0eFj0^
z-ku#ySP&H#7f%Ixm}+5hF>b?5q2c~&O+nxG2cyjITNBw-tP+3ANbJz}qDY-_!&93}
z_LCPRpMU0qt*wpC1gp=9*y-;eCFk0{E}`Pl+EVKqXEWM9&kIxYr&{(C#@V)-f>Axu
z(4=+kgq<kJ(AZc>Q&UPqW9%C9y+KoUscQ+)!&u;c<R%CuTUduksq~0}1S!Z}>vV&A
zmYTK>$qqUG%c~JBB|YCIv_9L{wiz>$NqBr|ZGgG2##z~7V=M>~4q?31P?V|67U*6q
z(<|SZ^X*bONew_VF)|{tPJ-mbxGNSkM#ezY#`>yc(eOBI4KR-LiMbHB0(k<qEIH9&
zj7-J==hYuC9s{r7_mSHJVP3I+fyk%c{x(TMccI@T%%}s_^mIm9e!FLD#HxqOOd&Vr
zl6S{aE0S~I@btzXkg@1Zlf)zEmESN8x(gAVcxqAKrHZdF6pF=YQnZ=>a;-6kBd|dC
zQHFZpm1R+WpDA<<pnE@gn%pa?cSHJL-;8;al2yV}4&K=_Qg@WwA^2uKUb=F-0##jR
zW|jxNEHOy_D<&l6<OE+XSip4azX4vC$BG><{N#Ig{9(K~3|)s}dFiR(mF{>oUD!`J
zJ~LN<b+i9mN$D-6<g-pn`1E*%B_!Ygz<`QYXzv}CazPl@K(Cs+?-N-)gdGU|J;o(R
zK7}TB4?y@9>aM0o?jQbx*ttL=-#&@;Br6!a2GGbs`ys1D8+uhzRdZPgvA}8rPL3ja
zKxA7p^^u_ie#C3eiEBJb)o2gH+Xl1jL*>$aOHZSWjw146_McJxgM8qu%61Cj5;d*Z
z0RvM6Mp(ifWKFcNiWT{yL;ee^l^&;UaK-yct?t$q)8h}>mR|VSevn|mTf_d3L7EnI
zPUuw!`g7)hb0^}O|JpEtnF0<bIuXYfEe__-eV0gQxrT~HTI%7QiLFCGl-C%p?46u}
ztuWLeFmn1235<L&4l~4OzVFkHOoCjdU9_2m7xW>3`D<q>(`2L$UkQviP|<^=pkXF>
zWG$!$TO$glkV9k&WF?A??u*X<?4mkg>wwi+5xhBzMI(s|c@hZJ0N?DqU|@wPvJ6Eb
zQ=w01-!wJUP(@3XphOcbQ4~+z4qTNQQHLULnWikn?^4(XRQB+)p`B|8c_yO0wnCW*
z(iQM#SDV-NKRDk+qmQI#`khUJo)8#k3yE0I5}*Dn4O;wfOS`j&Cy_g>e>FS!H`WLf
zOu^u^5o8Y`)6?XjC9SjX+P_3joK>-iku#nwm7GeiOozvIAej$0kks>x9LN7g2&B}w
z+|AunK<XM<3(ie$ieVVeqd0gYmY23Enty}pf9(z2OZ{Ztn)wB6f8L)tQyXjRr#lx}
zH~=LnsV6g{?^T&57KVtnt)Ebi0xh~|$2CX;Ipa10-MTz_{@0TEBUTPU&|^^JnBZjh
z@RJ6>ML-;;#fM_j(|c@$ScF)<LS6CqD<()rkV56Y5mx$VJkzp&?H6&hFgZDazktYh
zix2xQ0vO&D6okhD7{bp`;VP<?(Su4W4R$F}5NKx!b-BeU(E4;NJix!6tdtGf(y$HC
z-gud=xa0okc?@vJQk9oX$C5_1iQ$43pH|5v^R!ZCw$$3&^b@wh{Si%s$x9*-q!W^&
zPgH^y<vC_r*7oNdYA($Yu|Ov<7(0)K3;BJw`0M|HvVcTe1Y#vH3FQ&A??B2ONG2p$
zKw?9SuYZ1egc_j&SgA$rwm4JePqc+4t@78iV<t&dY;55`Uqht!VKxFT3<s?i`Yx@L
z9z$7;&3;t{<TNz5<P*ib?DQqV($IyUA>Uf%jZj8Wd@xcu$28QS#=-o~q3tDN8vp|G
zVU&u{E_b*qRNOS-*h<7gOp4PQ%K`iMHhIE}|J)}iCToD_ns%DM38)LOwie+w7T#PX
zRVa4`bQZcB0i7jlnCXG6!T;-v&+<uvU{!|;bUHx83eKh5P8&@iF%Pc)^a~?n<6`6x
zJyZ4a10&C32ChnNIdqAAyc=F0htW=)I+D+6FDWt5?wp*VPWii9nD|sUGBWmFE^~k`
zCS7Q!7HDCqr`=$;7zm`l+1@s*AjQIf?q7YT{%d=!?k1?F;cxf4w+?rq-Mz@L4HxVe
z5UGGG$O<d7pgao_9iVgHe6rOL3=NTntSkUbPJ5h~qehJ~bifgW6dwjt&1N`%_rJ}c
zTl)!S(Xt;a8y5+$QxwJZKa6Zl_z-t6W9wy?_E`c9-M|U}CeP~nMv!Ig1{eO9ls2lw
zQRd1h0>k?a)aJmSz2;2%2IF2&=IcsOUVYXvrM8whybb?u<#!8TeGOko$u(gjxSCBg
zC`HQ2Vb5BD?G^#<;XRmM2mPFLeb2~(_<Uh%D<}mpsDJTl3#Z{)<-d@*@qJhqz3d?A
zq5&#JA9MKs2%N`BvWCZyiwcJbi`L>RT<=E{ARo2j8EjUh{fz!pSV4t=A@D_Z`!+d|
zqLWy1v66~rUTuhv`qvbO&>Heus_`3rE+S6~-lbsvddgbie*VI^(-61Kh$-gRiN>QF
zqbD(E+Qz<Ro?|^84hGzG!DU{*ww=(Ph6Xd^Yvd1W2AeG`)VnK7#-TMmKJDQQV>%NG
ziHNWpyTawBeCrWCJ*ET&{+T5Y0@0-+CG`QSI+x4MXX<sD{B9jyO^3984H~Df(a!VT
z&dJIOnX*u8@HhhuL(M<{I0TVnuZKO)cgcXkQ_x}q+{|JSQOf_kvDA}PXrDyHH^b)T
zmYlh&QmOta17wfki^yKshBv{6n;3pSe@BoeJ+S09ubcE`rABGB=6Dca)zX%mn4XR<
zv~@mo?6oS)aX|B+*fDSm;P|sSm2--tGe~yn8p2Dwan=e-2l5cx#<CcQH`eFSgzZ<w
zvc3F<&Zt|f4iZp>MS)|lv2!{WY>w1=KU}HL#pPi1!ntuNi?h?%HubuMMTHjZUa~-X
zPw_{7eF}jMmQ+H~9d`!aG|q4TiC&GO1ioO+l<rdR(hUjVPOu1&23<p4nRb`4_&1cs
zSE95mu8lelJy%<S=GaRU;8Nl7R@s|>KxZ!_qb!-{X%X-d6wG75C<x!;|1I2De*2*~
zy7nc1&KVJfM}CPwBOH6pNd=;~*(QiH!jdrtFkof5w4}kUWd3de=qeU*L&*JxXUP^x
zTLQsDkEnfLW&s%ki1@5nA~60J38-IPrsZ*>dox?;Q$H_sgfKqbX?9kY1Yp%e7>V5p
zd~0ik4cO}GtIK_<tI$oCAn${5lPC>3dDmcSec`^@YRR0Ah)Uwy=k_c<tE6vNZK@0j
zec=q3Tzl4k>o7}f)#pzuTIkk2vHC;W$-`Gi8A&|E*kc<9Uxx1>vHe8S)6;Egrq0uA
zM2bW!hLA(e0e2xYiv3km^dJazKM^ouO!YpO#o+2tuGy)qlZ{O=bpPbT(F!6H(!I;1
z-3z*0IdEemRf6GdX&{1)TxgFh^N^*zYXU@|bA-x((x|<_*r6<&ncS?zb?qz)Vt2<X
z?ei1AcsY|syZW3So9oT}3-+@o#7Ez(8MIlz%fssL%V$c<$V5z8+!}L*_jl*c9rQFD
za`H<E9I7PugBX|`z+o|2`Pm&x8|ROtfW<HXT{aPki3Ly$F@@v2I$ma0w^|EC`a#g!
z(TpyX7T!syY-w($V`75vukFHdCV0zd{BWSxRS8q#X2Y#E#|zN(hEAg$`KjG)9f}=U
z%RoEw9J!}Y-wM?ZIXv+;&8p_r5y>1F`hKXRAFE^f*ahKb-1RunIgCKOh9qd{^Am4&
zs18Z)TqT{6(vS3Xo5<O5;)gF-U-*3pqoBRVI>+qAj_gsfX1=WPCEsyikH9(F*Sy)i
z;czk&yo#o}a5#22Fj_ff0koq(g%X_N^@kG2vpHJanPkWHH|l%7L-no{6yC>cOQYKi
z>91=w>Eoz{AP~YqKs4FZn_|MJY4?43y8PZ*N1j~hY)_d)&`rH!+9g-B3_0}kp!tJx
zrrZit0c)xqHgfJ;R3^(6?#s&Nw%_&DpI!BCo$rG=)wzI2=Xze8_gTo5(F^U;;i^pm
z$)OZLQTz2##fnLr08*Fb<S)Y`5x4(dnOZKMaD)0Z%d-AR9B}<uHj|God&{>s7DYS3
z8x#UuDC>=JvmbN;Q>WtnZL733Yn_T<6HPIk=p_KAh=gJ-Kwx+V{lY<8v2+OdC5k$z
z@T(S60yu!iF<pya_ybFecBl4c;t4Xc=y7|rKTw6LW2RJ@>R!Yl5DHgvM=zN!sy7#0
zFMvIQKzQB#FFyc*_}MHV0dFhFHAFwUW1$}waStesbb#yt%r^*PQ)eduA3l`Ytti=V
zPHMm;t}H-B@U>i>nWCj(hmBX;=VCP}Xllk%%O^4fG|$cGC7lZb#h0FeA+Nj}8ktcI
zW3yY@B!C{%3km5#Aw9)&V&RyXBcCON37U$Rx)We(d;$=9I1GQ_$R_X(HE=(p6~qJa
zZmPB{m8CStX=f81^cr|-BD|)@i}(B`q2cK5cB{~NAM6tGSFf@GyfJcv(xdN7Lf?)B
zRv<pSgkWkOmY56bnJJUw9j@m`OF?S7&dtD}5qEa3DstM^HvkiJezw11>EKXaWV_TY
z;g~rMhbIU+2RnYWw_hJUH#p-8%;3Nfgdm6oP+v2k;t;`cYXL<7Gbopx9QrM@^|6g#
z+~VStDY}e6uq=IhcLPpWG*pCP6fI57Q~{lG1}zttT7UXXZCn?;En|>vZkyj0t!fr3
zfQS+ZO~VjR-611m&K=5u51WFr89g(zrIlyb;waC8R=r!5%l_(}FB>*z(XqpX5@8b?
z!;cDUQ($3#SX6ZA7n$z-Y6G`aLCJB-hhsK>TaqCH8-dX8hQiz+$aGxnu}Q)X3P;6z
zRZ|wtt*vWd+ZDydeW8b>?8Vo&%N*qd^(UKr&}>gdO?_MeoIuIwM9YD}W36x?34?U;
ztI101W2k;_&3IIuw>CF(#IXE^%Lnq4216pZH8TucE-=lKNSB#6Ek0lw;p#BG#=eVV
z1pT$NlvJ4A3KE)DQNHz*COk;ze{;mqLr8Ng&u9jXlyvE640BJQTix%9^Z^HQw&~cQ
zEGPHzBbgBCmqO07X#8CRg)vYsM0Ip7L)_O3yH!P7I}b+6(>;7B0+YfhPGl;lHXEj#
zmlDohLvZo%%q?rTefa`)%=Z4C7wYELtBEayzKSAkqbM3obaW;#_qsl@G&R&woF%~d
zgd*hAq3>Uu79!Qyfd@u(cffS{gye>Sf`*vbJHd-n+q1d54GM5qSWUg~s`p0?f{(m)
z{RQ!amYx*7A=1;vKp-jua|^1aK%*9W6lgMFQU{mSa!*)*P&5GUbSPJixhtM$z|$x-
z`OG{}O@jUexZd{Xr_TG(o<4q-?9_1ogiyt(WXzB=j>EW<5MJ6N3W_HyD=Xl0L%`s~
zo|-&}t$d4GX*oDP`@o(**%As(1B;OZ0xGOA`nz=0P*tR-r&p4blk2@^p}cNsbaA>Q
zv|rJ1CJP^nqIBjbz2}<@Q0`iKv@UYztWXZ#OPaLWyPN|S3W{WnkCJIPB$p7vosjOU
z!lrz#pg`XA*beTp4rIgIJ39j%238jU>>01LPEGVUUO{@E%)q7@`m;U1AKh>&1{xZ@
z0R5pXhYn~1kAHdT4O=eG?IPC});0&?b<|%?T2#OnoKkiOCtw3|cCft1pQA#2)?jl|
zc-;v;0L*hFjLuC4@3jmYRXIt}J!LA-Wvs@r5G2di8*mjgEG!DZDAk>ufLuiu%$XlI
zFflep0yTtUMTFKJHk4-h>zF7;t;Lz)kV8I#R^jSB@=+kD0lpz=ZEal*EEf(4t~`aF
zZpUODe^!^g*xIO=(IrkKZN@;ug?o8O4kpy4NJ4}>WAwkXp<=>lD4Wa-ABJB}DC3Vn
zGzOr~0s^6nPKo;C_XCvq75xE2V!;`cZWx<F@&)aSN{!_J>9^?@{aCPhmzM~{jf&%;
zLXSJe^Or(#$Tna*A=vMNv84{D8}AAkSi!pf6j}&@TAH>GTRnsf50eXZSz)%_uUE-L
zYpw~WHskYwnWyD-SpSqEmyiR^BRQZk3uh$?<lH8_SOGX0#*^nyh8Xu1+yjJK@2c(B
zG9b>yAlV9+7wm)TmSM@*V1=dL!s6m8P}=+8rqMpCIe(Z!-F;{%g`|{>%phQ?43eu8
z6@{C1htr>Vj}JC-A=<h6`eZpZm>q061c3X?eIk$43RGZ~d(C<h_%gw8&{9#sI2!v%
zQn%kvNQ69Uu^(di{PjuXQi4an@a$Lm#bAdHL%d}7>&I>Tt!W)#qx0JQ@^6SASV4s>
zrNF>2AN}$KOlf;to1(FChQsFM3bUHJdX{dT%gTG2L<MjP<rY&pzzwZ~JoS?DYK~ly
zR)a?!B$W>(fewQ@hg3STxav~^lCZYmYc95cx|#zd1{ZK709K^{<>M6uPnuOWMvZIT
z@jR&zappqb6|dW$FLH_e*?Q+Y(W{eHB_BR~AmWY{HpM}`^pwWOSV2S_fIMmt@z|t+
za=B5FI=ev&))Yt-5ph{)3m<`I`5k<GT5zz`;gaxPo&%+De@W<K%(J%0pmjCIBvBE(
zX{T`s08w!0K37&&c9gfi=Dd2U5$yrVx~{z7Q5uLRW>;2nm%ScS0w)YF1B;dqlFgfB
zg4)nX9X5n~4#62KD=WM8-1pC754}t&{%*dnKv1_|?xlbLRnSz=#ibgk1@^Pwu3VIZ
z+8E?hUN%797yt#F6`^S;%|n-bmKQ5<Bw@#3i`-)3ar*rkjBx=(LVPc3AS5PnoSeOI
zw6-pS!-<kAxk54z6>U{O^&n&THZ1qK9A$M8ioc=ACmB}IaDM#R9asWUX<n|O;PgpU
zKt8xzXlMhvG(#!NEN29>w1Y_nh$g3FiTRxr!Xm&nmu@Y_Ssq-RL5viZ206wL{UXqO
zKOSj_qJ$E*Pkw^1`QPJ~0!foZ*xbDL-8;ZXkG6ZInb5P4f#RqSI@Kx&PDmUlyGQdm
zESyBREu6!DZ=bb7KZR%#j3AKN-jyZ@yZ`eBKSLZCmIkxoF*f+DetM;?7Gey2zsuB}
zC~&`YtLc_do;)6vPIYTmPGQ6O_{E`FL6vy_Ej54o69`CVFKP;zfw1CSH-RRKhKjSu
zucKtzAd_23aGbt2ewZnA2E0_6kY@et{cSjP)RN;s7-~EKfCCA%+kK<~45#eHodg2b
zO{YI*uo^&^NkesSex8H`qzr~!UNlTpSQY{!`dF<a_ceev)!Qxi*M`K&2me`Yn7r{B
z&N6bk)*1YB5>^ug@BKRq9s2;Fyn==ndtO~7r7!{}?JSKF<D<fUR5w3tR>OSNj9r~K
zyq@Q^o)cii;YToC;7$TjC;@Z`AeJE=42Imw3_t=%(e5(q4hkGj;wi{gm&X4A6I1H4
zj|86rO;NzCkOLsTy{ju<+4GInKRd52OEI<7q=!@hwDV!R0^XZbf4U=00*6HRa30l`
zjk5Eo=3bPygA@}^Q2UP`n3YM{yby*Af;}#Ps~nTz^MR#a`1Q;5*N;$07Gi;k29`Yo
zLVzp#2H*^1VeT`KJW-ioU?7&j(Xa4J3&60TCgHV{fF>(d;5ZHe3p`)1LHFt{Qk9u{
z_q9nqa$=c23)rZAhJ-Erk~dPPu*u`WcVz(i^${N*-?;3P`GF2Y*6oOCslRTw7abao
zpy-mXUB&vMrgdKN<;!G?nvHVWGAw^al)3k!CI^oVFLfBk%cD-5q!O4U<DSI)&8Zr*
zA0ZTq78E%5|GqET??68+g?RMhx)YGbdLqNZQl@Gg^$vKX5KObAlnCLH76{Q}c{ZQq
zynTc^PKclXbAR)Hg!}(D|8*{P3a(1j{x~oy%pz;4L;2|#0m7eU>Shx;o{Ihl2MB#l
zX`3GVN{$n4I6zQBro4{2&HvT`{ts{Hk>L1E1mbg%_4(`}Yj-@htH5Gn-JJ_h50V3y
z2j(*ulc~y?j=pt?4jUM+etMO?ETjo5I9XGw!jd0diy1lMl$Bq(51(9V2(H@q6tWdh
zEAVc@H8Ot?eNU=_?|FI0?E@DX&hYW>62H1c1yu5^@J`X~NlVpMlO9m-|E&4H-TA;q
z|A>#^z=giO$EWV21fQz4;DbxqqdK<_cpv0{H(DMQdpO?Z&m&$%ad<dO#$}7;{wr4E
z;0iprmoEeis6YR=RO`Q4dw&WPQ1JH3jX~gt{p&E{@KjsPKJ{)tMjg#!bp|k|*G;%m
zP(^F_Zs<HutD;-a0$jHXh`M^~UVv;9m_>*mB*UQd4Bsvu1cXdM_y(Y+_v#G~lFJQ7
zV6Y@aJ9NwYdLS9lI4Y?G(QFjN+^`)xj7wHbgid0B%a~S7U444CP_G3ZcXhh1hCi{<
zXj(p=n<Ak3K@+C#hnj2;J;#mlbWjLzzi|IOPk`c?@2?I$dJ)gWg$?;tIs-ufWOCjI
z8>2-cz!(IPl;4)9#lU%Z4k^UixLILjm$5P&7kH*p$1Po+1iAE#jMeY&@%HJpG(0@&
z0iZ|)S(+a!ft?~;51SWkR?0?n!<%L?kAY-(7;OvqTFumxLN2U_FJ!ctqT#0Tt>S^(
zObh1?ylId`(u4GJgC(1)9D&e6UI(zes0RTZysZm2i-~esFcAZ?TvKTuJ`lQ?$}#F}
z)-2dm?F}TIt3dc8cv`tSI<)Yq2nBMX%ns0e(aOpA9Aw}@dI9G#;=I5yfJcSaD9AQI
zdRnWu9^9n{#5b2_oKbNYi#`Yqx%>)!wg*vA4lssTV`Pv+Igt<|zXBi{3CYwx{h9%x
zZ=(BwMjcpg<XCa~wCBZ{gah*f`XrTiPqnqR8ygr20wCkeg~8m^j(6xH45A|=(?Q}z
z2emimK0E?6ZV<#-zyjM!NlGe$g8q+HBo^{yDwfZ&0EL-HMn(Cqe=1sq#Z!h=6+B*k
zcu@`@u33MYBs@p7r<Qr^O8}rX>)o6i%OTl98AND~yf!Kau`D@8Jh$~zXz3XN4TapP
z3wRVHSV~^kgVzb|ss|uYlnY?pWelud&`tnNgW{0GRG#0mXW2xFR+5lj1KgcHbGpd&
z<V1!#{9t1u6G%}s*BNgljeAr8bs{1>JQZAtAKV~41cN{iRe@Lp@&M4NN`Wv8<6Zpt
z$0qRjWq_<C2zgF7m_v{R?mz)VuhH6ZA=2`{gtFcc<Uo{^J2eBDt|W*U0~%Xi47+C;
zVpc#{hoO)T9U2*4;qNn`K{j>##55;PNsSD0ctF2HTY|}$fe#mEz)4U>QUmi>uSBj4
z=f`<oz&cUs@LsZDTHTqFl9hc9bM!t!DWTY)C8x@EDJLyW6kY?$^T6DWK>*jhSj$NK
zU?(m^1U?A7vdg-BaM(WgJVs%vgg>{P;B_G<w`z_AAhVJ)a)-QZZ#B1&o{dd8%np<%
zR^gjraIzwOr{V@MOlfHPneq`*UMC|=0EnUII5=cvh2Ou6)`5<$%Sw9mx4zS?oUT#H
z{^R{M?{-wP@A(<vm0@Utsh)qD$k5P+KxO9PKsIJpkK<Y!O;X<Wbu0JadjYt(xTu6T
zGtor~>e1Ce<+^+@Vowaf&H9B!R8-U|)GwYx5aQjAim3cA=~UZ=HqG{;9(f97V<f_$
zl&_LmEC>P~iIDw6Jm;dpKmdqA0<XPPz+Gl+Yx$I1&e==|ovXp2^*ZPgW)WlRI?Z13
z#001YaE*~l>v@0z{X@!YShCEm@+^=q7e+wEWGF0czBTxy0#vQKggydXW*n|*^;vL+
z5g-&Oq#)tcQt%n)5wRNj!=cLsHBWT+@>O1J0oQ{;(2~qUg-?)P$FdqpH8Ov2Lpj#(
zUS@@Ls$A1fsgPa513(ApD?mbu$6njhukEgv^w0w?!xY`&nQ{!zy`pz<91WCbf+#@_
zRn*1?q|dvYjN3}MV#-0QF)}!Z<~3AK*-W~zDvOBUz^)rA)DwdANVTRRqsKfiXUp-C
zfPfBc<vfRKb=#BralliksIHh7u5m-#3FQ0AFqr|e@sAK7*CW^d+QiFnIyfyRftJbu
zK~r}F)($Gfg9_5dSV^mpeHxQdy0r9LRNEp70~1bw4-My6Oy_Wz4U1+)2Q>k51e!OT
zs9XqALf@Cm0M`Nj=(h)Xu=c^enzr>j%I7DYceJ7VEpvVS^#Wj4nUgw|R!>kU9E81h
zrdr-9h}3HtmuU0|UAXm@xc{lwRhqH>(b+i&a<T;whvY({J`{xt;&2|KvP2Zd@a2nw
z&{3N_C>5~3XhgskaED^sU&Ov0!ctWbWK(P8OoFGX10u?kTZMtoIT+a$oTcPo3HG1|
z9|?^F$dQ7Cb%z>K*L4>2v0^b;^2*mFz)X*z%-5~seSPNdJ07TG4n1Y|ye@lL&;ytc
zZ>Qb3<Q3G&z~uD-_3c~l@v`Z%W3J=YG!pU_iQqAnK}TX#uooE}{kH~So!%C^^%EUI
zE>7snBBS&>KFh_1L5FgFuuXxaK-}8>JbG@c8`D)F6g(tm0{@Z4=-P_a{>5-Fh1NG+
zc}l9<24SX$r!*B8_GeaIvLu8MQM_t?8EuR&wm6->!)IR18QOtG6-P|(JOTklr$hMo
znc8pms+O$xwPmOXsMqSh=KlYG@j&yl8i*R$Ijx1(9;N<yI=aS^>45D@(q|=`*@qk|
z)*4-_n){H1ibXnB2lYss;KtnxmI+beh{>Z8V?DAM<%vq_J<2{GCTl68Fq}6Z^i6v8
zGt&`>LGr;{AFd$S_5Q+?|0AH{|G3#ZF|9W*BXD$KdS$_qSvZx>y>I0RVFn(Gw%Z%x
z?9>^iToCY>vSJw|roHR4xQY;!WDajr!AJeu|HD=HFD|K`zGKxl+b2dXsHB*jXu;DL
GZ~s4tJE%AS

diff --git a/docs/articles/vignette-10-tracts.html b/docs/articles/vignette-10-tracts.html
index bee61b142..e7ddb2293 100644
--- a/docs/articles/vignette-10-tracts.html
+++ b/docs/articles/vignette-10-tracts.html
@@ -33,7 +33,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/articles/vignette-10-tracts_files/figure-html/introgression_model-1.png b/docs/articles/vignette-10-tracts_files/figure-html/introgression_model-1.png
index 98387d6f1c502799527bb46f11ad2b4ea4f5a0c3..350be4d4fb1219d26dc1580069e0f4323d22c39c 100644
GIT binary patch
literal 21844
zcmeFZXH-<%wk}$Lz>?%r2})E1$sm$5N(LooC=ewh6tT!j1QY}$BN-%zLdm&M1SBJ=
zBosx?IcM%D_da)@_Re|toOgb_Uw1WavpH+bFh=jA_pg6npX<H4ssbqyH4y{?Ayrb8
z)r3HB2Otof>AwiTnGbD-CE!5tQbj=)9KnAlG7bLVaNSYSzy$)i<%<1_GiAx(0fF3u
zD9Jw2_DEevdm3vW%wFvrJz7tGr}CD@*5T7j0>cE$yxFE)c$I8#)ob4F!=C-cizx?7
zv|vRBRBMP8SD=@lugSp1#!KZ_rK6m2`4?H-Yu7AKE_fd+2RNHAlWcoQ((k<xUg{Cq
z`t)qd?>N<kQ#F5x4f{?Dv#;RTV<vl+G#)q-sAuAU10E}w2qXZ9+!`EOAVmD&@B)HQ
z4Gxb2a7n@8j&vX*V8Wr{IhRp2zZiDHm$Kx`M8_vf8S%-<wr~<~=KTGC=Qh}DNizTO
z^y6eY2xQtfRW)Do>fGf5Bb8?}B$Ohs60Oh^scj!DWS2MLfP_9z5fl(33UJN_7ln0p
zcK))jnPlvSb&Rk*lD?lPgrXpf>5bHp5Xl|hz^t_UBMWSXGEhkkn~%kA1~_A%6z2JI
zR#r}o_~6-9v!y~I?`C-8Qz3~xcPIr#u4<a-Cpy9w{6~b;km-cP#C$bf-;z$KSPEUP
zd?5pQPX1>`fujhao=DtGTI}>{UUftkdN!$_oNUj{^p9pe@umA+gId1>=?X;N$ALJI
zh05p>OHuc{IQaV0%sq(?Pg-QVh6&91&w8qt3a6@$rKt|_8Sn1CcJmTz9^S&ywg>Aj
zUFEzs7+-WSc%{{hcCd|n8Y&|g`nd=E3jL>55O|~z)|%UOo?hpBW&1W=_wmopY@~}R
zPZf^rM(znK3Wmz0O-1VUTRtEDtn}*nnO6nTIvhPw_xYjw8e?%Aqf!?)>i|Z`=X0(%
z+0PEfh*ihnJY$qj)x`yql1;udzvqQ~sidP~;<e)p1B~3IZZVUM(RkS@s{M*hfKRI~
zSWwPnT>7ehMzjj1t|rus)HaH|U{oqPTJ^6yK!^>C=R}%JHtK9$)heubdamB``*i8&
zt9M{Mt{)Qzt113w=fZ#9RZTrmluK_=^CU&##z=0}8=lh&h<Xz*a4}Y@(4j(yE$-1-
z^cng3R(|T6r@a$9K?Gbz3sVo7A80v=#f~7JNNmtZ>E8ZaHi#6wRAbmsiDm+e38kq~
zgxO<qaT}n)@KBQrL8S{yUsk0qc2@N>&n;VxT}k|7!+59DN|WE`X$~gkPh<AI{hAT}
z?q9h}DRQ<tB8H=mJ`9$LcqZ?+p<;3l9ysV_EcP-fvCz|%RUbTcYxQPW4(`RQ)Jz(W
zkp49s@<6H(HXC|+KA_>g3CvVXmSg){p?D<Q;oW&(+(DQRC8UT)9_*_HXU?-YCYvrd
zD1Yf$75kS7ouzo8cpt{5-|*h2s|7ClGD@1rgqx>lA+Y^97>ldXgzk9yVX@}p$8IXc
zl!lL&DGiv=abKQ+b<WOEOB<{o{@le?xq=Zo7?Am;*SYChgvnkQ_t;O0G`A?O&ajI+
zMME3nW1<TKOO0$ddz;+kF1$A9GVU%XM73qx+aPUGFqUDY4lx(KYj_D|c=e^TF&bl(
z%_4cFJkb&1h~(aLZT9n9HVM9V?7a@`BB6nu#9(C~vlmQS%D46?;FQN8>-0%rt^;a1
zbwLgl$6e3bp4G#tG?*!v*pTn)<6-1)k=~~llzA`L{G3}U`FuhCITWcBW9O<L6S*{B
zWZ;pc$ugPC+;C*$)^%T^$uBnj)cNyijcrzL&{^V7=jj|fe|O`tKHtXW?MF(3oBOfO
zsV-HKs%E6!I_VMEg=NWegh09$iXU{i&Yh{~NC_))gpv%4;ho-{QusVx#iV3MMxUS+
zqcs@7%;DNL=^k%YQEZc@!lbkZ+x)!FLO-w-r8c02lDB)-acGOlOxMk))l3oDQj8fT
zqshVkUKP<=wJKfY*WJtE)JW;JrWlj3KDET%E#@zAU&)qUQ$bR$btDFt`;q;ear!YQ
zL>u}6M3}AxYj`yiTed2l0Vidt<VSWT=c03zaU|IdE$bHn37cbOJU6bGOc}z#D%gG>
zNh7~CUG~l1Xyun1jVp>JW96!ODtrA6u0?7eDk*Y9r@LQGNYIW^syGU}K9=_A`~Jl`
zg-9!Q{zuYTfjBXJkzQ24H&zFiA6F5T7DpZCvaTQ05tTtz&QvWGV@y_jg9<b71B$|L
zg1AyQIfbZ_Hy%@+LzN7Tz>qxQZQ0y2iEK3kdFyZ9IB&FKY;^{q53t@RYG2X0=(u|2
zWn7%-ee{uOgqUb}^Q?V2y;ziW&b9fxRQF;nWNm-Z)FFlm|LlZ)1#z=JSBKalRkCz{
zQk>KdBX87B5?dCvrZE1tKrxtr0Q*|w?vG_sI!H!<%d-9M*JHbjTuYWuv|0EVh`m-z
ziWs7k*tWud0)MWfU*mwTwCJYXxrtlAeYi1^f960}Ab|I(CsNZ)RVjhkVL<|(pD`TU
zpmi{o_uNte#}Vh*3<Od;_FB+hNyHjX5>;UBNp^ljV}Up`O6jgzh6(9y{{<;x2MMi(
z>%jw=d}89>$QcS%wdCV``mz>>#4-k{gutQz5o};bKpTo|T0^>C(r;yKres7FYiU=G
z@j~M8Kpvc4Tw2OMd-!bN2M<ys5bXd@rJn1!z2fC-RsK-?{L7dI7>4A)g^=@|e6rdf
z?u-pnHZGF9`CMnB<{YKaY#nSOJZt||IsSag6kRDNWIATj!1wrIicn>w#Gpc1f*#U8
z%}z_vq9Y8^w(5wa8(Qp5aXZ~#>5}+3#5OT|+$GU2_odSIf!)49yw``T%OfPUfNh-D
z;X0fD)h-(UX21#S0EKiQ0py0GaNO1Y1Iv8YZ(jE9CuCs)AXoGJp#4AP^*;%DT6AD!
zU}vu5;QQXyK#u`TDTW(HZ|Tff2rqqc@V?GKP3<Co6;%d<(?VMCg=xv!+L}IJ^`sSU
zRj+G0mP!lM^8Z-Zf285)`0;d-{4-L`SJ%vJ(Vdu2BeVEZYv_$)6e-kgqKf5UWiO^I
zik#MCa$UcxD}K}=rP!{)r+}2Erby$nyrlvzB<|aatL^;q{nMQRtNMq6Mkq8F{-`Kd
zVNBU|va*B>j*Sz`z<T4o=;o>IQbtC`o=9_QB3FUE0X+AW^k!tCX`DMWMRdTFdZa=9
zepTe!jzT+!Q>Svp0~|=#7q0``jo%HbK^c7sZ7<=}9>hop+bLzro3+clog0&dYI?qA
zaHgS0g!R;HoSefo7{B6XJ#RfvNfZ~9?$r6h%Rv$kV!~;*^VDrUMb1;c5K-G#SZY*L
zl$qJG&YG7ct@mJTY^=y=LQT@-oGN`pfoS)r@Z1Ow55heTV~H$=7mTq@mvT}T>e887
zSXg|bYvU(hUTqKWcDX4J9Q16*U7#;N@9N~A;K=co8L8`ImD4qC)ZU7_lCNmu<l-Wm
zrWbsr%v4N<`l>!IS=Yz&BXT&ml;5Uc_#+1$4&>RatFp6l0=ww5q!i{8{oEEqsw1y_
znPgsVqNZo{SET)xK`Im%pM%4WIhD%Cpi%Bp9KJLubL)csK8i6-aR+0bgH&QUS^a5f
zGDv{bM7F=u5G-*$*ph{}&n(e|j)Pg;Xy-zhj{f~y%}2P<R^hb4hP9lJhMFk&pw)!&
zgTaH<xaNxxy`1qU7F%20i8-QHm%*ue)9n2ns*0geDv60sg;x1Ph{jkc#XuYiZ$3y7
zVdxNtf{QcpV0=p5(E6P`*2;w>Ivk1@JJFYySz~NPLPV~lF7rZc8#%5=y_$)F2;7Sk
zGMk^IxiJ4M4zJ_XJ+1!L2l&Tz>4`c&nL9N+l)V$(++f6+mKDNWwq=C9ks9BlIwS{H
zc^)RwwUSF&DJN!F`8p(29@?+82%@S2C*bj3WQlLqYUL%D#F^w=KPFctjv{%+>@|l?
zb2e)lx}Z<e_uITRcZCqKzZkuhBwjLKNuw^+EELzKq7PxQxNrIpHR(27$Q(tQTe%<R
zVz^A}D}!B!6bsF?R?^r8)x3=-njRU~ZwVAW(dyK_XK&2DVaT&!sS;tNA`zl|Q?FjW
z$h1*r_4m+$$!Je(dlS=0k@rmhs!xWT!Px<^@YNZQGn0X#z{G*4Tb*3J!}eJ0QJ?$d
zV4kD#GM^O2DlDdPhvA|?wz#^s`7k_&(Wo}P<DzDL!&~w`ly7JK)e5Q&b>uZz)!2=v
zSreKgV>38!rfI#=)p-N*C~=6b>EiAF911b3>u5M$t!sFGK4!0a$+&Smsnh(hnc0o*
zp1nTP#@H;o`(DP!YDDzwV1#qi$%B<9E90`>{&-8ZkABC7y9cqms&!#oXID#x{-O$A
zXG(f?O;@Bg-%~Ul^7hA`qE5f|t8kS$OZc;Y(@9TLwfUxnxR^JuU!-I>{pQvaglVXD
z`<?5HQDQhZDx!f`m+0+R9MAdr`t)60hM&WSt9ndW(@vqMyS_G<n)XN|{@}T%87IM#
zT5&)AawVO0RD5p=P^%Ibx61`_rBzO)S-$JOP@zrhpy}Lk>(LT)YyR3(NI_ruAk@^G
z#3{mK+<x$b{tD_rTTStVdVNXmnrCq_^m7K1KH(|18^!u4H0O6dlR-{xAwpwrJ$CRL
zou*my@R90DG7e(*Xo??Y{dy5*rSJ@S^|VXTC8CkcqeUk!ocv<9(z^065fAaIW`&W|
zvrB;Jau3Jm8`BDsILF%rZto6OW){kJ?{`H;Tiw~kNy`bwzm!i(hFDD{s%ef|W2)+j
z;ZD8lEKA{rz=D)Rv?*EoRZ_Db)#Wn{TKswfCCS<3d`=~l<ov0jYdNOqO6ll`=OlTz
zFAy%8n^rgo<x%PXp=63QJhbC%eqsx6wId?gN7f4AW_mdZ=ho7Jl#nGCL?f`gs$tpj
zsBNjrN^#tDwf-Z)U}i@SIt9v7j}O&6m3-6psY}mQfFwT$6V;Gz>a;PoAd_FKZ0Sg+
z_bb*|OIXq$buj6)Xr<G)rKxDQGIOH2D{ph9YXeQ)9}`13)1JAm7!3NueK$Pj#$E~9
zE}VMJE`KV9E7q6Dope`dU-r*VUYIOjP+hlkIy_T(U~C#L8M;ef^D{Wo=0M=`)2J)?
zGPU`No>j)x4*OkT*CVv1bDxXnlC%-0W(S|vQ)jgX#Go11t2ZB`qL8aqbfuY@d;BgC
zfzZ}0ABmUnuv^r|gQ|&gKa#;_ufn4&D8cnTCPck+EW0_lZDstqNy-npnwMLjlfJeV
z$i%iQkJ@|dU%xmld6y?R^~8R;CDhm@G>y#0H@jF|==CbQ2SX$40X=hs&uEq8<`%-{
zmdGb71;>60>c!L3Ih1CD9aO3!lEIm)m<A5d8Anami7Z@j_Mx|er;wOWdu#n^Pj@>~
z67F2U9J5~0Pnw<cm9DT32*^Ik+V8wF4t5`U)_IkBwd*d<vvHv`?jmc;;Iz{rT=Sah
z!j2x)Mds;_zNB@-I+KSmiHhu0|1BX~kPH)Igp@+zb$;AnPpmpu*u@=&IUr&;bQ`%#
z)jsa-<qYNL6nm-o1#0d^BAidg#0IY<;S&AAo8wjkpY~v3@&+4%ovSPY4i83%-Fv_J
z+2j<f=P|h04{af&jw)&g&fI@@`)#z*A3g2h?6(S=6@LFE<k+EXttJN#!DW)p%vPPh
zwiZgm;g`D2chx0vYSeaw580N~GufhUGd{S=5s6YPuO+fIYhF{nLG|(=elvv)>XfFf
zKP{TL*e;fn<b*hQlBjs&QXy=0Ni=W(g>n;kXSC^FdPMBl)a*uuI-Pu7D05?FGZ8Ap
z5N)hi@>j_QdpApDJ<$Q0r(se1jIcKEJ-^Kwb(1}Mq5p!Dd;gMSu9iyF(o<54+{#Jx
zE=Qw2y54Olfy-!~i(#fSmc`9CQP=D0JxRsqRMbX)=yHcI@=-8eKsZ2F`3xy(+1E1O
zE$-Gm9t6qfU0q#hGP+V@W~m_osxv)2#z!=xu@ftEfX?%^9motE(|GocJkpklg+;-e
z%*)*~A!!O{`UWTE$&~oOI=OI%RffZGslE0Au`7P}o=Z@-32M`D<xq)Y!e2*s$Ui|V
zvpD+nU@e7i=1r<98xMxt-yhL}1DPr^kbQ+uz`!RKvQnF3I=1V)UZhi;Fvk0=m3G;d
zNmy7PC2v}|bt5?~=QU;{cHH!rsxC)t5v0ZF;?=m6+oSgt9qRQNhsw{Xg$jRZ@v;q!
z^z>XERi{5E`E`glb+AT}Kjg+Z_u<Cs>WIN!Pb3~hI4b&qLZ{#3rzwf}KXt7E5+coa
zb*-s9?ESmxq#d_nC@C!VFOHT8RYuO;jT+2q;8zK^{lp<6l1E|@uE#z2lVuy~Dp#Jc
z&xra}b{9Mx2sf*kn338uFGU9KEA`93gj*;^1QF4OXSo<rPo&1vOw*BuFS}o5KE>qN
zA;O1CP>Y3jD8xgS$5i<BxQ$iPDubKM(Os8riA1k0S)9<?%)-(|xBVi<$);WcycvL_
zny)1yB4XK5%Uay*uRam=;}hNIw^jt^DRkc+HJ}lhnfLn4M1x}gD&{pUUc-U3w3X{~
z7C{?q`$EY`4rBc@$JmsMiZWRGUXuN%#MZxOr2gA{*t8uSl0Q_)gFM#^r{d9t!>b+Y
zS0180m(t1&OX<ZzdM&kld>Tg)(BgWxHQOKWZmv!>HF#fykkatGEcU!CyDRBa`|{N*
zi3BYFHWIXg?*+W}biJZ~f84ID6NS<%YupV>;D#1u$%J~X->I<d)dT}57#PaV&wZYM
zdMp<h7#J2AnZT5rmuCw~1zSvWQ%B@=z~ZfT4ztn18=$bJg9Li-bzXcz!fp)4^s_?L
zr~-rU{MT3J78W)arw7y`&N(Hz#czUw1mNrX%uGzRHA+fK8Xg|C|GZE4@Xck6-`R6O
zX_=XsZCXUo2<g6jnSFNl<_&IJN5@(?PqVtTbN~U#oy<Ty!mf^dk0VTjdzo&DewI?~
zLkc6m<88Cx&`|JjOj8LsBkDLEununk&&1YtZ{J?CpK3G!)1ZEA{_f?=m*!vf*;mp4
zE16mJxJkOkc0`#*$Uz${VtGZyNWK~a+Npz%h=e3=Xk~c5Gj^m{kEgAzt<#~#ZmQ8U
zS<D^L9>rk#+7$Q-yrJm_ZzO{h!m>A+!brDRmrF_0&#&3X*Eb&Udz<J6yYFwWcXf3Q
z?yn5RCnV(YzG^4;MuU>;0I-=2NFMfSV3GonXKCM~Gcw8~ybqUArv4-AV-*xsR4fb(
z4A0ZW4ZveU$>=n_yy{Pnws?w1*!}(eBd7(Gi}lJ9T4z{~oP2!4uf#wtQ=f5mc4oKG
z6~8)M7|Tf+{`Bcney3SkK%<t(-p3&Qk&>s&4~|-A9RGNp?Hw(J%%C@K=I7=<e+wdP
z$X6ku6V+z$JI(?;Wo~69zlFKE_4f3)O%!=N1~|gS{*c-x8x61!c%dKy(io<L<zF&U
z5{F6Dg0BnaI-;n=92^`F;JI(!yixAeLHhX~@dAsZy9t5S)({PS7Z*RHAL!`o>;H<v
zdZNy|Li~DAip3b`9!Kp(T%4biQ&ShoM^M}JrHX7WC`8eZ#Ib3SlauF+mKxce>@LLQ
z%+H&b>VJKzCUxQZ^+HxwmX(<~_qS{K1BFP|$B#8B!{v->97f75y7^zVlN{L%4i1vv
zx}}-@A~Ym~ou7aBqN(jAn_x2Fm;+Y(f#YUnWdXMy-CgLmsF?tNs?0lY9c-eFfS=f+
z>#KJ6_Nv_0bVbrs7Zw*+dxbE=O-EBkFHQZ$3%@9W4Kh;aV#Q@pE(@09)2B~Zw~&vd
zWm8%m$O<~#!j#d8!GEr}0#j%u^I1iLpE{M^f?2bE+ZVg33e4*1>51JI?H}$wC>|*y
zWbjdsF|RTdup8y{Jwm&OMMR9?-4M@AO1cA<xpjay@7bs9o}Oo<z(%nwYNKd(Olh_K
zgmMfn9-e^r;p<XpgKiBPo<i?`?$q|<2kXO!?Q>fc*j3JG>!f`f4wg{JeyqHWBj0J=
z<W^zhKGd><HBa}qfUTHEIG`RN0zVWcwtiL1>|x6oB!Q<XEPyvJ3h+(f7T5nD^!)!Q
zieWi}f6?*(?cn!+FZ(B-|L^sY|8gZ;@x3n3Pw2<60}H~sbTT*K4QIT5=N@CPt6z5a
z`I?)XFQ<EN7Oqc%0Re12-4z<@>K}=SiP4y|lan;}v!tY?uv}d5W21H*(f;>f;sr-x
zCM9fO7u#xlz>qOWSl0Ue`*4MoT0iqO>@Y<z?A^@=ll@dYCM6)uX&c@Hr(06)O1^g-
zDbk6L#(Gl0tu6uF3k6f(gLnX%^6&`Bemc}7*7qh0*s+sib)9Yb5aRu8xp01;W5=r3
zW11V|Rk28<zs3BYSBaUyz3BJ9)B+iOL74`7zi5*FKuPEMcOWZi-8BGbAZ1py(h3R+
z9AaV(kIS(qO!^%I(?#8iHXHX;DagsS<>|r6mg2kVcuk&rmZipZnjj9W4&_I!f;hkW
zTxZJf%s%ScpE<n;&%<0@>S<_b#N>d8n46pHDNYV9pDrb#S_WE?jyxIoSElkKaUkh{
z*Xd-}&fMGGb)EfjUCja;;~J;rD3(Fgk^^r9z!fQtpeqjyg-DBui4mv7Ub`&_Cj9kP
z2l?H*h5rn5ZHN|L>w$Z6a&jxv--75#L866t5Vx}2lsuYrG5qo(P(F&DO9`a@O*cNv
z9(BOs3An9lGe3Oj#mx+cPBW84N%KJKwE-8#<Kp}Jb%}172{cO`{1j00*f`VI*LT5H
zw0kNmD(bbMQSejjYfg1J2GR?@P6CNZZ}IfU57~n)4Dx9f_yq#9@F3f-v>775c{2-S
zW!uSGr}BvnWo6~TbP4acj~_GDBu``u3k&0;vC-yH+6*ooh~gJ#M|M3)ysHy6WlVLj
z4RnzX$HruxR&R>Xk+mWi*J5;sdmx6Kk}?kjJXStFEfDQ>o;(Ts-r7n*Prte11I7gC
z)fM1EnV6Vl@87@gESe_jy8JPcPAsoIg2tO2YyD|A?p4*ddrST9QG#M=svztYSac`Y
z4(7ZE$TuIPtf-JbW()>q{H4HPvpe6J3><;U=In68u*M;6u`jI{xL^UucU}wedgAn0
z({2-l<U2NQe<vQVbpkBE9gEWGbkL3A5(7T8OTR=+4xs9c(!|_HL243D`Evnoh(f0O
z#(1uv{dg_Xw9(@$NMQ??&ET%0KF43I2C)ll((V}OFx6P+?Bc?#WMgBa{QUWd>E$lf
z(dqj7YeYL35@6d6Y#`;=7(zx@M8Tn3V2bqHJvu@Opo||Vb^ZJ~@E}S&djf_;&ppB3
z5cUF?<?Z#`adB}u$cqiw?#|Ay4hDZtDJdx>#F~td&7kZ=wf&w4*6b}bcjyDF4NDDc
z9kX_KcjEvuNr5$wH94$}mSTa@QyCcv>RZ^kQ&PFnSu0_Y5846s{x-{GH{*i`3v({N
z8vZQglw1}1zi{RyX$CR)v*_q(EYi9J?D1Olgo8^N4Zror?Ck8gn3!y^@xAz2LBdIE
zof!Za)fMA=WHa3oX!|QWcq|%=MgU|h2L5nPSx+PeY(6_XOHM_V4`w&HE*z5sY+DS&
zG#8m7o5wVi9~oz6cyEeu>3@+1!6{)r`8L!Xq}0{lUq9!mrWFBK?$5mj$qW$#rZ^>O
z)LF^O>iUf|E!#NNNGPGLc-7fqrney%KJth|Wla~MK9u*oH}V=DEm<hN=j!#Qvl+$(
z0YvVbVQq<X-^Q`8=1$6icZ16%PrjI>TxM_5N{p^6wgFF8N#w?|>s`hSG&IF$TgX3q
zr^P@p<+;!SL+0>|`%h2k4W*dY`Db$LF^`!pE)TrW!~Ow>O!gdtwFFy-t0ky|KKBfE
zZ&m+Jl!7K~te4EKolS}Q*ORDc$wh*RHKPY8Qy82{s1<3qb8rwpw8H5fDk)h$_L_nS
z9|Gyadb6qt{8NLt_XDda8n<i{3Ez{2L`#K0oU8Pc@^Xge-!+{O$Uk}kHofx1y|d~A
zyXu$NpL^TXk12)mp4))X^lWdjm&3&u;BEjkO|cPUx&+`sjZdW$$#ipX%Ev3UoSn<Z
zc-2)wiNyX<rijKI9=B)5kcaMU=vCGDFvCU*)=x`T_N}19WP#IMJ3!1FY-~#2hwB<o
zpT3binZFGx2wizy2*j2t1^qC;>_{!#@!0T$%a!6KKwyQq{)qXt2cc-AhY7V{RY6=H
z2dEX8^K-NBnQ{zpASR_!!>Yl)G_jFJPnUmfZ>~Cru_nu2Y70xt)#YEG*;!fDL2Uto
z2?zNi*fS#_qwKF1m+4wd`I}dak%P*n$_`cJfUc+JefhTDeQVPMY1T^6n<`TA+b-k=
z1LF5@{isC{*foEUB|KJ5znEI(kdswx-Tdg#s8p_R+}(lU%B5xXWzb5G_Bwz|Fl<w;
z?61srGp%6wa{5P=RV3q{uIco72Oiz2RTs67O>b6ObvqebxH91>o#5f3XA4gMbm?>z
zs;TU?a&|Je#<!c3Wrf)1t*RNWScRVc27v%*F0TbsZr#33aqCw0$B%zUUF+=a9Z}(F
z%E5*|bN?lfG80xuON)<>Jr<Uh?7HIEz_!(Pb**G#6A;iD%2(3^QBK0|H3Y(*hJKhP
z=Jwq(_HZLU{UX$LJClDioo`|+9Olj)T5-H)FSxQmmAh82f`gOGKsb`pJLA<@{-mD!
zDUq-i&4wy!*JX^3Ig#RIu8l=7Clw9fITH9Ry)-{>4KO(t_Hx(F--^QboQn_wU^o9|
z4Z+ddk`m5RQ=|#l7x5#uI*yKC+9T<T-7&r>Xr`8?rt4a1jn{rHXyaHJ_1R=^APcP&
zGw^b3Bxhi0Ba7!Qbb&WSNk(_^K@w6pny6UNzOV(0kfUy{xY^hoCn!(j#xAd>;)}7h
zv=~ZXo{{Tu2nWtmHKz#0P_L8sAN*g`O7j@ksuPmY4uJO&a9Mm7HR-ktc4cDQOAU|?
zr+oH$U#<}m5n=s+r}_9P4h|0WqZa`LWVHITwcbaYBViOA{5FF*FS~CrNS17GJD!}K
zH6#jvkP7bk5$%vTVmksH5}W)$LgRS!NL?m`MB~}B51ewyU0l%c!g@NkJof-P_GERU
zUtodpL}4Y_Vd=GAUk2<!M=jAYhCqpwI$@0&Mz1o|A?Ek!PNQ?Ba{qvZWhkp$*+zO=
zk=NJPBKdAzey87QA{)nbGGR@cw(4u=6DG_3qhFt`ptJ@HigWv{i1zgep=XGv0E?^)
z<rljw^?3msU}L4E?<qM)zFe0#%Xi=jYan=_Kuvitq$WjSwB#x(;JIslusWjR<b=)T
z$wIJnY$^njT;lIhldikT9Z?K=fCUKSD%R;fU;rt_w7I#ttGiq4XcH}9_Wc@AlxzUv
zQdUtJO5%ML*5m*jAuK%H#UEBNHovr_hd8<pY022^B17P&CFuH`U`{i*CgetaxG#1H
zSU<NbGWgwVI)|+UCw960j2rS)K9Q)v&63F1eG;fgxqN`!;uCwL;t)8`*}>&0Z^GqU
ztKzl5VbgBfkKD+tjbT0=a*8?V$*8IFhMK}`nS(P<N`Gc=6&nW%iu<Idrj~$Gv=}5$
zusKqkot@*NqcvtVOt1_EsM^aiGh<|q{@=@KjZ5!GC#RHG?%(T#-t2(%+ZE3-QmDlO
zr1s#qZ>?v4yf>_VZSel7iN1b;R=#Q-!%{;Qz^!>4td71HbJB^ePVn@K4e`C?i;G;A
zjD*XYF@dm%<W3B`gy-jd<O9v!ojur^V|tpH{aI_Hb*PqMwMx+jj{vdT#!>y^;LdB*
zPY?<{u6Y5(OfPnfMOdFc6(XQ28CtP`J^1i}gp^+VcT-70B@l_%@-zIL)y5o!X#qH*
z1__p_5M)iS^{*YnU(96aq)g7G;C~8omBL(T3{X_JB;~<Q^!9xH`t?^17uU7X=fT0j
z@)$t~#6Pk(#n{xSio3b7(Iu0)a4ABwyE}L3l1(7Ic6u@5;$q~>A--Sl6-%M%GRFcS
zGINS;g#^UjFm*a~>Xo<U`_ybUXXOr;r}JMB7aE|;Ys4_L%T35S#Oi&ei@wRYvVs>^
z?isFDiM#SbXgO#*LJ_bTgF{0ga^-`VHY_2K-3VX*on=2!U8k~D$Luj$Vn9Sns`FVs
zg5Ro-_Q)=hUZRLiE02|xRS~G5*pi*OPQG0a&_7@i>VNYBc@|)c+4L<=C51>@4RiAx
zz(Z6!po;(?v;hfZ6P1&j%f`f%)ANQD081Cq9JdUmiC896sY1E%*|lylLPKXLE^L+G
zX7bKTVg^s`$*boyYhRa(NeB?M4EAMZ_W3=k8iR-2TD_6?I^D8@|60SnFiDT#ipiT~
zLN{rjG8eCJ^c`8P&ADu9o^)`)InSO@)A%07gqv(;6YT?JF90eUI$`y5!x+HuG*NQu
zX9Bz!XD8JfLOS%ttaTE@Anu{gXVnLgW(8B-=EqpT2Z5=wv$HD$rod~jSIBLviMgME
zd(FHZ(0oN;w;Pt5zXt#+4+PuZG%>v>$<qR+LYYu9ubE(~%2yp^ZmYveAXEQ4`XQib
zfu{-TlG2|R!|NGVESphWDKgOH)qprNPEJTvX*yHPDBF8&qRqv%*09L!uhCxhBX@8~
z`l#O6t1-V4o?(a1pJPJ0EKXn=v_(I1Z;0Km;=*(mS@MY!dZ7@|)xq5Gde@a~P)LCi
z4SITcwpHbG;yhAjY6>h1VkVW4$W%v1zWz;sMzIO3`D|0Ht-XCUkdR&-#80pNWixA}
zR^BrWOUpbg4hEF^+GfL+E`YMkN`RJ0mhi5|@|@W2z$mH9T0rHI-@fglu>~OU5P7eC
z^|7KJJkG=ux{3z6WS_6`vTo-Msl{;~+v?DU4F9-VFkN_l<weq(*TBU%B~O2hs~sT<
zrPIK{Qxj^CO3?py4OA@HTy^t-d|bUBp!=}Uijv1<Jvv~j1h7lv<rd+3>CbAtm2!3m
z-2pkAWN3ye*qG+B?0^KYF9lPX4i9zsMHAfM73(_j{i0ZAJoOaIk&P)aDKTs%W#+UJ
z1DZOuE!&@@hfgdxhfhlw;ZTNIIuHWm7beILTm0h_oPVq{Ie;<*Br2g(K)ZDtOF2vm
zOB_!J{R8ePOL=N`bqR{e$Ma1W2?>C|JdI|g32wq0-)XY0wIajoX2UqN)m;rzHGglw
z0WYe@3Mv)hwTq*v*wGR}v3C5E-*lU|4`(hw7M%YQzeY{FSS~g{XW@?kxdbq`+EL=W
z?_TE`^9=$9%k=6r5g0)->MbbY5#A5q8jp;v-=WfGS&**dVS?;o&3lk+{LsRCrs}e<
zYU{0Gjg9Zj#-%KnL8`c?CNU-Fz-(K%fa|h4)N>)8h>UC$5Y3g=zZmI&%Y*VU$Hebs
z;bock?0Xh*N@_?zat;XJMSJs177y)X5s{C@R9C9~ibXpji(nW{Rl~P>fOGfq_#Sk6
zloB4dH}BTY4Li2#WK&P`oDkTEaPLvo{4f*ym)%H_n}*O05lm%N5zrVSW|*ErYCS-D
z*V)3siyz%}9-F9)E}^K853(XBS4?S65a4_kLB=q;#89TaHaaDs`EIXbVuRZIO2Y8)
zz;MA0#yCRuNGzs1Pe!w%zUTt&ByskGTtalmtl~DstK+L?kb1g)0zd5g418VmxRtvq
zMZl==>`+OSr#7Z_#b@ORGxBm@|KfQ1V)tr}t99l8u+vj14boiA?}m_)scRBLB7S1b
zZVoxjc6zI28wg&p=hscP`}H6Yb^6Q8OQO23iR9>!bylm01_d{s!snidS?U@)HuUJP
z^*dG(j=d*0?AFt$@}|R<K`HltY_L_&!c7B3;B6LF4|ZZOc=~|5NScL(g+o--@Y}a<
zw!gn*d3!d%H$ZVY|J4zDw;EpK%qUQM|NN;~<$yNqPUK09>K}uWQ1dGSazxZ)J5xah
zOVLby-;U4*#L~YCvihLs(RcYQ3qUU`w)h5!RwD6Ft$7QS&51yCm}w17tdkTJ)CGJt
zLLC%_fE>8XJpd2~NNW~b104#2DW(6$Mbt{)_|NJTEVBVZ(0HXy*t>TF03i#2{W7)z
zXJ%%$=}(tXR#Y5d$?(Y@9v<%ciS4OQ>T?VvBPAWVyf_oE?74HWF`*YjPC>z}WIIw6
z58s?D1Vo58KNA=^%|${5Dq>>dArP;y$rx0bpb1BMW@d)y#*HIaN$mR;gJK1wsu&=|
zfZ_?z?pWISZXNguQf7Y4^gT6|mdmhG5p?XtM-zgR+qQu80h}6INX1E*Ha;55@N64`
z?eZ)`5{e}kXkd+5wY9<rD?<vv^32t9s&B!Nf=IACLH!MS&<@s6YB6@o3aYA+9YI(=
zVCtu1ph=ajj>SCq2c+n2c>M=xG~gD}zy|d5^TV3-{QO)z9&7Gx<q$Q-xEwH()zh^y
zy);p`VNgwhCDR8Fqvo^xpqeT?v+@Cq3_v{+$4yI5M}S-fC}sI@N-j+R4C6p<+CpBL
zfSs)lYDPC8zJMTrd<?)Zz(Rmy%J4ljXxblC@!I}I7|w;<X{R~ZZoRQqg)$kc=$9-8
z84&#nbT9)N!wi6Rz@Y;~#_n?Q<Na+=M)UpY2`^%9b;;b@+gn&Of^UIzp{%A>_Al?i
zBOn<72I(@sLS9{{2~9f6{FBI9&0dtQ0mLCTtN_i+4OGG1wpFW38UCXHDHH)O(4+x!
z#fK8?>{h^fzI*qVJ*F87c%We*GU>mD0*G|$)~#W{mq+3MVLdwK(2C4JUH~K)wj9;e
z)Wps|?|DiQ@M%EN9J$Jao3+@**^%w%D8(v^*XVQF4&_B$o}!SeF#od@P`9%7gCPZn
zfQY36nIEh3fYQY4U_=iq69L2QVf`pzwE(m|VCzo~MD6zrUq{DneA78sQ65PwwKkm#
zt~T}F5Wq?R`KoDyTs4!~@f><xwpF@tfZ-r5YjMTWXQ!t{AjZYyh;7!&)J!_7<c9pA
zPUB~23LB56yf>Esnh<rGeX9VpRNyu}j)4hSS>^u=LAkpl(h&Y6h=2`X6T%zx`~Sx1
zHAi;=iYy<Wn9_QG`B#kgy7~A2;m%)bYOjd;|BIi*P9tXcf6h?;u&n3=q9<S!bgium
z@*iW@)nqMuQ?}M=PF_w{wg4=Z?L>8{VWo8z(2vnxYo(_v`5EGAH^IPdY7hzl;WhGE
z;fCABxDJEwel#4Ai9kKF{*^5klLP3oRUn(u-@wK#54=__;dH-K8z7;^)qYY2Zom)H
zD&T%s0l|q?zQig2LF9qpio86v1K=v5&Tn}jYx4cy;QEn-8C-Owjn=~ad|cg>LrpPw
zwLg#o;=$31kHu>!Ej0<9)s4n9)qKmIq%2SqdNa|2E2r(d1PHO3g0nDfOb!sDdOz_g
znEIXMVrlg%AfN(Wkd2KkEGh~GfSJdXk?$Yn$N;DUtv94fzJGsubOhoMHdCZ~E`GW-
zae8uM1GLflv(1KoO9%_>14}7;pqzpXGf}xcz&d?q9`;L&|MSNq(!oToCMN0FpjF5T
z#s;7o3UJdvA-KG34Fu_8v(}K!1t25<;vuHv`ya0dY^)@FJqeJTBlT`}T&B&&l9E>L
z|B-bMcSgV8Htv@=O2q>9U!N6JVnEXNMt``=H~jUL!t*pyo%iqGr>p#-=&!dv_(KG4
z&$M>`Ap!vm{sX-J$OYD>rlulkedX?e8A$2X{TCk}3J|>K-eP>6BrmT9;LZPcluO5x
zupj3K4rD#o9ziYSkO~BCpc_<L_0t3MFUP7$YS`M^T9^X@5b_MD9zf-^y}G=>2Jq9v
z4GyKSu&@{=tZBsDH?={S{x^;{gt(Fk6n-{9b?NHq83wb`QCH6dGzS2R-l7cPZqqsd
z4^UmdeqBRR@m>BuRFo}H##m+e=H;p+V*Sy=B>EM&0R<6OVe)v;syCA)<=+hOc%}V>
zASmmKGcz-XvAIs2?k;_x-{r{?*dN?(z!M+ug0m2l-tE6!K@m7!12g@Xa0B=SBk<Kt
zd^C1W+g?-x!5~S|V_6MX5Gy(E-OC)_0Hq!~?f;feKvP6Ez%kax+cQgOL`gCD{0Rt_
zpJs3+YzCRIGCnp1xpgup{(SjWTb>*I*unwOh1B_W5Fjb_7l*N59A%2~`P?#F>Pw5y
z$oO|}qn0B09Nb$QbR<~JWBsNv8Bm-6rs#rB0RUu1fck}Hg(M}Jy^jISxDg!y1_cNK
ztW;w^kq6vQNmEys67?3Wk#xk%T4~sLW#MCIFb1OYW(L~`4v~JYi@3SmuvxDHxZ<tm
zGsM|Ra3k0f6lB=7^GI=~SZ?w@s#n>o>jy6okep|Xe>&;e--FdsrJq{@4dDD_Pof+<
z33ZD2V(Ezu_|dEL;~l^Zw<V;V{RFy!#L=YISRMBDwc_lE6x7w#zlz-eyYIge`LtHi
zgv_8h`5DMo`^Hbf^)2Bb!9?BR{eI!IGusCI>{%}UpXBlhz-_%!L!G-KU{nBmpx909
z>HqYDuK;Ig@aM+>!C+k&V1ZcBVi6%URRFvY+r06);~j_~Kz7loaezkL;eCp3aCa}m
zd@&2M>`OIR7XX@;9-#0#9)tU5dOd2A?oSob0(I^$fTn<ge*o~zACv*2*uvLWJ;?$p
zZfm1Dz+6Dg7qI9e$0s0Q2Ou{lE;+df%bEF}9ny<E1UD;U(aLuO+|^ofS@R&EkQdh0
ziUR^55p-g@DT4NcV624mhkvq~pxq$uU#uo*E;$kep$`iOf9`-=sW*KR{R5h&r(b~Q
z)hFU#yQUZe_z0##033IAc1~x*dC33%I}-<wz#0f2%$sj(-M37zjSaC6pR-q=g8%|<
z#v{lK(8Pb$$ufC2Y<_+oizeC<E}XEcm^Lhrav4ZaHXwbpCHR7!7xw<W256oEpT0d*
zOD{2L@?zuWRs~%<ASMh=*0~7yogcp}a}iAgTp)9#)A_Lz(Bw*fmm5~uDmh0407Svg
z!=nZafx_xd@$tD}F*vxmN`KH%B?1xGpu&>E2+&dPzqYWkLnjJl9Q+wf6p)Lj3yG%u
zfG<=U1O;ClfRO-i12oo!Z4r`>7zBbVP?fuyV5XSaD^*9)w9O>|Kn?4hb4A@&nZUOr
zfcJi$Dy-3+z{R2QGero70K8@`gQTQn2zC=?uJT<lJykkgOo8$HU;0Sw56IN778Or+
z4-F0i)-Ma&)RIW?-|by7`FFi5TL0=@fz{3spdDLDH@6zx)*gbu9bE%rPXK%SK5kl4
zQUL(#wukFu%tZDe^@9&3aztDfYZK`tytJ__29VSNeNej4ov0EM6B}FgvW-Xrw9&L3
z-UMQ5Tpan4kCl=rP_IEAviPcB37Snnug)_d>N+AHa&bKevZn>*YcSJc<2=<FVb9$h
zU4(HV2*TCF<>Mf4VR?Tb;ej%#3upsC(kt}dYO48B{XcF{x&8(`jP)Dn#RAsmU}4E0
zrbz($E0%S`Hidz0Wa6_UH-dl~fv~I_MK3`@X9a`@tR{_RYk)HMAK02H&m}sEv-JvW
z0TM!TH+!nd+k0LDgojM8ufqKjpUFb7Jfu<e>u4drw+)H1>^#1SW4vL76!3ra<dHxM
z21Q}$EV_JWz?(eVXRbPcYO_?l%EX;IJ+X_xoHdW}D*vXGkkd0AT>nmU+1a?rj>Rj|
zI{&wE<UcfyU>gY`i-0g5|N1J_#A5~@EBXOcW~p$hb$onW0BwNRpYL|}M&d#WvRMbD
zNh@8JR7q(Cl?`wX;=aCWlDv>BA1#EzOQb^454*FN+}{0#+C9Q(fNroJ)oYM~m)L9e
zKqg$ty9|o+b59*cG%&5ib$+^ZRn4#VkyP*{AkiJnJJ{VZ6QqG;o*x(D3_q&4+)F6;
zn%v;-7l>(NG|D)0`)!~AYP3gUq>#)o@a?cb4+`_J@QW``#Ve2C9>Jzcd!ra^$C3^S
zK6QZ^V6d2x5bzUzzG>zz5oiL%`ST*c2Lf<XW&dh1{?w6smQtTC6?YHn?Y(nwJ6RdY
zz;jUryWoTlN_&;Zy1sVbaSN=<;ef?#ddU#k&!0>}0;sVcv@MRG2b%M^N=W(@&U-a%
zte;9rh6yW=JzR2MnYfY5@*o^W43WkMpOm(UtL7K;n`)Y_4j7F${t#fKyQS+q5sXy%
zc7F+jnOM=>@Iow|h#+=-&NKJ-ObTrX`<DnI!;0kA=Qf(EO8x$Ja3b2DBK^t@&%M@(
zR=P|EDTEZm^OUP`3%4!?rwe1Ebl88!Ca5Jgd#QCv-mj{cufJf>JiEe7rG@J0U_5F4
zGa8tf?%z@TZLW5|yb_08<!@v4mh5-<6kNCz6k(#(nF(?X^5&KWpPxoYF8?UR<84ze
zrX&|4!X3fkQfJ2@dH>psqV-n!(v-_$R7(8oBTF#@$+9Jpeu=d$pWED%H;`vyOOwm~
zG;*r>Ek_;Gh~VannzCRXOdxMmLRs(j9C5{<y_oT|!$L|K>&NhS?}k6(R`Mrp?cJMs
zht~P4ZsfI88WDyr<X6QsGP32A>P4^5^7xIADg6E)-@cal&hU%hI`mykNk1Wfsi23E
zpOj!;8m}Ke@<RI#@Qw>1FN4GI_Ji!ZLzY)Kd^Qpf)pKABL&OOyDgMh8)<cp3rXSzE
zQyk$ZFqG}G!{B*ANkedgI!;IFCWtYc`@u{1=kDLXZ6aZ^TlIZ)?6OMZ$Nu+F9vS^>
zt}9E#^D^zNW1pKLej-Cg@DCL$r9KSZ`yi`v-y^>nEt*enZS#nmu#&jEfvD3lAiqBC
z`)B3F@DMqD%2d(i6r`VR=P85nnzuwVlBrgAL_S(Z%y%wsFW`)LO9krs6;pjJsj~L<
z^S}>*)triBWW=GZ978X3<VQN*zk7F08pn{jN5KC6MLt~bKC1NXhJZw^by6D@?AyTm
z=KHr4u2zgtS$Z>ngSBn#{(WrNdA{J@dAw~=Yo#%Ep9b7hkYxGxQc8c_JWb-pu#`I^
zqo&aQdq&2~H3@Rjqa8H8gV$S_$B#~POf{j&Au{ca@$ckm;>PQfeXAhOlTTv*zW@dn
z86SjUk7&p6LfvKdKPICLA37SvZ$4yXyu}(3NO|*0By~Fl>Uc)`7ujBfPUKwzV#zml
zo7>2=oZjXcdZEk4^H*&!-=3pW)bTiKqB(S{Dw}r%AasEodh45PGp^qz5)Ve<bn0-Z
zqVO}L#T47|_UVbPEv@;*P+zU+D@h$+w91iA<0Udu+rI=DF<=;vYw)|dr7^2FZ^k>Y
zNU}JFIl*IR3royYmsPHLREaWIBw_CqU+j0@AN>Bzf8GhBb^3OjkMP~*$k66vS=pTT
z8Ud6xX7m}nX@U%rHw>uOx!X)^l&-P(c7gVC2X|&h#=##2G|D-_cf~y(sVrWGSy7TM
z4E9F7gugr7qVMhvTnUcbxYmGvYp-UGo_JsBfF9p0`nUTA(x(W-{u09cNP#syU;Rqz
zhhY{GxD8J$?(ec)998{$rbCO{rMz7KDpA(2;@lz+aUNDsP_TOA*3FxBA8;!bi79U1
zZbeEoJj#5foA(`oKG0U@5|<(%kUOs1W5AT448em=?CxVG0F4PY5h=bkKngdbA*|A1
zPyNGXK-a4*6NG30EHWb_j(PwkcEt4n+)C^KMF@E&C>UjBaWe@E6~o~xgUq^R%F{ch
zm&B0RhtI0(7$1=YrT;>ooCImDpvp4P3yP!$32i_`(!ZR2ENFlBq0?Qpyej?UfI5P~
z<#y~=#)pZG>MF{c(Rr#khK40TTyk3Ds=0~UAoCOPVtdl3D_cYV!rdS#L1{cN=mU*F
zAb$U4a;TN0xx3P>dr=*&{STipiixDHq$%(@w#0wvNB$a3=C-idNwiq=v{)gMoT|sN
z9b<-3?-aH2AY5EA76#9nt=KDOO<sCzIl&XP`UE5toK=uYJ%-{RZVHj)gU%m#A`{N>
zMtHZi%eB!+{jk7q7gBfZ+?607*5DG|u~$D+75q9^Yi@6FE666_?tX(@4{|?z>LutJ
z6+l=PoQ2*%l$cnPR8}ei%sy|K6!BP`>N;#Bp@Hx+Y~<=J51q@;+5ML5in!2p45}1S
z8s;D5&#+BQLeikZwzQL^<Ws%E2V~`SZaLZHSPIjY;)xk^f_cw-3*H`!uYM=JbhkKo
z3YH4Xt`0UqKcFqNi;X7kMA9o)G<*n&J!KZZO;{%R7fjN^T|fZ7Gbkq-DnAC54^9mU
za?o>~ZP;qAXdu67!yTC(dT{pbRRR=d_~jh7fD*!^{(he;q1F2Q%AcQvFu?*`bkX2J
zFxnH!RqWF(l@8HL95>Xo=WkD68l>e~@MR|LE9d0w{?6a)mC^|0?`n*u9a8@z-pQxt
z0Xu_S6E*|E((Y=la^p{_xWaJy><2Wf_P_4|_xT|?g?qg0zP+?`j|H!OKbTNgdfmE#
zBi3+NNLFA``j@_)l+UI(Xv-#`;tSBLcyw$m$e>O`QE6eY{`fg<@QeZVjnLMaqxo-m
zCf!1U-)0~+I3YM>cqBxo#EIb#KRX&((x*0RdwwLlZrhSkZw6djd-@y3RHf^azi#cu
zSVnVa!Q~OI<5HMIYPWdnS%umyja81bUp!UJLQ>&T>EbGD(;P|EQXd~+7+8l?i!&dE
z>^FX5fnCsP%;=3`ZVc&foI<92a8+}qi$@*%8Eb7sg;PJK>+XItIX8oLmfMuWYLra$
z=dR1#75T(@sv!RGVV=cH!-Or0a2>X-*mU==$EX)C2AXxQ$z-1ni7*vy-@#mtr5jI)
z)L##%J^bo=-QLvl%qk!5+!+haD|K{dl`t~xq<B(F+z+kwFPw|vi{b^}I2d#5H3f^?
z$bQJ4o@2dQbqttcC6UQBrJHsPY<TmE-GkJv;m$z4n{;4lH7J2!59|q4jgvn30{eb>
zHjhb)KMzfA>gZc)>A!@WOqH+E5x5c^J@3Va5Pb1G!YBjv8NqJ$>x5IvhkIVcT_t7K
z$eu~!*C&h6OD9(jh8~gN?q6><dmDVYIFU8xB0{$>D}M@&w8V+B6*)-MuuN_Fsx9U0
zJsHOoO)gzE7avQx<{Ic5_w>(aERiKlSiYEkL_!;aSGKSKm>xghcc8g)(!ssrFY?~p
z!(XObENm<_Zw8FMaRkrQr5_N(dJ|D3B}aRT4@Ly9rPp7-bm4mufVx(5+nDSarNO_e
zVK0yk;nEC_N*6ZDv#P6>^n9Eqc)$5<YmwZXS3ol1_R<X}PKi&;hf*aKd;DH{Ciq!5
zKl<VrJ1qN9xGv+XCT;~`SV#C4JI;M2XkAHc#LboiJBJiMc$0<BYzOPow(Sj7OLRx~
z#QF$(;cJvKLiE#iamsuu5*oZCU;di@B>V^jJtIl4x36K`tzu;6GUH3@^jCzUZ!5VR
zw%VgCKl<#jcHYM1zM45(tvMUro=o?kk33;3BQpwU-LBH{GCn+*u-sWRy^;k|(#G*n
zqM<XL6zL}ec;bD{(SoVJ(PWaUA-Acr+{q`$7{-Igp$1JLEEhh7np!VA3WqO2$*4e}
zgj5o*>96ooX-;>`xf#kw(Af~XB7#D^NZz7UTL@Lm(0d;fuH|*O&rD~2#W_mqop1_x
zl1SZeKW69&ogBB&51x1mZI;@z`+Q18xD5Yh@3D|lW&E@Mg|sX`gpe=N>G2XLF`-1y
zPgR-U9}h()`u%!8`);v0blV7rM<E#t1Rz@8n06QMS&<meTY7~@Den}U!W{^sbkAX?
zE}yaH`PwZSNBqlZ(Q3|hc)hHzlt*s%5p7Ze)zD+7J7npnogA`ntIN*s5p%*8QfOjy
z5@n2nWM&Y-k9MCN&+&b66A<?o`w<zK!4rhQ5vLgZ;l87C+uZP>zv#Ihq(0P*-2^GR
zfCt%J69`$}MnD?B2xUDBIIQ8id9%<6Br~0De00lDv0R(k=I-IM?i^p0#(*!CaOm_e
zo$c5Tvjf`WQ!Nj6%64(wA%E1pIiZb#m8e&acgTpIsqNI;1m8H^F?`6g<sbeo)A3Yp
z|9i^|pCcN5|JO8ycWw3^hI+TEL1Nux1ED6Xo_F9I;%~^VICCo@C+Dt2oph((x?EVh
zw2?j&9j32Gr`6K|%`e#7Td+_2<mSLmN7V`2sw&M5^_{BtDk@)6&T%5sCgS%<0=4F=
znUnGA7r`T1W&+!)2X5Y6w_#?9mjclj`hSb1(TG7wTl%(+Vs(zwiA*-mL@QZoS$rA-
z+k&i4&<TpUyUsXxZ-O=uvN@dqu@pXadzJaBfplp;xYZ}QGNcgIi}wOHer0C(Fzd<8
zTk=R1E{8J5XGMYE0=S(Z=1ZpE?Y!9OC~uA9c%d~WGIAVb^Kcv1cCLls$ZWkS?s)P9
zan*sIy3V*-Rua&Q@LNA$Ig{jS-rjO9V@7I;o7U^yonG*TeYGCwdA!wIG=7V1X`p8N
zJ8b$zIjjyQmDB+7IxMf@j(&)RUqloUpdWGCB)w_xImAzJ|GMN<%FnLeJ<qjni?{eS
zGskt>5X*@3DE0|%w1b6=Bz1}gotazEeh49z>gCF>YdpJf-QA1|W0>UiLK;VEvhqAX
z)h`wEIM;$u&n^4wg4R{9Lhfr^@Dei^ZX51qc)|yI2nvmoYx(kLetl{t{7NpdKl|db
z_+z|Vqta(&hodZmve=}NmCZXar77`-H;rH~)O3B}Xd->CTP7pslXu_~blK9S8PZZV
zW3?ed4Eif*h$_nY+?ZYB^XZ7^g&WtyNg<_+F`w0~q=e2O-}c1o{g9}Glph~>F4K<Z
zkA&m*Brqn*c+*FOsET`V(I;FhCbP2)-MD3h;wIZ3_bvr6b}ON1kEUjr|1I5bu`c!;
zIyr1uBT9e&;D6i%b3J3&WpKYJ?D8G$8P}URJ))PorD!lnvYzs#8lX@&x<~<EtKbQ2
zkg0LK<VBwDGPhqHT&E2Q@?Q~p@858GGk8hV+<=lkDinN5WvF&?a^f}!dLYP5YX3Sp
zX}-6<6G!^9tvh}9NgX@=4W^A!jA@A?ABeGBgP>t%EO}pYPhkT3o?oG3NNvP3rEbc{
zWHO~CrVMirm7->R#_-kow-14WiO$ke5nCC`&bzjxxbT-t-M58niQPgX(7(yAWn~3-
z=T=kLyVu=2ftYJM|8h1rZ@na8e4d00fD_QMjejcFG<>jK@O1u*jI78{!NxUga}P>2
zN<Z~e%imPhaP;nYZL3EKhRI>YhpVHa-%)t)8h&2>Zs%^(qPNOoEK=+TU~HWl&~DVa
zT*XdN3+{EswdhS}Ek#Ae;5D>^(32-m;vI<1i;Cfs4nnBQS3KV3<#vZE$ICMMhi}R{
z>ehJ({hW<9{5u)B-eTVgd@{wLlI|VoqsILtSS`fv13b^Si&TrcQ)2xG?3MiFcTC@U
zVXRT-hKTz|?;1{4lRnm;FKsj#Rbp@1jX1o__U{m*jU<vW24AP8g?vAC=w5I(<B#<Y
zi+Wf4rca>hU~;~5UBV;a_8Q8p*5_u+uE&bk(ltMi{v2wIyTn8O|JAs;$1}OdaeQ>5
z66#QCdAY0=uX9}5lxz%ztZXi?Qx2)6W;R+Xx#iML7l|UL=CaUni4u<Mm`iPQiIPa}
z6y1brWs^qE_vw%G*Xec6zkBxV_xnA+%bxG&^ZsP}O37KzNIYZu`qQk?wQu&l^5;$&
z3;iedE5-K!-qizcQPxB-q1gN;ce-&)x^cm%Yhfx+D8|P=i3+rqSEPFnz3*Pfh#9&*
zXL2vZ)Ti5~9%EFna&Xk+uC!szfsY?Wer(>WO8ZNp`Fp%P)3qCqaG%`l<fCHLrdsma
zY;sSqGy^W(`|!w+&prHBGSMv=kbSfl*S*uJ7vu@uDzBAt8V>?R#qKLxD%_;C#b0y%
zj<{kzHgThaLiK9R%tf6&ie=s7Z&GGec81Ya=ZUm_*7oV92#vA+S@fUPl^Z3KUEt{o
z!B|$Z{*qy1CcjL8gEc<SeWlj(!kN!(@yU=J-IwwK6)ml~OzQ7S6<72{o>4BZ9h>U)
zNK=?WArUuoyVpwgsb(+KhaDY$OD>gpINwQncVM3&D*COYs}XDEJIXu8vgKG@Y#L}h
z&yl)D=duTk>9dbS;t5hlOT)RxM88hKq4BXnkwLwa{|<ChY2D}$Evss4IOWE+a5(Xo
zg#3i7Zgt%*zM7u|rPN;hVwRsoU{|D9(SMBJY8VoHEiEtW@Kcd5V9#w+*KsACE7Gst
zYP&Z)m(mbHV+Mavjto3x$!|0WWU%B!5+^2RH$}G-gP~FqgingTF2o(3qBRZZQb#{D
z*9ED?m+*~NCa$(4K`lnzZSQKaKzPmZ@F>@z8$%KPn|fIcS>w(}BTPr}KSdb|u0Uyx
z`p>$qc;31lwm-xd5NaOo;VpcVuf-j*k513h_g{e&g>N6qP_>{a!$mTMf;Bb027p7V
z`~AH*pXLN5OGmrsdPrkWqVL;T;PLpL5{3$B(hw8<8I5j%XyF6kC(skxs&3Ip<+xe5
zLT0N=3jYJxdJRZP)f583mz|qCA`L%IyamQcQgU)LP%%r@)ZSjT6nlfmWTDL{JslW`
z2ML-k9;i~dCCos_Y7>KEfThqhP0RHxR6}Z0zah42JF|^@8?Xe%3osrP5;z=3GI<FE
zcbKo66Ty%`7j?Ie7u7pQMp}W+{L2Y0!!xXQ=*&~ML0Nly`-n8-^f$t<?gfUWMXe0E
zVe&EuY1+1JGC!Eo^@cSOQp?`V$vrDH1437Eo+YKE)h*c?H|`U<>P@i9{ZqQ|f^mIv
zZ0Zxr-_u<(2Vr!wiwhG%Oj;Muc8>W#Jd=`wzmQ_a{5mN2>LSeb(y%5bg_O)wl7=;s
zFsoBt?d^RF?~n7HSol8a3>6c6_s9y$#G%ktpth*Pz$W`E-#tz-W4vH6nRG2BknFkE
zQK>LAv*RP8kx7SF<}fixI99!P6IG+NBze<8MHKKaBo`}*efhh*fCxVv4kr?cG)*D1
zrP;ap@T}MO>CF|rKn(&EP&Y5@$J65hzS_ZN8(louUU9^$u;$Y)TEt}e2puA>VE_|=
zReRed!x{lRP!q%T8DM(-<^Ay7nKrk<5M&M#JJ_{~CK${)@I*>OCifb(jsTG-YHBhA
zPYeYpw;he912(a!#0V15o>hSxfJ8Owc4r6)cP$-*`@d@nyUYq&l;+G6AXqgNHmYiV
zbco{PgH<O@yQauZ4^#)`e=*SpV$~NYtT=5#$v(^B$j_=%wKKFT9oA{+iL7P$I`cRH
z8+H?YVRzcwr|R}E3PQ?$Sf(AH-fwI~*6Y7&8uLMb7(sSwzOQn6v<~S__DQ!hG9f_y
zF2i+Ep1@?6%LDq#E`z~T8YUT0%A+6+pi>pgbPM@nqCg<<DDhcp;LObnvH@8FV0Z;B
zQh1Qv4FjY}RkA{#I%JlNXf??N1qI8d%=Tlgt<$r!v5>kI!10YP-i|y*Kf`MEvJ5L7
zby(p(V70S{>u&LQJp9aA>rk<F)YobKty?+K%IY0CPHXs1h%%EHo~gsKsoH09)AWGa
z185dUEqyF=xU%}Vcz}Q~1#mvV;N$X@iMF9bCk@8CiwyMj&jQv{0st?#{aO~s^gxz>
z^Wf~F^XJd^?9W9u>vIjMtukw@BZoH_0DJOM_xNIO$V`XT(rDph8!j-BhJBqCh_>Jz
zgKtXU5`$wQG^|(*KuSQ3WQVwV+$8Ou2J*G9x%%2$5)u=)Q7FrqPtzMVQ<n?9!G;I2
zuUXuIJjCqueC28-m{j5Q4NkG9ST2r!QBVrKAgkQgd5^|ZlOQbv>yW&2C%AlSpw4`M
zq_*(C0&59UD>=~r5DxMoMit&Ck>A9k(2M@>QTStt>qn>Q%LRojH#2mZqb9fnDAY~|
LXZynK-sk=XWxssc

literal 21366
zcmeFZXH-<(wk=ws1Q9_IP?4aLbI#E~EOL}Al5@^k2?_#|CFh)!oKX=Gkenn`0Rlo5
zp~yun-Ymaw?|1Gy=bp3gY46Xy?b@&{s@9ro%`oN|z4tK|Z`4%et`kxcLLiXq3i3}i
zAP}5k2;|ZN0Ur3~ZMS{}IN-g2$~^@~@bCCZYY;eGb(Yt2gFr~#vHvf@Eor?WkoypY
zr;oI}Gd5s8-ed=hJv&2Nzkka=d`hHy<7x_N!Btj@ua{r2$z8LBT+PW#&MT>pBuHvt
zm9f+QN=tT0gr8CRGeqW!sv23vjhl)0zBs7i;JJS2-5Ny2CWYOJQK~T&Ts-&93aAm|
z6Wg0kits-=$-J9xdM6y)tHdZ65_^>PcpQT5w^EMuWpH3r%ee#&x0u0*Ai-D3tij>a
z6+(V+c#eZh2@c#LI7U|+=L1{Qgq$2bh`=|7(qwCj!rt3CHMO;sF`#+ywbwY;!HM_M
z|LUzDzncK<LVCD~0`^DbPG(Jdkb7J@cFXy(`qZW0A{}P58s}Xi()ev&c#uI>o`TyZ
zYN)B@%=+!W+w6H_OG{Dqoh_|&AVyiIr!19-<#fH;eM!n>_@~Ro@H;<Lz39zgSM2x{
zef+-?Fx+MEZL#ty&o|`9yc?Dkk~+U_l=;5oaC6px8vm0IH$Ej~K~qbsz$@iUH@^z;
z+#@kcv3=WFBI$ijX?43LwgcF49Z?B8OAm0q(fN>RLUMaX`(WLsr=e1(2ckxNcKs@3
zd&gVc_2<#VbKD3RW9+OMn5mUd^YLI-{>-X$^e-Nw=IdR%L>?Fch0S?+u_Zgm#62*>
zYWZi5m=_biI~U9DtsDdDzjnT_>|`7W1S6yVn$%AoV2M^dn?A?~Vb(O;8{HZoK7Kh-
z-o0{Sx4Nz+_0W(rRu0@ybE+tMqR@pd`rN)g66Po|oSP}@fn<&tU{nw_HmvBZID5R;
zxbY6gc)g`8TGk{e+BV_}G~RXh8ozBfD+lbOfyd(~fyVRd4~Qe^ZHxv<d_5;p`Ud+o
zP2WJTrc6%N3K>>69B_?9I}J^xU7m-=bNU?5>-(XMY$^0~xhjmy8#`z7=JFnTylqzs
zIzA9Ujuynx_aZ#YF>IRr3w=7siM(c-bLT=GlC5H>os8B|D@3i92aJ&UcC2a1Sh|~l
z*PftE=#}!Erl(_*LFJOyl8+juefU*th3qxfPg1OBKeP%wMA%BSBB2~}4Vv-Gf<#6_
zelf-uzWHFdNlg;I$3c269)n`>1!mpdwql#g1JiZNxsQmcC0rg4Rbvzz2Mp*HPI?jJ
z)$<wUP`j}8^xV>uEzF~n+#Gp2DU4El-HQck<YYxDJz_gMymW$Lq-C-4V(oas1kv;5
zu}oOF%?kV+nY9xmInzFr()x92B%XR2+38%$85^m!J!H|&JC7Tj7Io7aBHc_SoW-R;
z84w^Bz3HC`x;GJEQzJ+{$Wd=J_Pqr;d|x)+-EL3#++yYS$yQHSOS|^9fXuyDo-_lR
zu@otG#@RGA3N0c5A(K=`#`t@C2d0OT<jGC@L+Pz+iv{H?f<!P;Ce89g>S4H9kG9xP
zF~#%W-v(yDWVJ7)+aVnqOO)H8{yi5yIR>8T%ImIy!{2C0%B`aM1sYXJMC=7C&oS>r
zI|Wo@gZ4{1Xhet69c{N%sQr&F$%Ka=ZGwg2Q^%UzXn`Yj`^36Vd#y)=@1Q|}@1m*I
zTouNqU)5OiWLbvCbCo)Q{=mj3%t^6FtP!mRzgmoph_G87$);Dxm((ldVUtW*rlw9V
zc!C>dOF1h&-G!QH8<?$gUv-qI|K%oe+{=1k@6!Amy{X68A43N}hk01o&F{p&xF;Vk
zxJnbZT37B-=X67}YS;cZwo^P|mmzAL*;^7h_vW*zeQz%6`&&5;AU(<v0sIw4(D*w0
zDVn?SL<;S{IW*Esoyxwi=p?jw&?}5NtrsoDN|j_|#0^!AQwknz&XhTvL$Up+&)@~^
z+Jb1QgU%iqO)=|0J<Zp9zU1-tw2GHi(c8d(J9N~vEk<<37g)x5SV)N&a<J@+j|r-*
z#-ebf54s8XAvuPJ)KZ)Zk>v)ug6YaoFLP)-QPM?Pzrx4sWYiR8#|91S2@<E|=8Xuo
zN4en_W$h22JQ;7)zIKl2N@_vFNcJ0tCfB@OKE`A@+sd~#l#x?a+)EeK`OsCxB&ApQ
z#%_x3>!RABs!t^GHEs82B?X%K@NUOOQNcDPj=cOj=UEj2Z;!GUc2lDbbaD&|Trk&H
zg}@Gxf7#*Oh^X=9YWqsle#!A_xVHi5A+;G7c2D<2MfKOm-iUajhbA81tkho%?Po~o
z8J=%3EDnN=%EY@Xc)u6CFnzt%dBRF9v8eUBE7!`Y$v7d%t(KnUp|Ct?rUh(ueeKPy
zJ`Q$`iw2H*OZgz&>Gsaia3)4(QFg8C#+hXfaxu0C?*e^=5w<c{@NIx;xI1+9*BuSQ
zkt4rcp`F(t6Db4S>>Wm$0+XehOqF`Ivi@NZ+enAm1_VjxdZk5sLS>AGO*s9mPa`2!
zx?Akpn}O+*g}tH57@P(1_g^kSwq2SH^=sWyCV0;BYF0!xva2&aU?>CQT#?Ss%6K`5
zni1BPCFNpiCf>x8mYZD(y+tQyIJ`vO3L_-^J;kd={U-pp4G2t?C@8&+;fY1RJpHxs
zpk7s}MilR_FyjC6)vJSu#V&zg30ztmo(9ybc{{L0!}b=_gm9_d*AwcH#LE}e1Q4~p
zlH%;i1i2Vy9<z>?ZUObpoLv*z<}Zpk&R=JLzbiYc?L0fCgn3#RI7p@yH)8lql+y((
zN0%;blT>|CziJqESz6h1;nSr`<92-}m5h>^%Xh;K-Nwoe?~p&Y8_|d<uq)rcgX+BK
zz`IUaP-WE0N?+jb_2|lT2|{-U5XV1}h|`E<`;H^YEQG7pD{>@RKy<nh{QmNP8Ji73
z41^>u?s)V29kQSMA+H3~TjoEs4J*YI#_9CD414(S%f#ky^K*+Ie73_mArmg_GT2C8
zcqcq}{IeIb6-K)@lepA^kz}cuysDt>uk3t<-zv9mCF(BHd-HKX!){PyR8*~Kyag6!
z<RlhCQ*6&CNhZ*xOEuFKDzBt(Qn4}|II}!N%3b%{rfF|nHY=yo3m5rw@qRPkMAa!z
zAKzffz(M$Al~MVxUoXn@pM_(iqvG+-_JfI-22os`qyff>vU>$)Ju!OJ-Mm;JqlTA1
zB%Bg;K7{`dAOl+1l6Ac!ik)C#t=fL?o|p3rB_uv7wUvmG_!5+=Mv1*}oY^fa=^N}p
zGqghGMOF_-yMJRMMrj4DnR#*Z=33KR>cK8G^)1=*Cy<g}aark@-((c4W!z4$CV1G@
zf0cdRW@5Lhg$DWbAjt$kd|x&9Hk=W!ZlB<Rj6WCbDr)fkj=W>x@3kPMf3o9TZuacg
z^DOsnk;&UM3CW!gHlMAw3p$tE6;xd=c-;bz8)6}Wy67q>&nmLPljk^Dc7!xm#oAV?
zn&eN8IBdpl!xKhG9phUH<R6&A&sS-aBr_NsOqjyLv-MF&&Gr7Bn);~^s3qu|Ge!$j
zphCrpJjw-n`Gw~OZL<<m(kESP8t07h<<Im5OlF!Mvx;*HM5eUtC;Y0~kJ0LMP8qK%
zN7u5aGhPo+FC+KBppb_1-P4EnsC>7Ef=1!WqwxxGRJj=m;-sTk1uiz_ay^#6BbzKB
zb-!n-Z8jSvCN<?1S;IUmFs>@L+Q0InazBu$Ycql$5}fvhyFiF0WtSLQpI`Xv)@aL$
zhDw9`&>q|&#vjG;;!UF>opDOXvo#K<gL}}*>VzVxEf^!cVa;se#I-`$z4nz+J!WM}
z&LrVN{WuNvj>e-c82$CYK{27DXr_WgA?W$VFOF3wjLgaNBKN6hh^-*+VyO*%E{Rl5
z+%RvxA#aTn(^xVx_OPiq&!F{yxdUC6s>;SAr&Gn5507m}Pqg%^H@tTG&u2DQF=tHY
zO{%`@Hl>adEEi0OgvR!HlXAQMpt%H1VWOetSf36QnT2|eT?uMRqHFvbBzT_s)W^Q4
zKNEgawGm30{qX!l<G!al&#1*f`;fs|pFvL0PdXW6B!}w`bzveC!*s#4SzG?Mo>=cF
zr#G6&z*u(&E~jHZyNX8X)pKv#ulwn-h4~`zxG7^2U%6COQJ=8-g#7w>)HY44g&bm#
zsXc!I-%FlPRuMJyZtEh&tMwM{=#rCS?oG@&r^-C)Gc_1O{lj(?i^u~B!G5v4HkZP>
zHEqOppi>AZ%0=}gaH=`>E}1_DMOLVu<MKIl(xIxK^x_3TuOZ*e<ZGu4GV|W{ug{lA
zc2b&q5JCO5{MrgOtqgDLnIQMx*ZI_Y6$rHrrlM~eqsQH5hZ5$lD@alAN;<j0N>|Qh
z{h*dljhsdj+BAhFeaBAc_VzA_IF1w|-=@%{aysY(Q)LSfHaK~n+(=a{?(v(w8NXd>
z7Uqq7f8D@4hz7n&J(5(=&&d_!liYV<l+iK27l28iGz>%@6uQ8(nK~nOY0w3G)MGcz
zF=FjMvE9EC)>C4<?pG0W>HaW)<4E;#%|!K16+0t}6fh*{IB{lfE%}<nysuoS6D9L?
z?-{X~Q<&AtyjLc)P0Gt$;i`~tcJT#F3?Kc5pN02ZYX!Os-rN-~rysY-F~p<XhKrKo
zl5Zr6*T~+f%09o1;(zn9O_Q#vILdM1{1|2hXQ3BNceifz($jjSNjJeH_oBbNZN3|E
zDlxxGLVUm}7C_j`<)r`Ou=DeAOtW44F>3#X9M+MpBzyfl<=fS8a+&QoTqT!1hR<^w
z6P9y0oXm;gPdTCocQ(D#y%HOgnNYo+QN~F+1<=8Vh^HOy2OJvC@5d`L8w~@wp(neY
z@KreyAM{=0S--al`4x^S%LK`k^@*Hsk{T-FjFk>3`sOaf9UR-yUjlEho}Oj7X84*P
zHm!<T`gCX}QoPnB#sRz**=OP>$CGlF&A+0qAn5dn!^O9H-&r!|K7!;{f#4LBp*y~Z
z4(6SbC_dgQ*%q@u+}X}_pJE02=n=g9(f(1sSAcT~`h17Fef?T2RM#VJ>ompO%exWr
zv&#Sj`k*rMVY+?2dXph6)VKHzzvBvZdo;@1v@;5<Zkh1bag1ITFve-3F6PzFyF1WK
zQIF9y0XwqSCX2s)u3qvYHLb*(D$|}9z{WOfq9;{ojGad$?}p2Oc_;5@<eg;sj>@Ps
zc$?zK)s|!<kLPRScf+>J>(ob1kb4egCY>D<>g=D%h#|rC_JX*u$yp?s@BlulaulUs
zc5-`7F?n_MBrCvZeZ4?TR>$#TLJtx{ZLIbunkbGn$r45hKldQ&cM<BWWn*<N*QwOk
z>G?q91`vicOAz&`*Jj_Z4W5i?Or^$>Y&1CyjiU!yKbB!jvy_&;T9za&p1AgSh4rK8
zb)4`Ao3rd3*;J*y)dL?yqYJTE2BLWN>eZ3RSe^BqI1;;)E)+BU51rroUnd^DX_a?W
z{W!tXM{unw@V1WR2Qk%n_sxT=`ockJRn5OCqi*T_VT*kwn#F(;p_P<V#h>8f;_CF2
zzGIjdFQ`TjHi*jJRYlJ@d0ZKrb8P(5)kRK9`obeI_z!2dvZ$TbYV|zGAfcwXS@Dh=
z>>F*PPSuPrRR5tM@LvLo|42u8@Er*WH(g6ml^mICcJnwpa)J5p4YaIg`7JvOpLr6c
z^cU9E@pG7T7^<tQQwlrhh<a^p7<9U>C@R0_&K>u|Bf685m`K)X0XVhvz;mAv%yt*a
z!DoEhwIp3*bv3nj#=Z!Kfov(0^OJ+USsq3Jph`-Z3kwT-5$?A8Ya@i$u4!s$#M85g
zi5V<?i`JZNa5(o=0yi61N>*8PZiV@tA50zmG>c_<`0(+A2M;L4y+2TJ8m={(QHgmf
zvFTK(+1i#-5l1|H_|P2<^J0^bfHe*bW{Q#DySLtG26j%)?;7_jCq1-5i9Ji|ukfz>
z_frbk7QB5+G`|iY3+M;}un-+<>k{7sWFfrWuiudicHZa+E0c+$uvz{dSD1{Nu9f@v
z@gw2Yt2-Qab4?#TH-0{7oaKn<$_0R!MMA=;T(?^1*|YasZF?#VOicZjwb~<0_*hZL
z!Y9OWBrgmYBz6+1r2-9MEgoY=Ua-OsAGq&^=c?z+B8t)-M>nQxS3Fw<9A@~HGexz#
z!mqQ{b9Heqd$#sdEe+>Bv7M>2DYdA#ortFw$y5ObR9aFZpDE@wxxLusvGJ29`ks)A
zh=>R!uep3F8Vgft1NP%k2faANK<9{xDAW|rP$_P(pU$HVJbIMCsEmD{MdPf1?Km^&
z1gzDIO-M*r!tXG#Yf(*J{xufI0x;OOw3B!${OD2PMdtH6;V2()hJwqu0w^?Kge<(g
z8g#O6hhO4e>m%FUoNcsQ{1Vm?(X*6`6|fA=*{MT_Nk~#)j=J+50RdwSCkN}rUT`0|
z<KyGL-rkXse7T86CzE+Um1L9RvC{KX^l6d3w>S;xj%2`ze0+;LBCm(WPl<<zN7Q{y
zt#pZkf}#&RRb2e+Z8LJBd{s>L>C;drpK9ZFcVHi<#>GWNWd#KVH6~d3tn4QAnmd?9
z1|R4S*Xlide0(-hQT^-p1jiFuH2wS0o3q8B18m~rhC4eueLsFA(mTx5DMV3ljB2V%
zXxmD;xVX$TyIGYYDE!b{+K(T<v=>arLR{Qy6xL@)J00ld<z;4G-WsLPF)=YZ?(TJ{
z`T<Kefi$oIgVTmaS*S)vMidnlH7qTQz@`R2zX#Hb4IG<gOifKAV`Eds5aToTb{2QS
zf$<CY?3zn#f4XdOc+j6fs2Uj=sS!a;&F=&k2XoQZCXO|T($doSg7T^s#o0!uDKL|u
z<?h~|fJGl!VNsDQ=>st2z7J<Aq4~1W8tUqC8k(8{=HG9KP!UV)4?l^g2Zs9Q78A0(
zx<f-zF_OlA`w_6xa%c)?Q4EjozV*t=N~uX_AlMa&TdgQbU}lQ(WE2!d3JHuh$Ga;n
z=ZCWgt5O$wy}iAPDk_JkR=B{l53Km!zfv>(I}YHuQucs2z}g41_IRsdE6}7<c~bRJ
z6c4iS@Z<LSX}u%bz1%Eu$3FYh`gQJ}T_DDQeJT3KRQ@%s^S|ZVLjD)!&wnbb|MzMC
zPHg@&CC^)=RkI~wT(e;LAQS4%gmtr<hZ5n}?=hzIYuipEV)Nx<k(g6dHF6GI5X|th
zuU6gLy9sOs|LRw-UJW$|xVpNoXn>Fkxm0F|>-+mh>aMqVe8J~e_;!;OdNs2{;@+^F
zN0&V|XC{KKg3E#-8dW$F&I{7>d(CdEw%^}9(54^&tqIJFR394Fm@8;6gHMpMFxP15
zi_zljz%uL>?TMjwZG~CQSx(njBoSg|^(zm@aD4|eL@2qR2Lr*pYO>ORjh?=M)%8zT
zdfh`Xr}Gz8EBz^Mh1h`_CMpeuz+1hlTRhfv$o0TC!NFE7(o75th@+h)?KrRjAOzk;
z*Y@A@n7I#0pi7#Xnpk(imyk;X`&4Ta<+^q|zq%uVZvb(`y_X>+W2cDKsck3R1P25p
zNXS?Bkid&`bkh6x%nHC8s;j8P%s;_)C8Q5@<!o=L{2>S*5H*hXS$|Xl!yl~LIZ=_3
z?nm3EK^E8xag4(V_luO%6B82)UcV;j{`&PO89Ym0PfsuK8n*2f$}!xNPa&5T<>l4<
z{o8o?_@G81`C#=Fm6XPTNO49<3a%m`1M}#=)x64Y_KoN|jrg;85cK`mzs1s4dv4Cy
z{CLmP_x<~}i6Xf4zVv&VkSwX74iNC2Cl6RySoB-mZ3-X%qua!Ugz=_#tmboJoWK}L
z)j$*J9A2ZGQ#7jDW~?aFBS2O*-1l%p?;ocm@O?Laegq1{SebS?KXCHRFh>gu3!DCA
z4n<Yfsa8}+`NqaZ*>`M$=h>ip2ynHNAogHAn+FPR6z|`7h>fj+_CIqh2hxrO=)rY2
zn3Zl@D`3TtWDbKO;5|WLQtj^Uj*O1hwzse7-21cMg^zJh3ci1rmywn2<0jgKw;SLS
zQ>g*xd6tA-eF9<fmRpPah_QhDfpAl15fQzCbRk`UO+3~s8~YIMC;xoZTG+MQw&Nx0
zlzf(N=@IMIos$u?fhDDy#kRl%%M2Q;u-K;)A8Rf+AqW<#A*qWtEEWLf2aU(<WnYwl
znQ6JeoH8h2tph~591xdIg;=mdfhfgFJIytza~L%2w`YlYRe`X)>}S;IXtX(9+arYl
z*91e%TBL>SCY8d%!r}`uGN>%9tk#7(kE5N=cb0mZDY1rN%MMKc=Lb{dRCP85hkk*9
zQ@b{RQ7wlvb_@zw{VhLMvP^8NsSEW8Oir-$b6?_p04tJ`k_zW9FzA}u%@b76Mft+Q
za|QVN6CuaBXX}%ddw$sUy?+Ckgr_i7nuuEwh+A49TQlp}I?(92Ia5FL(JUe_KmYu1
zGb3*yuA>LN?A`QS2Y2tV0Fb(7maPqnz0HDl$)+O+z~Eox>(^>ZN^hNfz(YV6RvMkb
z1=2Y@MmZ!@!a1R#SGs$8*!cKr$6vc;O9mvqe_vwM>iN&L&whF|=Qd>X?d@GGN~A|+
ztR?(!Vai|}%nG*up)vRR^>119C$n&Ish%EgYPz_5d|_q=u&qHU_XY4-{8och?Pt6F
zNi#!S&4!!+&0<3~Sd3Ay_a^{`nLkNl)3yB)dbJSF_xuy2u%O^b0BaUyQot;Ra~`=i
z8-gvaW^P{akdE%0wBW-BHLi}MJOBokQ_eFTe|>#}MI`$AX#hdW#?n**+{(bfpb!s|
zi~^2x!u1Zblc*lxVx<ROoX~%H`4VR!l~1lGmKJUIZpzqC8SFpwx+DYQ=SMUBk1xUi
zGFfpJCcAkv7i@0>1On0F;JG!YhZq4VS%@JOu=t@25lu1AjbY$f`AxrsYIr@YC~@jI
zmIV?ya5-t>KCY!6U_&cHp<@8jR}2rU9Uv#5?S5!Zx4BA#Mlgvolp?O?r7M1i8|q-o
zKYsie0Dd9I3TiRW@0*_ei_X@7VPG8$K#}AC<%pr$Gy$@ynYK^D*kdC_>;u^DK%-7i
zi3W#jERwhUNycYolT)uU+ni&!u`Q%tb>;+5St0^Q0u0|{ZTL}j`(c{pP*$ZQtR;!H
zm?;UZ5(-q}VSu*Eun6h`b6z+Jpxpl)>ep7zZgSRMJc%GK1@53Qd7?}k8o%r=N=`;r
zkOxwCV1l|LNTxoT_3(p7t}HKGgE&|SPf1NJovyVyHEuQJ#Kp%?S$D$$)LZ8)(aIx~
zRlB@N`2My1$m|9^>Rd(G!~$hGot6k#^06Xiu;eWVlLibv#qMNeWLm<Xfa9t%?Ft9r
zy6s2HzhMIdmI+;pz(!srrlpnP(|G61<<A2;AoUv|^b$*ne|1L%RMXMX3B349AaS(t
zs>6nsDJe0LSxV~S?dc`R!s8}RyH}H|x@TxkwTW;i=+x^?x3ijRfj6p1<c#uJ4Ob9|
z4=L<=zZzdTP1nfy`1rK^PUE|P0(2?|Jd-w-IrOU}>-;MO;ruUxkomC=s?qLy;NSE5
z_3Ld8RRskW4~rO;OwrPUg0CCEP;<Uy(8PkD2_D5aA7&?bfBopxtkkNemv=>Sp+kW)
z`cn&_pP5%7%^up!fIG7){wpP*nJCsUJP7mF7svfW&Eu4hVl03#H8(dC5))4rrWhmB
zDb<bFr>ds_UhpqO?~{tV{CX-QCsz(;HI3II5+wol?%3K$KF}Vjj*h%7EG-pxW8QM<
z*U76&9B7rAeeLeHSsl#mI3A3z^FkRj(9?gnthE7dRR6>C5a7a4eyRW?fE_xd|KSqk
z6N`hL%v$-=3loBz8j3(8)%D$>PFUEXM-PvV9P`q_t8AN}lZ}-z>JD0KM4#&IrdTB<
zjStqRbb&H5-sa;rQDtO+WseOGE7?<Hok2kypd+cNselQMvsJg{fpE&g$ET^B#;*>b
z1K9Y0`>=I!Ydw2b(wo4v9)eH9eEMjRH<ON)RgqfMeYnkM&m1t4-CceUq!wT#it$w~
zNLGjW4kIupDW-S&hK8m9@d8%VrvRoH!rbK^0&jB37xmmIiJ=y=MT{2CfR7!=zwd#W
zDKl>OwVfzSDkU{yeG4-#BM_q^R?ZMsHwidM($LV@UwB2r3SzQRi#t#KJeI}**9f$Q
zu@d!s_xS*Hsdl+eJaA*&p!*8_DO^*Vjq_Df7so?@yXXM;877Hkh_-_1rn^c_Td>x3
zz<DNtvm9y)1T6amh!+B0TZWxih|K@NZFbGS4D69=CAN+r?-Y*PI#=o+7-0(uN+v}s
zjA)&iZZb}kQoqv_#<T`D)g1o#AS!Dkn2wog{SGodSCC{e9t!yf<)X?%or}1ZAh_N(
z@y|uW{ZJL_fajG((D;^U<boctmQPJ2Ospe&SLbhV_wJL$2s+FFTzOQn4s#UpIsl!P
zj<#r#zCfr3je?NP3P6@+rs*Bu!)dF72&uErT#Mi)!O~{FggtR|a#%b!kSV5%ofx3Q
zlmd*1#W`-je~bX|xBvB)>YB;LiEb%c0&}z!cnu4Co^`v2&;Sm&;n@mYK0g5Dyz2n8
z3ji<7+s^`$K9G{DO1Ibg<{Iu{f8Sztry~cyEWG%Wbox&#S3(uZ*MKLZmh}H=@)w%c
z&BbZA8Dj*Y1S{(Rmfx`gytj^yPVzifaRV~9dx)g-LNF;QDH$zo^^YG))0@SfBgLw;
z42+DnQ&mQRptu8gYr-+F|6TFxe-iZnO9YEuDt_yChMXz9yu4(0@0Nc5{=Muw7H3|e
zA7i>nK}Kc`*p6u6nYZ>bb~<s=K)iDMtu#NNAx@*JZH{FjILQE!Vih-FhWj&i)i#G4
z)5yunbPzH7?0Yf^Ks&Wtshq><f)4ThE2EW-aMoV2!u*_^PpHqJd9bv3i!_&B4IokA
z6J+9X7^j)nD1e<F%inK;c*MrYDEs;IXH4Hed!;zq5pcS{ZgQdTu|BC?YdHXrGhJ9{
zC^npl`X7Dh<imFBQrA9#yQ5>3euI7K^0GM=PJnc55|B)U=Kb^U@>!PX)mqNfTE&9+
ztxN8M9UT<(a!7pdt+D@(3|8?dOrG~%iuPTOliTlOt6odB${GRD7}@v^Tq~W!ZIrVL
z<QNd=vOUgU-=rI_v(fxhS!GK<=9ZJGQvCAeAiR~DIKtJ<jg1`3J_xjmuDJuHsm^}7
zrnc&0x1Y<S>$;E@D;YR<qlA?ozr|<ox9_hvOo{RFd2CfpkBK9m;FHjd`=Q#4nX<ee
zv*}h<L}JU)f>*!>YHAV=0Fo6F*4k+7y%?sk4wTEptgI>!q>ef6V{a)BwpJ3rx&Y1F
zfxrQzsAAxRLD^LvAZ0A{^k3)if?)>-t5=B()5*oK@6B1(mI9D?vgs65js@+2;}wVl
zkCF3@O*bTq$1EJH{hgipW04{&D{K51LVOkL1xw!a351z_3ch4JRuql$c>46K&%D_k
z*=|kjtYR7Z30A}H3_MTF&aS4B@U;f4)>W8FJM%A^w8A?4=+PtOTyrfwqN;UEBbG*@
zjQ_7mQJR-u0T#0>!Ntlty4V$QPWd-&zgG80Yx~!szNY7CsSzM!a&O+e0is!o$&)`6
z`pbT_R+itE$iRD^C$I2Hk_$ALZpF|_8H1QP2C!pdT-;)3A{d2qFt-GbL=eXR`W^0t
zigh40*SRA`6~q4QfFuMCH5r-vpq=>arf0KZ)&Ki73`9j>cAFLJe7wBlwN^v_JS8#=
za@pF3h8b8Qm-{55Z_|^FjSXthV6X1$+X4?j8-ZohmC*su{^Sbm2B?D#PV-d{o`jbG
zF)_a2>({3W8bE5SGws3?u$%l*{Q$d`(xxraPz|6?0~_N7f}9P&KCGOaIsm}}ynmia
z0&YjZOwKyY1W-QMGbO+hfWC03%Lq=~f3Cfz5(<L52Iv;xdDWn62w<hUKzj*9x6WfR
zC}t4509S~BDgfY~P70@yJE*O&2BIxC^iEcrRExO%mXF7ZBLGaFNHBw|ZxAS{E_x$N
zi-CC>|F&cWU|tt%3~_O}Qs)R&EH;^c{P$`Q>|13A6ahJ+whA>>P3!gW%HIJ?$Y%5@
zP^_(tP@b*Yp!Gi|66d@}3Zt$ndAxkrY1QFQIF_jaymZV!g?+Sjp8&q^$v)zbK7^G)
zQ+Z{w{deC0t%9cBdgL)ptelwFmKGo;9j9BZZgAwEpfuR0okQ!hd=sEhP@;v8ZwDfU
z?+K!~-~JvIRRyrab#V3G-+D%3xpcPOARl;|Gf(Mov=;a!IEwt=6eavmDkuMc&|)ka
zvu9-fJ%R1>VEQs9fu3_q1A7*agoa%K?6wY%rJh*bR?kg-%O7_v%*~Iu7_bv`A4`@%
zTIDhxk?{Hn$iQ$IwbcDV62Z~{C*b4a_Hw_iMGZTq^skg@VVCp18t{`qX`iaI85?Yo
zq@#NRB>#$koXWZ^;koe)P%9h2KPW}rOMq-$Q&UrC(N9rWT)YnRn%DO9^wc)Tx^s3Q
zhRMmvVVx7mIlON1LUAeppOZ+BU734=_G-Yn|5IQBJ6^zgSQ^MWn#RUiK*1>nX_xMA
z-&h3&C+Cl`lOC**ixZ&nHw8_?K7ha*m^!WhcF6Cc>}p_-SqG6+;0EPd+}Ls4NbMl(
z0WKjaBclR@5|BtL13>95c8}yR@=D;G#|E7}xhLc}uB~gNP44sO5thDJL;+?837g7D
zcJPh#CqAPbr{)z3p!}Dz)_`vym#QCY3S$!wSmN4wGEq%NN$LA-^VhFmtkfWp0wLfB
ziXVDDs2Z>_JK!`|=eF#1SFtV2_yDs4siKzks#XwxNJ@b(2ZT~P?i#k4JZJ_gb;|)T
z&GbNG<w&)JRLM6V0O4~L&HOiNA|T0FDg}&JONItqc0VUI>&r)Xga)AXEMU2S)S(Rm
zwnbyg=HmaTCK@+7X2!?I(<=boX(<hh)tXA)a}H+*VSN5pNO3`I!#bKMy=K`QR6E)j
z_>uPG-{~Ezt9!rGopeP1-gVkJKt@Ee4HQdZRRT^y;Ny#j{49z2v7ao9R{z%oU=UvD
z0IM9&#QzuMU#qs$oIj=l#(7}!7&`&MV2=XfXBh3jHP?(yZzbXWaonkh1Oh-DHONK4
zRl&9)VmGm0|A$r6p{_<@`K(xN1X%Ga63KU_cUpG7-4pQMeh&QT(ZL-X8yo!&|K_~B
zJboZF1GYWg7`W1xl*VoPxD-uvhg%xx4uSpH=cIk!#gQ=ma%G~@K$p*IFmZNs63g;+
zjeTT*8!MQP;la{A85tH14rP^MO(2G2vx0xtWqNuVh!7bw4Gw=NB`s^*2_pcg!D`k(
z{$^ulme0-2Z2`XQ;CMBA{p*`sfg;}k<*$3uO$aDe-=Z@!0LotT0ga<$V*#@pbIqC{
zB^fA!wG&%Gs9cVln;SXMRRHmsuK}uvLOfva^yIg1vw_ag2AzF;{NxD$Ok4+m^8CoW
z@gL<k3yVU5!0-&HFYjhIlL7N@aG1pc6VQx}om~lJ?~3CS5(4h3urf2N1qKF|T6E4~
ztyv(w5Vr-(bos1QN-eOdM{pTAB_(SYH$TWQ!raG{(<J;FfIk6(brDt_1#)B|-0DvF
zoTFE&8I`EJJP5TM^%Wp&#kbKd089ZuiU#1r0{%zNK(|rCqEr<Xl~gfK&SxK_Z!xK`
z*Ru-?H+{*b1&^Ol{bU%H-_-OBMCnQF_5^Mo8t;jMSI*5D9&Ai&0yG8$PfCE7lR1r8
zV|z5W0`PH@FkUZB?{ut!Gx}}bA7>h!xY7rTfG8Y&hv)G%3U+4A@96?|ML^U!rtxfr
zX<;3JR#0;GQ)UGd|DCWR)$A&u^*+-z38;7S0o*(WU>^W)Pm(j)JHgrm)f=rmz3nZ|
zOY|h*WIvAH12HNSD{=V##fum2Ys0x3Jzz=`2P`$cy^rqhjhE{VdA$W3ZGQKjW78v{
zIp<d>k7AX~W!Glby%Df&)drp!=Kv&v1r31LmMY~&gX99paAI?S^ECmdhc?($IthvH
zUs^wd0zmDfc`{M^<C>~TGjaeI0G$nyj$&8%Pcxdk6}Z?qdYcPc>`W!%`ZVe`*CgMN
zxZAfHN?KYjpu+$$0Z8URH?!Sc?)6;=!SAy_{col%U)~2#>igh+#dq)b05B=xrRz{&
z+gOPNWGh*L*>bgQzXV*b66}w^vXy)wTZsjHfYoXOk>}sheGqm=L3i`tzrO?XS<P%&
zNj5m!#bRZ(+MjR7hDu(s{Eu}j|2|)-Wk)?(HFa`^1Jaf8TEz_=iuO((L|R|g@ukK-
zR5}?@=fC|obe@7+cM%K%Qg-?O^J)ebAOGvGUHl)JcMQE>ei`y1yk}_yB&-jxIjx`z
z4A9FOL`QK=8)qk4JnW2Iy2*9qE<v<3$*dP1Vr~0iOG>`yG1th*0oy(t9iQ(gQEp~*
z$1%mewbotRu!46!s}S7HAQKez4yfgXME!`B!jz=@9jb2y)D})=GAJBVShXw=J+YHa
z%X4f$VO`%NubE99z=JSL1=>Mwr@pxeH0o)YTV)%%&5*8K<rp+hlV7D%WoY9;36U0f
z5`xFcn|;m&WpHq+5tWtxZs(IQS=kQ?jv+Z|AX$Ao_17;_7U6N^olBnyh&9z+K<N+g
zKSX~+#sq>(&)87!_InuNt0cwHDVVzN`8MOIYQ*3fe2PTT0|`~FOCh%BZ0D><c3&A7
zZLd}sxqhB~58^!X;tFMOz0QT4Yl=7RMZm`m-^TUv6B9|5-3-(8m)Qju?Q{}xYQ6?Q
zC3*vL=aW1{PBvAVsx3$gn*(vwYj^WkHM4_fAbBE=#=yTP*_!+c*ka#KUVYM9@XDZ&
zQTJF%3bK&*G~^p@EuYCr&qakw2sQ5NGlZu?Y<*l+n}V&|D}JlD%dE*Q7wSkc*&4Be
zUk_Q_GWHP$qdQ$1QFeu57}okteM~?`Vy%&An{Law4eGFbRl$VwU3^3eD1GR;;1&y*
zKfKQ({NwaYDU1pdOqeIV^ipIa5kBYOLAdtSXL5DpUEufR7Lo*vQ)Asxw#Er_smndY
z%xodgl!Bj-cfU=h@2FR{H?0{wg|sC*h;|C*gM7wj<ric;Z(=(<cj~_*HDs8Bism%R
zL#3|4E)e09dyY#k(D0s9|26tj$G2~|2q3|3X2kr#H7|Y@^an9a+iOJ3`=)Hdm^TgA
zIffM43x)OygHTcjo#d8M`W*oo%Pvr;Ns_IMLiCLHH6INpM2`d<2Lq+?LWm5~pr;#o
z#Nkun5x0O664*n)&-wm-tGmMM2^QxcoXvxc>X|3SS=kvBrguisvu#7IM56U)kumV>
zQ=v@Os@0huh1N;mRTeRZ>u1L;<4mfDr~Aq*E2*QYKYoM=64{*EA`m9f1UzVMOy+68
zA7}9mH{lM)3I>iOYYMEXYsqtx6dCxnM}(N^fU}VLb?k7q?(b$7Kl2_uIn<DR?}|L;
zs*eA-uZ<Y&Ov<V;wIr?^VC*(o4KQ|}6XTlwY~lrhTr3C?2JhOrRFd68wTsV5sKXIA
z@f&@oVCS~SXieA0`o~3$+Z}ppCxfq5J>Pj&r_|II{8k_z(l2wqX9XjIKx#pedd}Gf
zvCWxiX-{^+us*!Gk3&pBwVjFKko6+4<NT)>EPwr7O(S{d!6VcVNwJQklTrqHHTAa^
zcqio1Tj|Tf-8bn6&bzH0me`<N*Q)8R2Tq@stVxQ_>3F6mICc!33L_3sL)oW=sKK_g
z3(xF}b}$!T3HY7kdZi9E5ea}F1hZhhL+wt#cCpG?>rbCK!kWFQ`;bpmfo@<&`(cE?
z_@zje4CW5l2Kw-on*ncpmk_kS>vADc*1GRgy1)d=((UFO3C|A$4%SAxkn{jCg1qj+
z!?D~a_~&NJbd-CIb{=QH7k3S3+We=URb3T#1Pc^~hhIL4<mX6BF9N!QK_=nfK4<t<
z7mXS^-D%w5u3Nnss&%CUX!-s&u(&cM9iB#GMwz4~&J-D3T>Um~<t3)JvPTIQ#lUH0
zUnoN$XTa4r$=BOYYYbryY&21+ZBO)5v}HnA7mv*m$P!ma@zqM!0USulF=zaR6P7(3
z{o5+}|3*wP@WJXpu0&7tk233?=O-F43q;3nZ896pjw8*rUjlCu(!DLuW~nHKvGO2=
zNc$+C8J?hR5eq!Mi1-HD<ZtW=l0O|iYFHnI;N1}R^k<1~Ryjepj~c`pWa2^8X#8Rn
zql6SVRS_#RPtvyardI?Yn!=E0T&Q=6l8X8~{ii2slA|EW6Mf~Aco#QF69D}&Ndf@B
zq^*k=16jHn^yI64&}!3+L(O7izqck8Q>9Z@npzJxC0GTjy>c;kjBkLd#YvYGc~m>&
zz(uQ;yOQm=ev4mT;xG(pUQ)1|Tq4a$7Lq1$`dz)WV0rb|FP@#TJ-40XZa2+S;jRv%
zC+p%g(?Og=1XP}_?Q<k4hVir8gUTOf3!kT(*D83Xn~x1i1qBA|nx6U?b!duupKZv9
ziZ9~?BCox|x&DcAJz3sKf-#`e2-N+;j)f54{ODVs^ufH4w_8yu2YSYS;wPN_bXJu}
zz3?)(II*eAm%e(|9R2La)Xg%apRe+|pQyH)lf=Iq+RoxTdrLq(npAI>&T5|&AWRi=
z&9@Di4703FJu%xo`8>pLzdlQT%I8@pPr2QowZYjw=a|4*z7#Od4QAjr9O*L}8dz+c
z{XGS<Cxpt!(Ksh9v#g!AC&$RpQc)Sw5y~j{!UMV(0}R}M5f!3|qz>GCs$MuC$0t+X
zY?hx16R3?WjxGr)a{06OGB!(Se0V}w12t>)kt?;kL&zU|t=g20<)I?$nF1kU@(Tln
z6k|@2mu;O#4#m-Lla_oggj*>?b%%J}A4FyhSn^3whJC8@*|VhfH!I_Js`7di@zn0J
zLOiSp#__BExLUOwAbOqRN4R9<<cM<{x5A*oM8c>-;%L@}{G&Sv`CK$B)#7K3bN$hE
z6CLqqFtv7}BSoVCG>BuD*IRCrS$~QZ#2B!On|nl-)r#qBRM2{oO&O=TO6%{&*RA>3
z%s0mOtVp<QBj?j-c>G)kER6!%zLR{ecp>k4aVGq#>Y#8u&4)ufasJE;J`xanuIF<Z
zPcg_($9C>1;nm>gLdQm8Cs>6_=Xs@N`$<{gMeDp?CIKTB@lv;9^iOBQp-Vw@E1B<p
z<yFS4oZIcl=YH!HQOM@-F+pLhazVA}M5#aXaMPN|9=&cz(BX=w;x|1}HRb0=Znk#T
zq3)++qsnMSw{P-nR?s2zJJR^V2VBtTYW9l`@!hO54b7G20*Nrpv~K&2Qo*1DT_L#`
zs*&?p@w_%&;{;^SGwR|B%M`ZiC-GZ8YY$*Q>&pQ*txySdr9Aveu-VEZ%MdWDPhD71
zq|RzRua@2UK~b|vdE!+U#@ZwA!U;9Fl*fw3j|sBG!@DFiPVc4MU5JYwP5e<f%3o>N
zYHE?J2WN*WZze-63P*=3+QN)9_Z!|rZ7mhke4#uk6bjwfZ>imQ#PB&JRb&<2V|w%M
zb6NaBqQLp;DW8EKB17V<(6{?5i#{EymyW^fI(`ZU_Z_m~>V`eGlM$;FPqJ4{d%cg$
zZYdGG^2-j3qt*I;rX1Ud&hTvN`u^=k;48lqQgTax-=DRH`&sU0aM53xr8WY5@N%s9
z?rZo60TXlN7eL}-JrMWXpRlA};v{ZORmIfGlUhG`iXg%xKooe9b>mJ$>_j}^b(fXq
z@j5GBTHN}ayzaj{SBh5>vO@M#yRYs`)}F821J*aon#JGCsDHs#+}MLJot@21TIg)v
z07=BT$!unm+g!$9wdx}qehwu}lzvMk8%m>PxmtH4>pMh9BFF1S_k<{6;;MD~XNVWk
z%fs`}wl0`}b7XMwdAINbD5>?L@z#mij~z4aQwL$H>UPHoN=D@nMj<1&sj`v#^*QY^
zx6hJ>)_iyraX3Pq^-=dsdpF5lm^$0KM4~VEagnMfJJfCuV!E#!<MA#|-t>*2YkOo%
zC@!Qw`<g*&M1qQBNb>qf_N!L8=CG0i`Vfl8!j&=tn>S7jQkpVdKDvpx?2sv!O`6RO
zbn)B^I3R|uS*&$AbLMZ?_u^9ug)hcRH@4Qbv<cpO8%XNv8rd=BI2GdsfB79<??$qJ
zDGIl!15d)2ydm>k@T-03N_md#!0ts$Xa}A@9>qN+#8rtj@~rH3@<3r66B%oPoJ(0o
zxILrPXt$NqA)}Y3{dKF=hcCC3`x(x5>A{RW_KpkfRNeVf)m#Cciy3*ooLw2QD|>3i
z;NwZBt^IM@{#oUuog_%7z($8aZt}j^y$1VBt;*2wJ(AYzqeHf0?!RlN>*7{Y$1*Rz
z+)||o=p4Xvn-3FfB_B)}>}+h3{7JU?dJfM^5U;bXZlua~mCH&uGP~bvr7p{%hj2FV
zpp}03F~Nhri0-WZjPsEnCuTP({5Ui1=D)q<Ot8@{FNEikf0*5Dp-<V`>bK>nPO5Qy
zmr)^ca`ZCxs=t37ok{!2!umDFTR$tDdAV4K?JftRwi1D2N{Ye$T>$Op^E3{1({ZxV
zQQ7^2;f2tJ(vhum(NM46hA%Y%+<k9E{L5)0klKnJRX8RX;z!y!0RcaKmF~q2()J#b
z#!Fv&5E@~Sb)zd3wkQ8Xk>R`0yN{I-qveiF@im$XP4C}oe@yZ(nK7%9Kpo2jRhS$n
zLTiwSs2+cR|LubV4$q_0Q!#PW`g+u~J?BmuTTlJr-l7~WkHiWQ5oKEA<yG>iGQ8`-
zZ;76@DM54g27I5SyY9piH056-x_8BjN{erMCqxo`TZ(o^nB4$}L~?nWVgKoIq;)mJ
z@MWhF&)e+Ede<=VT%S^-xbxZ@mKG<?KAbmAHr|fAqEGNZ{HN31h`S9Qil5h&>?q<~
z`d_~_@bsYG@~Uhev37qyjEjfY!qMz*m8Y*^A_-~?+B2K8v$K<zTUu_-ax~grbq};@
z&bi-nwJz@yOd={9^r7N}^7ml4*Y75prV^@$GhJopH=3#8TL!+f5y@n#c|vb-G>uE!
zqJ1Vzd<7#~9?k4=1*Kqn5_%enC$PLFO8qmf=5}z=1m$q+*5>&2rb{OIs)&f(Bayve
zyd%$wO?O^feqZl^XX-6;hMONNI6BF@U0@!ycNtYe?pA*ek%G}?ZxdVIQfj+r`Snu>
zFOr9?%rnv5VI{D{He6Pqj?mXjFoc?^E)H=r0lViV68~owp5??!NgbYtW)GbWhMN4o
znck4C%9Zs(Yfaj+&M&;>66Y4HH`Hv6S?s2<kSC@>LXd>R7n6RV2J1C1JcSL{-E9-;
z+7Y*L31}*3;jg^v<PaiTNCxL}Aa?gp)0w!u&=EFchI#QtIl{C6f9t*u&RzW*j|sSj
z1q!Z<op8PIiS@G9sOP*!QqAbT@<i!EATnywzP;;KmaKMN>AlXaF$yZUdp9FXmA#TF
z_owpeE`;U>yJEj3GI$q6Y4d6^pXpR@{A|+Uw7yr)cJpRr88zEdr0)9mb6((?ZWw<F
z4tb#L%lARu3lm_C%NJzu{q20-5q%(oA3l4&aWMOC%hmy+kb}3aZG3LIRG&1U8UNI7
zlXEIRMF%|{S2U2L>-6*NmERIMywHZjVQV)zpP|tOq(tTTyqm@B#Vd?vxe5Kb(ubsc
zajBFex~y^wz0Rl-ZdyMG-R8B-=i-_>Bj~}aHXU_{9qG_vn<{pF-KWqy%iaAv)WC5}
zP!{2#+PfJ!p1_&gX4G-pc=UvPs_}M<NETPNs`@p#6yQM!I~nchexU2w)5I?mwQ2jE
z9X>zq-n9H{bgVxOqrJppp2&0uqjlcjaQxs}>9$D`_}j4cI<ah8Iqv7=&A}e}wt27o
z$n5gp`03Q+xUrOmo<Rm#F8&=RY@6E;<E3>UgJLHo%r@5~M%6FCjK%GoEwcAYW>*Y%
zwdO<YO9f&XI6#Y4$~tD84KyH;mRGx+NjK{GsQl)gFFz^i?w+2xy*K1AU=aV3y3cR<
zZSZ&gtNu&YH}2%wiOas7yG#0Q-<Z$(&08J6bX3PFv8(Ii;sf7^hpactC;E=fdVM1n
zljuXL%!#Q*pJv02d`_~Iiiq%PBR00u&BCJnhHeI0i8a`th7~Cv%s055H?H3mvcy^P
z^Z!~AXva+{&ZI68fVq;*d45~WYvb2mb+WI=`OLbytt~5zc4OdYILV&ZSI@E&yftc4
zkwe2t225-R7m?`%$2usJtAszedV9Rw;I`QdQU?3P4IY0EinH@}JGtid)G7aJPCXvo
zoiVY<hvvMR41A6o^500l=b$#T62vbb9e)WPWF0yfcz)JYd_at!KJe`)&(izz1z%A?
z1VX)+Pa3ws*ft}ScFU#I9#np!oI|6E+IK03glK;3RYv$c-6dME;^k{n;G&qvh?>d8
zjEGS!=7Dm$o9;V#R=O-|<Tzg4qxE*W?Q}O)2nh&sd}h<@as6q2G%eA4d9^}O+>!6D
zw2*?jzQv&?De&hm;Bko+Rt8BRv#{Q=f~xMsC@!T{mY6)bh}(|8b6xhNJ}|Prd}^IK
z(s=|u@ATZf7Jr|(E4LL{$0WfJy|+4qJ@W{pyAU%qxR?vfjS8LiE*4Iql;Bv&9KW9q
zp<!VN>S*WBFGcnUe9>0&EjeDyF`H0EZy#}|4pC*4eB?V8sw8<G9g0UGl7*Zx6X_!m
zOrR{g=tU~(?0D~r-=p5@h2WVppTi8eDZ%Y&dqEt3#h>z@F3;UidvYptW8=~_`p_{Q
zPNTjig@mKsouCDS&XDz4-;~SiVb{!QMCH{oTf{~Eh4<hAa;SHqQCo-r5)=Pz7S&E}
zim1Cp+7f1EPfx16e@vX&xbzSB9~%)xQwofV6+8r%WiKx5Gq}ofwL?gYV>|qgj7t_K
zAVK5+-a9lDs|pXwM$ZSNfTZ<?MN(!tq^!(!Yfc0dDhFq0ThSYT@j#>7vkN496ciO7
zf-E@786ppb=7UVvc-3a=E2BTr?}l-KViId;cz7~Qjo|hjBD~nLcu?PEl~O+@5z;Vi
z&u-B07QDekN=5Yp6gDY>)CB834f)MiOugSjJF0Dm=6y+h8`<E$F{SS?jZ0`J|9jus
zNUpNT44tOuYC_-h6WQu|IsG8t?7eXRC67cAS1bEDv$Q+W;C=dZe7(2neP7FM)tmm)
zvJ~hal_3m5QP;VO)hDV&>GEcrf7y@V>}dZ8akn2$NI{(j-z%G7>*SL4P86IM|ML0T
zRorT2Ji1N0`WLdTJo{B%esK+@DZJKAXMqq&6o%qe`hE+tBlO1Orr510g4Lny*gR(J
zU_cf@r{dY?ED68A2l8ebU$E8lX^*g@nI_?^p%3IIwc*7nsr~Ww@OHhQg9g)N_XN9<
z+iDYE9%~%7dFs~8R>};{Qtx|>xMo#G=-j+X{`&)oy2!Mam`=Bj^RtcS4dFd}w7q`}
z%!-7}=JenVE#b_-o2ZVWzdBPzHw>!z>_I&$Ih)RN-c$ktf`vCsS@%Un(?GeM&UXpc
zhj^-aWjD?*YmtP`8a(=8?C`Y3^4-$m?897J7Q6GnjLVNgX7-@X`nwta6d1)Y%cSRO
z9$ioy8QJ9OiEk-mS9hzgeqykrpv(IrFX@$$fY{-$-2A+v89G}?H2*akBk-}D5OwzP
zebA=Cr{ajt+_CFImvi!Xt%CUL&&tNi;hXHLuOA#};0gT>2sA`ML3sB=h~k`2nFOhW
z^7bE=Zl9Y$vAZ)!X}s*d3Cb+l_3J=6CxL7<Wj81SdX-1U#Xw6-I|SZhGwO)o<5?C9
zJQD_`hFrpAGM}x-18&SSFpb@3xE$^t@p*Bkd7v|C4*u+$etv(ae}Z~BnuvjE8UD@M
z8HJv*W;YZ-neBhJA7vF7Cu$q1MBmIBN<>fd-hE5t<jV!=xlmzk3Z&<YOVDExpYkOz
z(<jDxzG9MIUB>kw^u|`<I)S^bt?kDe`jC(rXG;|oVhjeOlqH@DYGpo{_mXaSie0<S
zNzcj}$zvAjcqON-tc$$Y0+wH~e(6$J`u(&0G1VX0Ql0V!w6w6|xYp&_XPY}6#KZTU
zo_q8%-qCm*u)B3Dl5yl#1O+1;A_3~&$Z^TE{Um*a$4m4jcXaDIP+?@yyhfG5Z{}?+
zsM7#9J5&BU>RMuvWzYHVqz<prRWp||^+n{d`e$xmhD776lVzB=E*n6?=iRxK%WVvP
z48ds!+$t5}ea!`bOebY)dOqpu+R8oq3}Kg-`}^)EXjJUy=by}Zc|c+s6e``3@U0um
z?<|CMrLG@?H$e^QbF=uYhy~LJO7Dn=<bAqrkdr1grzOY6G$ldQaDM5H&G7Ie;$mj9
zY0<;J&lfMIv{5F~-Xa_$9s!RVBbSDTlg?Sn9y{nsdmEMQt+WGIT%ou)^sXljjsK~N
zwJm7j+RK3nYij>nctk|j?zin!9apU_9}c{t6osV#kc1u+2m$qsg5U)?nL$u@W5#`F
z8;u6<>4dqh4(@<j&fA7Vs*Y~1uJVeCFGdR#Ak2|3%gV$-jvf*k5wSN4<x292t*J7p
z0`Gd(S&!r`cz`N|E(CoR@{@u?7O(WG3l^DmI(dDrCB_7;Z%|olhe?qx^fqtLkRHwZ
zZABY#Sqy|6?u}rUmVa`RXK1ST*6TsIjXuhFW2mgjlbQ-L-x|*)-L<1J+bj6t`r6o=
zx3%o~kiol|Tphcj(C7lc%tOeAh+l9;+fd?sJSPOV*Jc0X!4G?(Z-_q&l3jJx9b`X6
zJU8xmwo+@tV-%TGv*_8`iC%q{1}?lrx0(hdcU?aFk>0wMD(b-s><>8F8s7t32r-rL
zC-92)dr*H2MKiDaTGYJoTijzSH~0UDn}yKzAFQo)%eju%w8^;X(;!nmYPws6SR_so
zz!$AB)m6z<eXcpT`nGA09CFCCC9O`5^J8LL=yqm*ycGUSaEJlY3-7)d+G!(lv#dCb
z0F~IeJrozaq)Q>*>92Kn;asZox$o?VjAVM@0Ym7GFX!&cF5s`V*i2PXgOXf1ZEa4&
zCTT4#@FKuX2<R`!7k2?~sivK@v0+ca{_GPdm5w^l9(9kY*;1c8=zBOtxb9_h6*YML
z!RyiDNX;d4y@lakyVjnSUeBww&)Zby(T@qTEjV3&Jmbq-IHB`uyYlHJM?&+(g`@fR
zmugvhc@sBi$CfHJPCLXevD)-Sl(Z=^oVi+lTqO-~mtcH#v)@ciO>T=tcxKZ6QISAY
z03sp-3Ea}t<Kgq-t?^RLX6Hp5kZm;sNl)>EaZOO9MG8si{R%?&bi1DrDCDpJuQJVn
zdOu)QDI?{3Xy@O)79Yfn)3eW{t+-8ihDGXH51Ww1cAcuQwnrbn*4d%@Z!wp(KRdTP
zwV->+fvj8V9NH+hTK~K}!(L!4tS~FP_~ULv@`KDQuG$OHiF=pdCA!qjb>J-;HDth}
z^H{~<ZBl`ZZgF;I1|+!mSp(~7a<4uJ{cffP)#qXA8!v?324!PVL-%JRpg?qOrc~r2
z(^&<jrYPX8<AD+PZ}~~J7fJU2Oj!7^`}*2*Ppa)M>$0W4vHhrQ9~*T<TzlFpP;2L~
z^s9*$33KlyPSL3OEivui2UFn8^F?bBV8h*OSHA=I_mihiJ-SeM-f!MP<7MT*1qqW)
zcZ<E2b(@)z)n4x9WxuwZW9g?U@xNLH4n=!1#dfx|v@pz@xL}$~=)<5sS$li?*vB0F
z{L4@F>YmD5vH6VmOP1>^TuK?&D~naHyQers<MAcdrY5bEdG1ae3z<C^XbS?hynA=8
zF`jSg-7x*z-=tq$*f;l`wFPh`oak=D=~vFU#6H&O?K<`SWA{t*4gL2Pa){gh>^~k^
z^~ZYe;Q|e_T>Gn8q07^JSKa~^-QmCy9ChF_`-F`VbF!*GJUCd?w=Ctec+IihtFt~l
z&S!fg7qaNqI)^iH%%87ceX#c@H?R@A`R1AG&(Dntzy3X<dc5jRNM~_m`V8UEJf_m&
z3xWG*fkQYkUyjX*JmdF6zq@ap>TzecMWUe(cQ2V=`DJ=$&4>A{Ko!6am#>(hRU&X3
zIFFpo4B&nb!_rq*nm+f53Fh6~GZVP|+!(lWJFfn3sqLYsz@6VdGk^Fe)!%$8_}Mb(
z(m%F@56-m}>F3&ktFwKccX0jv@wne-_Aj>f{v*{sci&Ihe9|QU&ZLR{wg-=d#~r(V
zJ^cNIBcUFCe$V3d{S!Y=JF8*!-)#1riA)C??)|M$E)oN7N8WV|=%#yqz|dH-e7Ufc
zlviG!9&l6imG$xR8uJ`(=a>QKK7h*yfP3xVECB8pu4)7J89`$$jXn<xoYqEdoncw5
z2HbRbw9INx3@}3jcTT&wxOjZh0jUb~Jjt<Ta*mlaaE;YN;1YXK8UgOi7nf{2cjAPC
ziHS(}(L=z+8r*IxTR~>JJXKP@;P>g{$HaSkDjD|tc+{=ob<wi-@^XK5W#wkg;AJpl
zJAuY_v+>EC0J^QSuP^Oj6YG&<$ArYi-GQqfGA=A&1TDTTZMOx+Gtd$HQdn-ePMkRL
z$FE<XfXjgsw%=xyu&=Wz+dUV!rPK@R(vzM{Zx;b~X&VEVO#98ZI}2Q;!qCy#xe&No
zf=9~4<M|7)uU#r9Ys!Qx8XHdr?xdSAZCcwR4PbK?xTMV;xTokVpD(Z>Bj7Z}Sh8^+
z7jO~FnOUZ*6BHj__P5tPxf$faBc)Dm3)W9sqyg+MOq?;}1aK^j1GtUGS_R@zMwP~W
zY`}#=5jtX7!1M0PHGo=L1Qs=UfcohmHg^giaIg&|s5b#TC<5e&BV<9NgdoX7tdM56
z6URd)&jsM1Z7FC350L@Q+2H^jN(Gx2qW~Ja!)e?<d8>&oJOBFR9Rn?K@O1TaS?83{
F1OTxw0?q&c

diff --git a/docs/authors.html b/docs/authors.html
index d07acc0cb..8b101ccfc 100644
--- a/docs/authors.html
+++ b/docs/authors.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -94,13 +94,13 @@ <h2 id="citation">Citation</h2>
 
       <p>Petr M (2024).
 <em>slendr: A Simulation Framework for Spatiotemporal Population Genetics</em>.
-R package version 0.9.0, <a href="https://github.com/bodkan/slendr" class="external-link">https://github.com/bodkan/slendr</a>. 
+R package version 0.9.1, <a href="https://github.com/bodkan/slendr" class="external-link">https://github.com/bodkan/slendr</a>. 
 </p>
       <pre>@Manual{,
   title = {slendr: A Simulation Framework for Spatiotemporal Population Genetics},
   author = {Martin Petr},
   year = {2024},
-  note = {R package version 0.9.0},
+  note = {R package version 0.9.1},
   url = {https://github.com/bodkan/slendr},
 }</pre>
     </div>
diff --git a/docs/index.html b/docs/index.html
index 177934cc0..942d6d6e8 100644
--- a/docs/index.html
+++ b/docs/index.html
@@ -67,7 +67,7 @@
     
     <a class="navbar-brand me-2" href="index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/news/index.html b/docs/news/index.html
index 5f2ab3bc3..5bead4dd5 100644
--- a/docs/news/index.html
+++ b/docs/news/index.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0.9000</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -81,11 +81,11 @@
     </div>
 
     <div class="section level2">
-<h2 class="pkg-version" data-toc-text="development version" id="slendr-development-version">slendr (development version)<a class="anchor" aria-label="anchor" href="#slendr-development-version"></a></h2>
-<ul><li><p>A new helper function <code><a href="../reference/get_env.html">get_env()</a></code> now returns the name of the built-in <em>slendr</em> Python environment (without activating it). (<a href="https://github.com/bodkan/slendr/issues/162" class="external-link">#162</a>ccc)</p></li>
-<li><p><code><a href="../reference/clear_env.html">clear_env()</a></code> now has a new argument <code>all = (TRUE|FALSE)</code> which allows deleting all <em>slendr</em> Python environments. Previously, the function always removed only the recent environment, which lead to the accumulation of potentially large number of <em>slendr</em> environments over time. (<a href="https://github.com/bodkan/slendr/issues/8707" class="external-link">#8707</a>b9)</p></li>
-<li><p><code><a href="../reference/plot_model.html">plot_model()</a></code> has a new optional argument <code>samples =</code> which will – when set to a result of a sampling schedule created by <code><a href="../reference/schedule_sampling.html">schedule_sampling()</a></code> – visualize the counts of samples to be recorded at each given time-point. (#d72ac5)</p></li>
-<li><p>The <em>msprime</em> dependency of <em>slendr</em> has been updated to version 1.3.1. As a result, <code><a href="../reference/setup_env.html">setup_env()</a></code> will have to be re-run to update the internal <em>slendr</em> Python environment. (#dcb83d)</p></li>
+<h2 class="pkg-version" data-toc-text="0.9.1" id="slendr-091">slendr 0.9.1<a class="anchor" aria-label="anchor" href="#slendr-091"></a></h2>
+<ul><li><p>A new helper function <code><a href="../reference/get_env.html">get_env()</a></code> now returns the name of the built-in <em>slendr</em> Python environment (without activating it). (<a href="https://github.com/bodkan/slendr/commit/162ccc" class="external-link">#162ccc</a>)</p></li>
+<li><p><code><a href="../reference/clear_env.html">clear_env()</a></code> now has a new argument <code>all = (TRUE|FALSE)</code> which allows deleting all <em>slendr</em> Python environments. Previously, the function always removed only the recent environment, which lead to the accumulation of potentially large number of <em>slendr</em> environments over time. (<a href="https://github.com/bodkan/slendr/commit/8707b9" class="external-link">#8707b9</a>)</p></li>
+<li><p><code><a href="../reference/plot_model.html">plot_model()</a></code> has a new optional argument <code>samples =</code> which will – when set to a result of a sampling schedule created by <code><a href="../reference/schedule_sampling.html">schedule_sampling()</a></code> – visualize the counts of samples to be recorded at each given time-point. (<a href="https://github.com/bodkan/slendr/commit/d72ac5" class="external-link">#d72ac5</a>)</p></li>
+<li><p>The <em>msprime</em> dependency of <em>slendr</em> has been updated to version 1.3.1. As a result, <code><a href="../reference/setup_env.html">setup_env()</a></code> will have to be re-run to update the internal <em>slendr</em> Python environment. (<a href="https://github.com/bodkan/slendr/commit/dcb83d" class="external-link">#dcb83d</a>)</p></li>
 </ul></div>
     <div class="section level2">
 <h2 class="pkg-version" data-toc-text="0.9.0" id="slendr-090">slendr 0.9.0<a class="anchor" aria-label="anchor" href="#slendr-090"></a></h2><p class="text-muted">CRAN release: 2024-02-08</p>
diff --git a/docs/pkgdown.yml b/docs/pkgdown.yml
index e269d68ec..80afc3cbe 100644
--- a/docs/pkgdown.yml
+++ b/docs/pkgdown.yml
@@ -13,7 +13,7 @@ articles:
   vignette-08-nonslendr-tskit: vignette-08-nonslendr-tskit.html
   vignette-09-paper: vignette-09-paper.html
   vignette-10-tracts: vignette-10-tracts.html
-last_built: 2024-02-07T09:51Z
+last_built: 2024-02-21T11:45Z
 urls:
   reference: https://www.slendr.net/reference
   article: https://www.slendr.net/articles
diff --git a/docs/reference/animate_model.html b/docs/reference/animate_model.html
index df71ff5cc..0a28c844e 100644
--- a/docs/reference/animate_model.html
+++ b/docs/reference/animate_model.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/area.html b/docs/reference/area.html
index e168b8d0f..6a32bd607 100644
--- a/docs/reference/area.html
+++ b/docs/reference/area.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/as.phylo.slendr_phylo.html b/docs/reference/as.phylo.slendr_phylo.html
index 18000373b..c9aadcccd 100644
--- a/docs/reference/as.phylo.slendr_phylo.html
+++ b/docs/reference/as.phylo.slendr_phylo.html
@@ -16,7 +16,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/check_dependencies.html b/docs/reference/check_dependencies.html
index 2a6c509c3..bb7e05f3e 100644
--- a/docs/reference/check_dependencies.html
+++ b/docs/reference/check_dependencies.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/check_env.html b/docs/reference/check_env.html
index 8ae01746f..8ba402ce6 100644
--- a/docs/reference/check_env.html
+++ b/docs/reference/check_env.html
@@ -14,7 +14,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -122,12 +122,12 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-in"><span><span class="fu">check_env</span><span class="op">(</span><span class="op">)</span></span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> Summary of the currently active Python environment:</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> Python binary: /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.0_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> Python version: 3.12.1 | packaged by conda-forge | (main, Dec 23 2023, 08:01:35) [Clang 16.0.6 ] </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> Python binary: /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.1_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> Python version: 3.12.2 | packaged by conda-forge | (main, Feb 16 2024, 20:54:21) [Clang 16.0.6 ] </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> slendr requirements:</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>  - tskit: version 0.5.6 ✓ </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span>  - msprime: version 1.3.0 ✓ </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span>  - msprime: version 1.3.1 ✓ </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>  - pyslim: version 1.0.4 ✓ </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>  - tspop: present ✓ </span>
 </code></pre></div>
diff --git a/docs/reference/clear_env.html b/docs/reference/clear_env.html
index 05655cd61..6bedadffd 100644
--- a/docs/reference/clear_env.html
+++ b/docs/reference/clear_env.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0.9000</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/compile_model.html b/docs/reference/compile_model.html
index 3d633ed84..771735dcd 100644
--- a/docs/reference/compile_model.html
+++ b/docs/reference/compile_model.html
@@ -20,7 +20,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/distance.html b/docs/reference/distance.html
index 7f8c038cd..e226a9f3d 100644
--- a/docs/reference/distance.html
+++ b/docs/reference/distance.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/expand_range.html b/docs/reference/expand_range.html
index 52a60badf..a1aff6668 100644
--- a/docs/reference/expand_range.html
+++ b/docs/reference/expand_range.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/explore_model.html b/docs/reference/explore_model.html
index f89ee7f92..ffd7d1d6f 100644
--- a/docs/reference/explore_model.html
+++ b/docs/reference/explore_model.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/gene_flow.html b/docs/reference/gene_flow.html
index a5560231b..6a03c7b52 100644
--- a/docs/reference/gene_flow.html
+++ b/docs/reference/gene_flow.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/get_env.html b/docs/reference/get_env.html
index 35acaab9e..44c0442ce 100644
--- a/docs/reference/get_env.html
+++ b/docs/reference/get_env.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0.9000</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/index.html b/docs/reference/index.html
index ab52c4da9..772317a7e 100644
--- a/docs/reference/index.html
+++ b/docs/reference/index.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0.9000</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/init_env.html b/docs/reference/init_env.html
index ad156fe78..8d40555ed 100644
--- a/docs/reference/init_env.html
+++ b/docs/reference/init_env.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/join.html b/docs/reference/join.html
index 609058a4c..4d362152f 100644
--- a/docs/reference/join.html
+++ b/docs/reference/join.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -133,7 +133,7 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-in"><span></span></span>
 <span class="r-in"><span><span class="va">real_map</span> <span class="op">&lt;-</span> <span class="fu"><a href="world.html">world</a></span><span class="op">(</span>xrange <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="op">-</span><span class="fl">15</span>, <span class="fl">40</span><span class="op">)</span>, yrange <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">30</span>, <span class="fl">60</span><span class="op">)</span>, crs <span class="op">=</span> <span class="st">"EPSG:3035"</span><span class="op">)</span></span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> Reading layer `ne_110m_land' from data source </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpFhhY5X/naturalearth/ne_110m_land.shp' </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpqG5m0G/naturalearth/ne_110m_land.shp' </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>   using driver `ESRI Shapefile'</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> Simple feature collection with 127 features and 3 fields</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> Geometry type: POLYGON</span>
diff --git a/docs/reference/move.html b/docs/reference/move.html
index 085f5340f..270da5d09 100644
--- a/docs/reference/move.html
+++ b/docs/reference/move.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/msprime.html b/docs/reference/msprime.html
index 5ae11d4ff..3b3ec3284 100644
--- a/docs/reference/msprime.html
+++ b/docs/reference/msprime.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/overlap.html b/docs/reference/overlap.html
index d5858687a..f250074b8 100644
--- a/docs/reference/overlap.html
+++ b/docs/reference/overlap.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/pipe.html b/docs/reference/pipe.html
index 79841bdc6..4a7fd8b61 100644
--- a/docs/reference/pipe.html
+++ b/docs/reference/pipe.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/plot_map.html b/docs/reference/plot_map.html
index 91811bafb..756b6881b 100644
--- a/docs/reference/plot_map.html
+++ b/docs/reference/plot_map.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/plot_model.html b/docs/reference/plot_model.html
index fa109e41d..cc3b4716c 100644
--- a/docs/reference/plot_model.html
+++ b/docs/reference/plot_model.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0.9000</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/population.html b/docs/reference/population.html
index 578d70ade..400ce2c9c 100644
--- a/docs/reference/population.html
+++ b/docs/reference/population.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/print.slendr_pop.html b/docs/reference/print.slendr_pop.html
index 531dc026c..3e07244f0 100644
--- a/docs/reference/print.slendr_pop.html
+++ b/docs/reference/print.slendr_pop.html
@@ -24,7 +24,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/print.slendr_ts.html b/docs/reference/print.slendr_ts.html
index 660995327..c91c7334e 100644
--- a/docs/reference/print.slendr_ts.html
+++ b/docs/reference/print.slendr_ts.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/read_model-1.png b/docs/reference/read_model-1.png
index 9b9b5caa5d5067193ae3d6440146dd8cb8b81ea7..1881807ffa06774c9d36aa9ea39df0e45db56d03 100644
GIT binary patch
literal 31200
zcmeFZcU03^*EXCPZ^vgu<(W}o6p?WZqS6!vq{cFihN|?Ag(A|F7O8P`kWmyc^e#$~
zF5S>#0}L%l3q2?$^w2{HkawRr_uTWWZ>_KV@vim$@v)XmjK6x$K6_vLx~`Llmvk;{
z-@11z27}p-zxb;j2DAAN2D9nfe|-m^^hTx3!e8HA)4cF2{0%>xa}`AR>-$?5P2DjV
z(IE8ao6wOtTMXuB4F1<M26qys`Ux?XhA9fO>_#mT{y^&K)SyeL8m{+#J!p7VKm&9B
z(pyQ3|2{u#hFN;l`XSU8cj@xUlbg>9UR?j>O4y;A)2W2B0^fgsKuy1yT^ELr{{t^)
zuhZPfVyj+X?)IIo8VFyxG1?U8xwYn6!B6~0I;vwx-o*dymEW#P{x7>V-U^^!E~Wf$
z`QT=~=-p*S=bmB<-{}%Osf2E&!kwvbjotYTx@p6UOHx9?BRkG;eWpw3ayMF@VX5le
zsc;omWl`fpbj00}tHEe^F?YA74Q@uiyu7s4>i?t<l&i+uGMH-%G+IcnYx3EYOFvj=
zFHJ_@-=TVp(0BX7qnc}EqQH^xIJYhQr*_+`=wh)$dGK!KMy&|v-Glq{OSz>V({$Am
z+r0XZWG{_b&D@c`DJ8Tp^LP(=C|S^+V(!{Bdrq*)HrhIArr)@^G9u)jFy`6r+EB4%
zE{lS*PxLejSbQOtW?tqp64IW&HSp|r+kOh0F71_3^IeuA?8DskS{RLYwF^CPt~EbQ
zb$LnFexYMk93N1|Vm2h?cm}B7p<5`;bSK<t(^kqVI8L`BSi(if*FS}zkkFhPye~M~
zuCGhIFR0$<H}mmesMPi6trG)P3XGN7<4OySC)&zf8E$poU=EX6iGFT!_Fgl}6R&bU
zWzTP<RD=z`)OC)U$T}NB{`}zaUiqV2#QbidYi~U-;^Cp3U$t!Z-S)i^*lzn|e>Y|#
zAxh~7OiseaN*!+0oUt03Q|)5trTpx`q!Y>S*m2H>Hz(J%%QqcSIMg<vo8*i2o?6bz
zDH&BJvy&+H>q~T5!b!|sk~h1bP}FJWIMR0@nTl6&(w{jcnHpom##tw;$2%wS6n#9%
zLoFdW5KuSXuCp1*uKts_!jZvi4>s)2W#LxM^?4-?@!1hZ<Iy=Kw+olTYAu!+uFSie
z*tDTgwWoQO6;UdCF{kmWi!JAl2n8!$d2wbWr10HV`~#O$$Cw*KV(?+9!`QD`^`aQe
z6<d#ac6IzWo3aae-#f?<RDw4gC4&4Y>b(7I51FdPR{qGv5*Yr9_`>C*RrLv}&N)RI
z`He{@B-%1`WeK}_p2CD*`9ZdN<=TGY)I-13!3PG0(JI>BIqOuL79&H>LUO(oVHdo3
z_QcQGmf1KTu_sTS*y^11)tmhmBlU$Uk*r0^sW6SbAJMHGuP{o@NRlFKgC{$J<~KOz
z`1RjNspg?_TaFa<BvBjpoM;e*Bw_MMr}V`>%;`zbu;g}gm&En;O&F=Bga;4$*4{Oc
zM`M`_AA4DLI{xPrX6O>h=B|l;OZxG)!5yozUbNQ_!wV-z4u)el`7D2Is3TU@*t5K9
zKB_Jaq++R-L4pl=lR1Xvif3!|;Q=w2H*!BI-qu?d@peDdoKVujDcE_)Zt9PRoa76<
z$omQd$Bg&A)-5-bT}#>FDK;HyPKoe$n-mFd8hYfMkQh=Yv>h$177baQjh|Jly_(LB
z%q%-@`*=pgHKxR8>885kL``%NyF)R(;N|nQV*aE%)(jIVv5hgxy!K3@PPvDp_e$y<
zRjis`et~UyP-){``jA6-xB#VOxFGOsYXwp6R)Y<5Q+Nm6xqlx9Q=F4x()dcm#aYG~
z*PZV=qwHtWIn`0*-KdryN((sXszXwpU#{4*^hI*dGIe_cDZ2SBq^!FN3gb=b^CbmW
zH|UPK(%yc@1wt)*jBo#*JS-^`tzMd<9xHcX54UWX&A=yBQ-)ndG0)EVuH+KtTO$PH
ztV7kmgay|3V>xSYJl$7aR1_)d?34?=(qf6>yI~fcr>I;Z9GlildO~nw9~K~w-?M4y
z#D@-yG<}1)l$qzV${8K{9Pcj&Gjij0Um2n6kO+ISZLK`!1Cy?Il>dANPFnjbu5V19
zHEbgrCu`WzF5dK&;S8iYhy;mQR81`iWs%zQtW_KSpp!I-UBz*V_$2?s-2utAM0E^?
zdd{_~->;JK;C|F8mkt+8_b$6+i;1E`jWMx?%{=X3`b5<HH%_k(o1cUMOewS0n6Oh6
zb$#tz*u-R7`;N~Usa{cv_S$9kI{#8+PQjVN_38zh_Fjq~Z42gSZ@bpv(V-<Dwpsnh
z&P+!qr#XSBK>k~gS@T>s=9anGexF0Uz5R=;DVM2H_j_^r+H_7`_70O$-{6{Bi(Bm9
zoJIZ%^U@@~x=QCe?bTn&*NbBbVm{tvR&Jh|H(k5a(*68IIYE~2SV8~9;^3~;e50A4
zB&9dYVor-VywpvqwBMLq!bKX$&X(nVpMR(gchP6bbZx+U!cM|ov}eAg+7Vl4-L<M(
zb8UVjp8BGUhAW#}&ZXcSdSqSZACK^Aw_wIheEeS7eWyKjYBTF*+kkoc-2K&@IAdH)
zT&hEG;O+vS(3w5NF9p(8y>qSECtSXyjs$oO73pRanA1bWOpd5i2;`kP?XX}^<9F+8
z6_}Xf>`E)Wt%>EC2HK=E=Hiw(vy%nmGE{d#MNu64QsH~ysEvf$2jatK8acHO0)sza
zyFE8oAr`j?gYCWv^U5aK%C~9s9(#aZ$9(fd5n`ejm9e1YizJWSyn3Z`>P1~KSy9Vf
zH*4fgA|L@B&g@NI)!855y(cO$?_Wti*Elo2PQOPM_Dav5vFn{WP8sxHi3*jy6HO;Q
z`i$*fU7*^B9&UorOsx4FPn0uOTelz$3-b0xCr@vT=k&bO=yn(Lk)rVEud!5mS<iFJ
z6Rn)XcfP^&yy@$v3WPQjtKN1lgwlKcHs-eMBpumFs@7bg*2?sg#4P%kIZ2fg;}oO9
z$pg+jC(PaEYaicuE<8+9le=2mcyd>8H#1ITfqqV~{-)X}i{RTPAgdT%fuBzlEjx0-
zK=}v$yx#GCYDi&<MM{lJOeS$I6!r>7&psPjbErIP=+?6&kZ@QbvYGc<-Fr4YU7x+C
z)IIVnHOq%H`VzbD;}RB^)#Fx&xw|9d0)1m7<dTtPR+iiPfpbONNV+(_RwpLLQ)c!0
zcll*rof91=1S>f@wZv1o-fK{c!nL_<Nn;D;@4r68aA%;kut>c6vm9|j25HLcVpK(=
zGxeISTZ)6=pk+{#?459%+IE{romsWoA5<7I@<#ouLtDPZG^7SBYdhRoA8@NXoGnVX
zILP+n1S_qqnlF6rBjZWOT`NnvUJsS^X4O;?QgV;LxEi%I3Wh5tyRNU8wviqVK{AuH
zO!qEdvRLu-w;X5w3o|z6Woe&V8Sv#enZt>8Utf?NeXGc5ZdY<Ix=m5<wr_c}uc9f6
zV?(JP3BgOGK=Cv=L&M4OUehE<JBjf^#>|b+!^2hd_>kmEiS7l*G+L?2V{vK8?=V+l
zJ@PZDKbSj_4wiPMcCh_!E_KLpDb?!*L;E>BpCx~MqLZH02Z-=!|2ah(Zotzd>h;n)
z9WD(Qu$GtZyiX%943?hms=IyJ`veT;bj4*T8##9~!%0+DBeAMGBU*JiT%{S87av}^
zt~c%`ePdmkDaIf=a~r>hY$J$rPR?CJV@(}Of#-GKMHla>iOJH9JpvQ`6i>G3T(~V!
z0dv~^)(9n8o766^ke*(zeDP>FZxj`$mLtwB_8nO5&8`=YWml`?TXW0H8}T=EQOZ1j
z@??YxJoohHnB4f~LLaZzVL1s05nbJc1bLq&E$1iq%1)qXmHP7evNokT!DwLLh>=tj
zoO0S>x^Q{#xUqRWPL0EIyI^`%tT{3MDXlszMP$cOsQIz{nxCxxa-4ovf*0R9HI6Tv
z8PS1cu56i|y|8XsO4~h@)TAw2X)$xEAr~sr_k@A)ad+mZuZpwr%&FocoegIJ%$qOY
zDXgbj4csZORH~nTVjr%XV?DDAgGpP0#n9C9<hzlNmkGlX*Z0?Ri|is*TID$OKOT0w
z4mXiE$ifR1{<!4r+&r_Hi^WBJ0Bh2{XnD}oK79CXJgJGGomCN`a+xc_bDCeJgs=08
z`tTTxMt1q7j~!e558x_zpDku-5^fs@L+xKw`?qF0M(?T@b@|>ZNlsp@F}7{ZXWGzb
zq*iu+n5@(50M6OV3TRA)rZ^?T^KB$^8Lb^FUV3d0H3lGBMVfUxyMFGzd6C?P!9-Zm
zCnge>$o+0#c9=7xgsrG;;RCeB@>RDwv**<%C41I;M3g1~jZn6484(c4OLpl{H_0tR
zXsGx~qBd!hj+RT+VQ*$66m^oPqIS8L?pDc3y^!6<oZVWx!c=(|=T{Q(td8){IE^g7
zWDLgR7ssWE$f45StiIz{Vo{cgx{^&3&(00+<1&Xs7y!n&W1R=PHw37IpS9}%(bQ{0
zTlH#}qjQAch&}bOU|gxp%XrI~U!C()QXFXg(q6ji$6vY6C;<$ML>*YmRyFQPe@}z>
z8NHc*!TIwqj}j29!C<07Arj!{*;%xT;pZsI#{d5Z|8K^lBSHmpiJY#Roa8%0mWa0W
zGIc2*PhXo2!fDZx)EY#yv8&LkS$cJg_qtR#qo!vFjVu<{Y(q+QwK`>EDc^lmmKeLI
zEnPkt$4qe7>(4gc6CQc7Am%~riSTR3&0Q;;19%k-TWDQm2uIP`n`K`)?cKZAt?P3T
zvsT3A;xg{op5+>wmS?Fg#Y39sVizJCr@E;xc#RZUuD<$-JGWdaH&*xEdh^YQ4r*{c
z+?EG7M^TlS32g>5NgGAO5j+iru2<x~>6V+0;r&}OK9><MLMV>BdM`+G%&#(<F@Up9
zOVGbJ+w&~fZ{>3kXT;QAe9Lyt<*uT>+x0KZ%T6XOhqrr<lZu*^v-=$2=({>PZv~2b
zED8OQrYGzVS?{%)IXqIL!D?q!x#m@UesEGjAxZ_o>XwH`?)+}#yWDNB-~YnCa9MdI
zL@yxTDK0Zyf(eb2Ea68u=-Ezy@wp22%?Spc`Py3p-7XxV^z>k<Ii^MR`yiMxV~v1Y
z8qQJC6?6WDnhyB+5AgsT0cPw?lCJ`1d2Yel+T(U1H9a`AVkT}_P*_ovH`QoU)*m0j
zAmWtPf^(BuL}Iv8_4x++qZ*r)1Jo`0qsiRKM0{CKCr%75dqw>EX4y+-OG^us=8f}u
zP8NPz*Kcpce9@cXz0UvgTv2xL2|XOYQ+PjbWypEe2(74|_<O=PaQT#|K3ux0JHxx~
z3)AIw={k@$P&^Jm@tB@8W@3{Eq{mxryIECiu?9u^>(|{Dwl7#48_Fm8ewnjbT}(C3
zAMureGfv0CkS$*}k8e!sH+X5KD94$9I<3fFpvgBSVpXOSa@2n&Pj{x%1#vM`WVRcw
zyxK%-y|wfKruAK|6uKk8lNR-5&e{3qUMfxuzp)Zz!`nm}$S$Szm4(C+{RZFS^9@IZ
zD-Y&drQ9RDzp!7|%HMtSrn7aWv>Za=b@Dm5hzF1)&-@LHB^bOja%*Aff@a*NKx(Lt
zb*cT-5?S0@WhHF*k8mNJwRPF-Mn$g+Wp??$WvgPek`fjt%bZN%3K*krw^|(ze{k(M
zttr89=BEamYnGNqH%{E^*n$lrijzlWsr`gQ)|{J#18+JR*4C6^$0wRA2WAVd$8R0M
zcDry>FNCHK_2c6Ec<l6{<81leAJA2I|MphERxYht-CTV&->uF>!BzL7kq4#w1tqD^
z;~Y>2Zq2gcjJ0a#LTReDfZyw$-?soDa2;E_$L`5SkRNmUJmm4|>{12J06V;h_*9jv
zHUD8Km%<((O7&V4*@+9v+@kFenEJN7Lt!D5;ho_?9F(V<l`;|W6#-@S2GMsq(L&*Q
zrhytWnwsPrX_s_iztzgp?_6<1L+Q{edyo0Z7aHN$#S4wf^19R5M@K@=NEFq6=&|u4
z>%%oXv|w!4cSAGOd{IYLJ@(}BJ>J?5qvU|$cljbB(Ur!ml7r#0MEN7TG{SAg3p81V
zZtsQSqg1|2Y;@LLYI?6<p9r|(*>$sbdJ1=c?ERuo-Jv2cDja>OJnzp;{qPK4|LUBW
zj<%I=$uvI%Bf@r9p5`R1z8`*NyK<|jO1*4x^4#*p0Kebb<|+j*X_DmWS{rW84cYFY
zO3L0NBk4mbuDT((f+cTwSWJB*Jgwh%+b%v38(5uNTGHg*yb*)*8a7-AjV@h!7yW>S
z+wZwRD!YXaHu)Z6cYbxX08h^Y@WeiZb*)z7Ee+CSyr=S#V~MKvUh{oL&haQuJ^RDy
z6DFcLJzku>aee$c<fe)l2aXwULySn&{Lb(B5<9niVjj9r4Cd93ul`fAe(#K5v6Az4
z<Bug~7Dc-cfnhrHFUio_FBi6GuFNXHDR;e|4Q@7LNm@IvmQ+_)Dj9vJuWm-|%Qp7r
zk4a)2%wtuem&>Z>6s11?y6tS&3+kJt@cnA_iTsT7VOQnpNoTTQ;fxRUPfca^(E<Bp
z{ES%mwtfl%DSu@^D?>xK`6E<!+9Aps%ce8OP2whzrWT&C{NCpNcG7){$qXPz$gRfu
zGnmtX-Co?`I(sZL5vYzoZ73<!+GMSpjtz-9Cm-SW6Hl5FFSV6!V-KLTy^x)%8SR#V
zLQYu;3c}QKUqxnn)%^VMs-p@`W1xhN-xxBd9tb*wxFnPp@7@zO(b+`T)!{j`wuDvU
zj>MW9&-|Kp!q%;DW0iWes)7GrR}O5ox>5t{dXxI)nO$im&B`M>a5ARl7qLP<P5Qry
z#{IwK?Em9G$V%_~Mq{KC$kRW6IOw+2VVqG~TH1G;gw;>iqHb603j-eV81QSfS`pWT
zs^uQZ`&$Hc3_HG9a)A$Xd%Ia!YR=m+k8Bsf?PWyC$)Q1h_2XpNX4iA)&gGFu<4z*z
zuTf-d<t;9e?B@nF?*jSuP_1p>>)Yh@<zDwrQ~R`ZyU)0gxuH7QCAAzamLPD~7B$z7
z1Jg%vdHwl;g&S2IPwH6Bk2-T-;jnvQ>li)X)oZ|(xzB&`Q}>noA)Ni+jHxPXy)~B-
zI^jOrSh?Ke2vi?0!J#cZQbn(*yu3W&cI_VE8aZEtDKVel2f-^UgiGi)bC}7=8OZ_4
z(H3s_5ZuN};EXQ}fWD<=qKIqd@xZewTI1hf?uJa|mJQTLhpxEQ_TdgWE?VYu*2=3&
zFY=ToqmG+8><X^Y3UD~vT)|pPSZXMmA8HFoNZwe?klrv5#~9q(p+CDZA5Ce*@i;Ww
zYS+#jQ!-G_ZVZ36?Ylp|Hzo9Yp13nRI8{VbA0Mdljkl@bZE#P34lq^RrlYhv_9C9f
zW-)NM^>JPIuB$;KAu2N;KV1+@+^^<y(wfK_i9T_o<>wKor<Z;VKcrnhle{M(lch@|
z2Mt8@D=Hlh^ax?$>>4@8qBn+wIUE`>i3*71oq((q0gEw_NnA}n%n0C4`wfXOKJGTQ
zc&a#X=O7FrRa3Sq;)jF34NqG0e1Jfwv~xv7L{2z$y-M)vanKYcaHq>8@DBvVs=G>h
zdU~eHC-d+m`y_9>6nhFhfA_8T+oPA9A*B`96RH;ywfqQr0fgb0Z3DxDI{Gt*@HCey
z_b>aIBOla=k-;+MhtX7%4}SB;ZT#3~)_6v8lDP|6ysXbn^S38^mkNfM(DLc)D2bo2
zYy9|7A}QwS!3%cnFE2xMInGDs`-z{hZ*F8$c<b#Q9nwom+qZAe%+I&}vUIH=M%e1k
z0es<r&t%CdyM-o2U6LwqZJ~*2*BDdK?hB!6^kHRg$Q>q~8PiXUT3U70n>l6c{-K%C
zAVKdt#7y$FBEP_($1*=D^-|t^Z=WaKG8TB?>h)<?d*I}6sxw|(2;Jw^bv5?b)w92a
z?oTYZUT0S)O0bxbf<>cCGBD;ZsG8#o8hzjIJajTSUe();lxj~);=tI(eR^?$*`&y5
zy!*|jLV(sIZ#D@w9?i}vY$wBb3-|nZEs5;Tv@2>-;CAh<hYVXM@3S}t5jRydVq7np
zmXo#Bu-i8{f0)f)KUHK;v--kz#=I(=faT}d=eN*=C(#0yi=?d>Fu*6^Ox5+pHc5lo
z<|H*0d~1eovUo^r8EYm;ETGG8riwkFNm^s^HeAsf<tQ@zP8_@t(g;day!F}_VNQF7
zdPzGc*?+lPyxG*!vcFsr=GWttYvtT;KAzU{{=Pd76Cw)D%UHi<X6HBj@%>J@TZ3<a
zv%hQGdPHCV7NJ(cspI&z>ete_E6z?%S|qJlIorU|IOFEhbgkUXSYF*6WFc|<a@Y0H
zeE}<<O`S_R^bYTmyZz;Z1*{VCWL|~a>UaiozCqUdrZ@D1Eu`Ao+I9oN5yIohv)Z0?
ztm(wEFM9ik&8$AMdweK!H(65-qIq=aV92TG-sfH9D3$Q|_9%`6XGG*|F6t<cr+Q3w
zO5^1`C)z_3Ul4g4P0StYD}(#|X5R)s5j}S0w9cu|e>BF)OJ1*iOhtXb`yDDdHB6zc
zfsgl`uq#+VuV32d))4B?LaFafeXhPxr9fV~sQgWX?Byx#IQ2`pDJfSU(#?F-T)BZt
z?xQv{edO_;lF4iA=H8I2{hoAZ!;n-5frrPh*ACYRTU&gIIh)d=%DZ>d(J^`j2xBOD
z^`Bq_UHZ$@($fzuEiG|CXley1$*7+IYM{(wc}mKG_Ybx=q{q9&Ei^lFb{bcBjFU<_
z%_LnZ>t2VnVp^^}<bK$xi^VJSd$bu*g1ocCeAzSo6KDlMeoKQ#4m^7Rgh%6<*Fz!j
zVu!YfdbC`-?ov5dYvDZ-$@wfuj7cO6ct50>&eFyAD0|$HArF(#v;erQE^8@)R6Fwf
zmi-sYT>8JNlW-BY;inhUjq|O@Yx52IhK40AYLL|FmIO<Z4Wtj({&*<FCf|goQ^K$E
z%9g}VFIT*%vuP+H;7OIr<FWgef}G>>pK#_9b``F$FLNztA8N&SEGiuq_FqbJs+zGU
zVRvR!MV^z{SNwsDw7f+tiQXM+79m|6-|j0*(q@S}F-I*s>IBDQAWtP>Ic>=s5nJ}(
zdFI`4FZP7}Xg+b>QKn+5aB-p|TYGD}+H%)Hd3kv<tD5Tui6SE}PX?ZrRk@%OFneDN
ziw}VVMspAPEKkKkQ9P!fq0^Kfu#~Npq)PUtFwpCN36^bc)5Rqk8e4ik9614t)jZoZ
z$8pJ6eRcmK2eZ_|mJ)F_%2j@gWxbf4chzMeZeC+U+3VNpd9L&VFR8jbl_H3FB{v=+
zc_%zw@FbV28$e3ttud9@eeS{)R_5mM?#)UP^74`!=IxMm5+Ih9mfH1CT>J3n<DZUS
zOPQ=-Rr@Xv__6|<6Cq72&3)Lp010InCg#R*1_+1!hd40d%B1AJ#gVB~H)xCF1wN49
zPE$Q6ylw@FdVNF?0MtQdO3?m#$O<v?bc++{2~K#7wRo@ws+vZ7Y{IqO#?zuldmQtd
z-V5STk+sdvEv;`!IadU0L#njt9HZLb9UW55ooN{R^kSh^Q#LIt>!^s+t216u_om3q
zB<DD&@>fh%J$&YdwnUybOi>n0Fru`OYp<QNN|?2&_8!mBr?)J6i?h{z&&lfUxBs^n
z;DIyILg|-pPghL5+RTzJ>ub%r4R<$#viN!*p|6l9Ml>`u%x2=_jJAsr*XEQqR));w
zs~zci-M*|UVe@`q`dYR9#m9Sc`;J^0QE<FQn)fJ{OwD?=qI2IcaKfpu)=SBT%WM2d
z+|VTrN8)wMKq);Oyu%zgu@>*DfD3ZAMr8dr_0#Lbj9&x>(#DPSd;Ka$vnHcu{|i%m
z5Xfp!>z&v|<uY?U!h}-?82#LPUJb1#PK}c+T=eArs#lyodxZf?|LB&xZir~ZylR{f
znT)!VThMi>)mvC9S|s>Ma1;AUKkcK>x}xc*J%xRvbv<`?NSDh^W6f1(D_|ufpb>|K
z`>3$k&3mlVx<L`gT^RA1>E&JA>L0CW$epangae?x#}D{Rc#pjuP9%`>EW@=Gw7q(i
z#z*x%pqKAGmr=I<U3HB@*RGXP3E2j?8TE3-#;ff=e)FbY_pLzA+GwMW%qoTITz07D
z{6;GypE%#IPwL+iEg^!DO!H)4G%}xGe%3O&cY;KBktRHfq~8+@h{A4vvXq%QQP@Jh
z_9#|~`CfeBgzmp**S}~~hi1!LvCZ?F?ZZ-8Gn~uSl?_LNq4yC5>DGkY(IL+B)R=D%
zm$pt`3z!cVli)4A7#29!s|dHut&TsGoo_a?%j3{e(e6skwv4(7fvTjH&*r8@v{ivb
z?yb(@lR_f#*7*azi9*SShA+*=JP!mej+JXvmrn}D2@GIk<%X(qs_T0sWA<T;0u04)
z|9bX#!;eF1=-<4aY%R{Jv|LriV5EwY`hD!*d>c5unLxBp@%Io3-cBBK<#nu;`$i{?
z(-(F&oT_`r%Q{gO(Cn7J`ELC+Xk6ywR20cBHj}lTbj2rX<HN0X-(r5o7jjLdS@-p%
z6w)spwxLPypy>*h7jo9h*s3`DzB_kF!yCdbRCOks7IJ@C%hOPMzrUwPdQ(!X>hcOL
zxvEQ}KUDwlE+OqgE%v6<X%H=v`Oux&HQiia>D`D^<yq(%Mj9AO7zT3cIt=kCS>sF3
z>kK>aDPac2iizg4cOF_V<Oz?8$u9PVCMmIAPrPvKFbo!xjq`kKFEgF(9pcS${e38K
zG={yArRU^yvB1RXv&hewyS7DxTl;Eug$qll#*&wc+Lm5ad9eG|URBDEj9B=w9Pp%A
z_lyx1jZxf{8^rQ+t1cdT*pQGBcDCGrpl$gcZie%ZP5uTGc@e^>@(_jM1PBhF<XwK!
zRQb>N)?!87luH*qJw3Csv!O0@_*ITgj#PCT(Z+AkcTHkE4#`p!Wvn;Wg_OKmye1bB
zRx(z}r%cQ~S(-wMPhL5P2FWVnm;&VI`$mjkRCe?n+~Uj@GtFm6MEf>RcZ!iNSXE2n
z<-TkrGpA(xTMn@{V~p&eA4JTZo5a{Fz&8_<lls=yyt(@hCj!r2O_NU6^q!`cSv5AW
z<fV_wO1*khdC6pW=@)vIkQ)0qJu7Nft4jHFuSqUF&U<Vz&DHmqLr*V?iO7nL^=K^B
zT;GbyzOX;cQn5!CRy?uLh4Q4RHfxi6Uzqji>jM<0rM^u*`3nW^5>(#4+xFSX-(y4r
zA>zHAtG)H~+lnbJj#CB3^NHc%4{B2jjvrA7@aPNM6}%7%Jjr}V*hwK_rH!$q2cb#b
zugBbh2^*Z#OJ=;Dbr)W{xiS_jAgWkAn<S)Ub#n5@B#K5#n7m_FTUS@ig9i_w`;zj2
z@jahkaP3jeb4G)l>C2z;TvE1K&A#?ILRU}!Q%6!`LR{!m!zspdy3z2!`TUZ_#?Zw4
zS8rI(hI$v(R|TJ)Fm;YBnJ>D}OSoUgH45-&KJ*^1${=9{t37<tT!eIuMEE(EiDmfG
zt!#C+f_n?S+0t0-RE70C-P}wOz?{)~z~d?-R<cNGmj{_;{py)T)V0?o^qvVNQEBsn
zv2)jq3<?UyUwG{mAxYUbT<qz*H!@qe>r<|Drm$sZ)!;r|OOExpX)iI&lsaHV&&tYj
z+Hx*iZEZf0b7eG@r0N^L<VCZsQ%{OdTJaAcGR9-ox%b=U8=H0uCM2pPmH#3!y)-#h
zz?t2(rupBPoMW(H-MFDc;RU)t+>jS<TdOH^h7V1LXYQ2TwKX#{6R-V*$GxFjbL}>S
zhPnP@%Iu8j9^S@_CIMQofYe!8Az|;Tk0nAh*%7m$ACt(%j?zu^uAjuJ2flk$liC%Q
zdcc9@IY}Fx@k+Wq@=1uBldqNLNzO0vauSy@C5}!LqZT<GhVq=&9TGOQrXc1QU)=_2
z-s<&7N9?{7m$}tG31X(#$u_f>Ih5{OZ!$FEUdWI7U$HPu0YMFOrJJ)u)P$T{s4<VD
z;G5SR=7+m#^6BkZ#)_TWN^~zyq&R$YL!5?3ufG2LY(r?2N^PN}%%#-;1A|f%b()Gx
zjP<rMlf8;kddFn#f2bN<>J;je)O%T0=NBBs&iRR(6MX@e=K;gayeu8a+^|41j~b+9
zT}qtDn%K3~Fw;ey_o-xaC~hzPOG(zmsBe67(Z|>4nI^NZ%Uc$GKHaWxIe8L7!#u#9
zCfNF7^5{aH^f7D0s03T9ne(AcHT$>h3Wm`I`vy<-#B8w*oVV>4N;p=xGQc{d+@{y>
zYIgo;g|B6pRf&K7GLzlBB|jZ*9Z;oyxpW)mRpqYAH}&~M-tawvaEbl8`(NX{xznlU
zWdKUT)z~`UZ97aj`|RM+lj~hX&yEZE6%q2<%*#$@s{Iyg3aKM!s{-Bp-Z}=k5VcpA
zyKbbJHd%e2J!^U#5=qU;)Piq`vb?oHZ`=5ag08UMq)MrwT3k#+h<~M&?}CtiV?VE2
z**v$xyu8E1mCgC)%6Xsw>n|O~_Oe<QH=nLEys>(Qeb!UYFmXhtv#jrCTTa15r*mx1
z`*`C;BeUaGg^i4smKal(VroiP*tLS=Po8MkJ$bU-0cykCiy?2i$X=0FlblMq*9HSI
zYiAG9?#Mburwmqc<`~S_GG6S$gV397@;A?<k*Idpz|as7)x^-2*5|9;)aw+BW$vRR
z0S4c~#phnkm+h{ZeK(}OxA)!a&B<+g997TGVaVH|`i6&x?>dJx`K;rVcmqPCye=hm
zAHjuD^A$6Hghj1lX7d{}<2%-TFB!gUCI)Ow<@6KoE3ED0-0|G*KF<sah{SEIXY46o
zqR+?b?r+Ik*mT+)$+nmX{DD9N<Bn8i`EkAU+)l#46;^`R`xMT?Ii~H4Z_I7O<;yhL
zGHS*v<c>fNw@x>1*(+%uFH|d5>2}KT$>rWt$!5$pZ^FbRXDLx4Z#F^9zBWP7z3#PS
zQIk&vv*RIc_`PZP3BL^1@CP-TX3Dp?!-AL}5!O1LQR>CmKA+*LO5!F}9A~Q$u3Oz<
zTgU3-DzuX<8OIdG8WSR*T1k6Un`^Td$H_ih1GCn+xfgnD@;YNqPETX0nye=cEx94h
z8@+x9MYeZkA!Md^Y4&XFVrC{4S5~I@^K`cL6JGZgjF7RVVYr%tSQ7jhI7Eww9P{`!
zuhD`%7MKZ0U!8(7{=5B5yU?Z}^Wy>-=MRL{ONZnG&w5#;t>b=$aneYuE|5{gZE)9a
zH`3%hJGo6vUk^)hYi^^1l2ZmN)lu7d&(3d8*S#17=J-nBZvDnG8Ln26H#fP739Y`;
zjSAjJ#VtI->hiuzwH}z3+kC0h<Z9rq`EKnXdErVG*Vno_JqvWy0o5mL!ze585vlW+
z*~IGjg8WeGY@$xRjL3}n%qbA;Q7mC@?6h>exj+ngUDbCS^GrYU|AU@tpq(iZqwIOp
zinzX131pr(h(2;4Z?$)J+P%5^jp_9#@`waj1mkD>anx&KloLq1D)XyXZ|0RxWUY<s
z66>?!Uj;+P=FTOE&QqHIc(7=x(>&wS;fP;^&>5$LTGP!>s;&&)@4gx&TKM)mk>SNQ
zT~|YLJQfjN(8f(VRZN$5I~6q{Qmw=mX+z*R$<#KQ)rmQFz?$1bM}A3#WD(Unw{|Fx
zJ%b|^OyP-(CbF{AfeGyB=x}gy0<EzK@otrz(fB=JJ$0e1FZX%%I(u($m?vyLhD|{?
zv*-A=Uln?8yeR$iztBxuKIdF|+uPC-#@JXc<IE5!@jyr*-i`w-3TxXQIc$Hqr};#c
zn$O}zO29fAimB3G7bR%2^^A?9QNs@hHy~r>6UBj35=nQiHG`uU8ZH0x&mme6y&_qB
z{=Jd8b7Z)V@{w~vJDOhK9)ckXwSiY%?{XvH%c0Xpy{jN@?N*ZaXzV_pu@v&y^WZ|T
zb>?2wLKnHa!y(XU)}@*=N`{dwf)Ok>OtOo2E;Y>qKKHR0P93-MR<LIc^@gCby990>
z0Inr3*G=jZ@OtstdF5Hd3;h2xh}+%s90?VeyYr?gyDM`q1LNw__j+yGmDtc#aD8E>
zf@L?+p6LzZYB-i|;jX>RAo5U$-F}&W0p$piP^GRq$7YA>>>D5#L^Sugw_m;>j{hxW
zU+mTQ4<hAupyN&l=@m-vP4b(kqPc1saiQl~qB*FkU7G4X0mxBlIP9ET&2Hm}qkkQ*
zPr65FPEgtdQ%tf6Ra^e(PE(6EEwUQ{K^q?}V=1(k3w`m(;}gvnNq*~-*0f@WHjN*^
z^aA?K`y=?bzNqVkcD517&}e9*RpOlr@mX1BNm-iFGJM{2C`3gD+@j>E!iN8GB5;Qi
zc$v!Ccj7`7!K<ic$C78tC;xej3eb&M>YLv;nfufBI_5mZ?osh_I`RvDAU#frAGuTE
zIT?ioe!9Vr)fjuCSNX4(G9cY!yC-wO4wjjbbaP1!R+kmY2Kw~F=dtL?!B&Oc%({F*
zG!m;c_3FJ@)h}#&9M^O~4gDv^uM(D;^~yWpLKtUeTC`O?%y|@;%4LBdl@9R-@9)AB
zSHFY8)_%8vYjSTqUKjBTlBr?%_4P?Xo$bRgWKHNK#0sB#uniBbHN<ZP(5&M4>XjPV
z(So}D4M3HD@?!u+9tZq&^xsz??XCoyZjU2#B-*-pzwgvjELvpOL)AL{)<H^ek|A)}
ztYlaNHZ(Q1l;J*fWAs9jcbKOBOCA4aD-p-*21nRTbrk?@qx0R4eKA18$pbra2d<v)
zG2K%V@mFvqKyWQM71RhJueJMVtbJ)6bh_}9c$Qjvga6yJVxt?IS<YqsQP&E(j7&*b
zaE24()q#wFKzLyWI&s&ltC`u^j@eT9*3m$I(mH*?_TSd}lCzR`Ais4RDe<pn@*9w;
zo$Wa-)BJztV32_dehgklW<lv5jb8w!4g^i~+3E8p4$CHog@X4eIGL<2PsalhAFaZl
z={FkrHk)0qU%w7)cO+6Cc1q~1b%7ods!M`+$S`<{E`Wi>v%AuIeZtti!_ZiI!5?KE
z%`pDOpAA6_b6G@Q1-TC>KJ6l)u|~np+z6y@yo_b}@a%g*T@rAbTM&&ZN0{k@oGhW(
z@1d{G?WRTjXVUJmWr-Y3(y*g7*X*~)KRrF05{6ZpeY44ZGPi7$K;c><0t^fEIt{V(
zu*eiaCUV%%-x=|y)iCl)3J_B|B&4?p>@Ecs@e*kmLhq)!vA%vdXdEiQ{8{f%nR=y_
ztX{S2ulGVVLAG-CEXk${64cn*cfb_2EO`T|7cduWoeET(v9-VFKNk!V1hAQTc~)a<
z%9A<aSV(;KLKXIWh9B|XX@4Sd6B2Yp>L41Aa4ZmZmSbx`vqRsaw%Brw&8L0mK{fav
zkFjOnt<$7oMGB|sEE6cw#EFRsKIIPz!&&+Ok++V;7nn%>bNL;w+ZT>i0$v?0py9V$
z#I#zA=`#m{_K~oh4K^hinlLQh($Z2#oimK3LV~n<bP+upMpOsCb>R0+TSkL$Ygb7&
zAO8HI8{#r9ze$Do>AY~7E2Ek-Wsi%m_UDe)OXh2}8m#e4!t9)e-s~YNuz86{B!POd
zaF$sHt1`i7A`?$SbX;T1aimd@hT|D)cFTS4G`;iuQ^zjd)Zol++HvY<c+EH0xD!c0
z>2||Q@!2{M!)U*WI7h$|RRE1o#`doznU=blr3y`QfjfL^G{(PD&Rvr8r(-nGP7wKs
z3U{z2m&RM8!N!Fw@2{%`V+K*p0&k0)-^y%G*?^DIXv}q#w51eALy2qG5zU?hC;(L}
zq2BthWkp;P&zJEa*&}utPlDjJS(*JD{6uB&F3cUIF3xfpB@xu1v|YFP<EWep;R8|E
zd@xj%1oO2vB1@z~cmx0Zv8J{kq^MYcC_1?HMJ#o*utgY1LMi`@JS6^+(!XUlSZ}zM
zdrg;rD5&yfH^<s1#ai=L>3CAGZ1wAZ9?xbFi6m5iotM;*&_fzt?sb*@XGkFC6su`T
zu6PM59TsopH&>&F&>t9lZo%XQOz0<TkSZ6NRdTECM;l}DBwfERJO3H7>#fn%5m?0S
z>Prckd(v#9Eu(C@OFjATJJtx-fP`8w-kPR~VAM7d1*c3&@5zY?x49Z&<{vwCE|L;F
zTJe8hu|48TivQ{&Ra>z?0>B3D#cxmQUtWGuNIeuk3_7VF(k}pXs3FDoMW}ilN<r?-
zW^Gc!oi7i&p#c1|n8jle3;{79_MD*F$+mRuh6NffnVC!scsQi1w$z>g9Cg|wB<Q2%
zXA}Iu8x)u1O7?&+DxFB=lJ=_7O;%SLiO`>i1U>=-p=oG{AL759$6;~vw<@~M&dwp~
zt79mzpe&+bMuZIz;2<{Y>GHF6%hkxrLCZ_fs^(i&^|))G)9>!cwiFoH+!S;C;85)o
zO;~jb48Mg%7K37OnSc4CdU(Yue?WutFT~aln?tQQ%LI*Nd2yy+1!=ZOip?qQxycz-
zq{M>q7WE%Ivbz!rruBnudkVWa5CS+D1<R>5G2TjDCFmQRUdBcg>}kp};onE}H$0%v
zBvmCKoe3nHp(l#Xfx=c@e$K-mpK9+kHbk(5iU6nARM+BkuOf_&YwhDb`u|J~Nb-eB
z)DT_0n5GH}0HX8b^DS?^-Ha#g)H!tk6=aZG3L#}Ozb0)^#{Nef!W@&ZXlAmqMs?>P
zHKA0jLE7g%{8XSHy(*Rp;b8sGn}T|I2oI8cq7Y4j3b?bRr6fts_YmF+RG@#xu#vH`
zqFUw6&vgtCMDZk1udV|nB_#ruu;)OFRMY%*`U_!U_BxkYl7{;U3}<yZP*R#QWGvnd
zgluUq8l<jp9&Moig14ox&$mR{srQfX?fYI@VE2iqEjteskXtvOd9&r`AFEa_Jv5#w
zHS_Rf+od16bEfB}^MXuV(!}2EBZqGsiM{aQX_}P9<t?xKJY8=nq`fP~3eK-wA1x!^
ziIx0DuGTM`Gv|}|{%Qb&!ThW?%i#JoGvl%At4or96}_4xse|jo61Z3_f*OVX+~tUd
z5y+qiK{L=M0XhySQ-UbhBqczRIsbJTzdRMVs}kWRRLS}PNg+I{XKkH)L_Zzty)e@+
z=r_2@`MMZ{?_*IF1qP>?K*B~_`8fRWwc;Js(f%R0lTK$smZ1{YLV3s0%``Qk@P?h+
zK?NLI+0ccO!jVV~JSpFQ+7am(!C#BorQ2JN+zB6E849II=2wye1%yXIP-%F2VSg+L
z%)_tV@94GdJF`VJ;_{2%Y<r3wI3F*l>lu7a=8yDE-o|4AK!bLv1p!w?KOV2>ssp(0
z>G8i7$N|D^_0D=r^sl?P=Hx2WiDEE=q5PH<GFa+gj+VrKyKBSwDhBi_df-Y%XX*U4
zQ?dpBml0R_2)_vhe~_X6hv)GM&Z{p*JKEc!u;<-xL#JR2($8)kfMl6!aCH5$`{d-L
zdyhlL$cbc*Rx|iJ+uL{j_yaoZSsORF;`!Cd&Rp-Qf;zdY@BZ)`J116O?m4;9_jP19
ze%Y-zTfI7w9g8Cd^fHmz0YT!pxVS>QrnnrllB2lQPoZ)kg=48`jJy{{N6_e@w3n%$
zkq-0V{W(zB4Q`(X0sUtTx-BNr>;PK~kfuCwL6I1KZif<9(kSQS?K;u21~w~zU9s@H
zpx@8QD<32ZZ+v?Gc%PD-HCR51uGKtBfEGq+GN;H55SuOPh^zg)QCkCY-Sa<Re!aHQ
z0f{>-XtZ)HD-=M&AIR-nKZn#@Q9|ES!)&7ws4lL)0kC9j!8?K{!4vQPVJABEnZ9J|
zU?N~*(lGQjH|Q&nh;jJ&_hK8jqL~>9ZW^Rvz~_^1zGC3Wl(2hjh&Mb5;CqRqALze8
zyFpcyp$S4I#^Uw$^|z$+7w*e#5Z<V1C9MI1aiy>5Wkz5#b94JgyitrQvKIL<am5%8
zyIvQfP4Z)^FdXPBiN?pE>9vVwB0DhbY5Q*5*gtQcm7S--*Nl+?rnRT1oL)dvUC;u3
zIt+_6;@%|J?m}@SH-=-0>vW>!uV33&o!s<4&vOAW6IX(WAKNqaPk7G{`;ACiQ~xp<
zR5I0Acz80<W35M!=m<bF9?O?M=R%2<&QG9H6;D5UpzA-OfBpj>8IHB{7tsT-9Yd|j
ztC%UzF)x#0tj`eG%FF9uo!y}GR;@GhHrShLR2urlHfDpo=@kQp27f(Lc9`T|QYwg=
zbSqzHlpOo+3_U>r$E*K!xiSEyW;J$SIC_hw&p~3<4Qym)K^Cb|sE<dE{(|Q3bcRva
z_XCL_ptEhEX+|go%E8lI^Ro2va%Hsr6L{`rWfq`6DdoL?opI-FMT4EZ{CH<>Z?c;v
z`ta@}L--K#mMY{3Z-Dp<z>nB3O?1F9=Bv<SrJM`01GkmZ=RbBn3gN>nNFma9h#L#Z
z_VBrBh1awU4La<2gm=)o$%6%i*bC&!@nO$c+WeI~^i1AzVSnpL{>DUpk2(Pm!@=Jk
ze?*B#J6<8Jpx}5zw5%>^Z7~Da1ZyK2Dv$a)mo0nrYg2h15^aB>t!8bAXgn6>AH=Yr
z3SaR=GE&2PA~VBnB+}T9Ge4p<UEGGCL0WZn^`Bo0>h84iKhGo7?>X6NyEyg<F$`=I
za1z1nJ=iW11)E|O>=SPP2Cp2AME*a&V(i@+VfP5jQ{8wH!pa>T+=1s{p(tYn_>vSG
zuA<SyEa$=@R1!GMpu^AWcSk}$minb8_@C+4+{}!Gg)BKOIPWUqPtUiu{`G4I{z(7g
zTkfaOfRp(7iR}DR-{zhY$8aowtt0;^FAxguK!<?yp?<`zV$^TaoS{K_ykD*SA5X~!
z`+yD*6>fks>|0au-atsOl}indjvxyb5C~7P<S&mTAN{)q+<asLn`#A{5a>mT9&$sh
zyrWU`siDWnzfrf}*9ow2YhPbN^7@oL=dEBsSW~>B1il*Jam7~-=UF0J12zCj9B$nB
z2zfSqyNH6#+k2Z$UtM`?RcDP%H{<E?kG_IAh#B(ps~N~WM<y#3*)`PzDN)|b20Fq7
z1q?z#2&h5KK6=bgBbP33Re4(hZaPc8Sp-~sN1_|N0SmRZwvu3v^eL!A=Uw?=?oz>4
z**gq4^=Qz3pRvCKyTp1k3%#Hlx$iRc>X2)ulg@q8`L`FKWr;t7M>_&`SE6P)(_rEy
zwC$h&YTMI}E{Z$ykJx2T(V2A_s3gB|<Nteh$F%WVbNvMg{|h4g4_rS#;`h;ab#79&
z4xkR<g<EAVuz5Ht-+_Zkq{2NQm}vlwB1amc2?j})$V$d<{N||=!iD_fPTQBlBcr8{
z&?E=;Pb8eV|J?}MSnTLH8L;`Lvvbq)^TCJ-+_vx3F|=OJoqHdU%O|2PJzUNnOavR5
zQNC58WZvsvXzE-i)VClvRIIwcPZAhDoO+5gEhoXibO0F&7m)i6cCR<iety3ITkDyS
zW9was+}>?s>VCv){GP)gHE!K<5XDLvWqsNLxD`$>WxZDw;3OEZG}5(VTbI}22@_s~
zC?7?%3dBvJb8qQMMa7KF%p<_PSd@bIs39&R$>$~%(MSNFSBXRXL}HM$E^~+A!{aiN
z0f(eR3P25kAaUhAmRkEnr0CAv7yoA5*t7d?-L}yqkjNUrzSniR;Oo2@1?{d>+5Xe9
zpP>L<DRt^5f!|Js0=&|Eu=Uz66o}LXfVB1i<@tOc^w+6i&_eDbwB!*6h7v6*%KyHA
zXV*L-3$_8sO$ha$y3XmUQ#Weo6Udk?#p^y)e*VM|)&HL{%3oif@P7w%pzhIdJU_Hq
z<{gv&F8e;B1sS2QgU>Y3AX~Vw(YynKSQ6zMyD_K2*3VmfdIiH*Tw1z8zRo|%<EW_H
zJ@(j((B=?Pl^*qIjWjIPcVHpOz?kA6)YsD!H#$0srOwrfF@CtsKVipziR#VfnHHtz
zw)<x%);RlWCn}d5{POh){#zN)0MrmOGczwv<pRq;aZsBCP*N4S_`a7&X!++w5My8A
zR}ITs?`WhOtVypA^Zd(9oeA1iNgRVX>KXYJ1Qz%a%?IQl!dUt3hX#H?%?fd=f9CJ3
zx=Xnjm@v40PDIhAF-qE8PbQkJw5-PRrX~N($7`pV#wKqofY-HfZ<0sMQz4)iFPfw7
z_uJ5gPoLS$g45<x1pN<dYipmE;^%_XTaVtMEF`juBP)b0JuWxncx#pHDweiqi6oI9
zbjlhBHnX643dbVf=P)%qxQqW>gLjQTKJD2pX@teAFT~Pz6{w<{V2->ToCao+gy80%
zBm65|gc~FN3fukq_Vphc_}MD#-qwKznF=Yq8_=vm|2bDw-S18_kW7A1E@>oj%}0OV
z46v*$Wz7wd=peHXLuG72hGOK=LZA?}tB}#Ennl5A_&4xBuo;LQfBmD8J?cP>uam(o
z+bdsYAtU@38U1CK+j;);U}_|X|HYcjr&xU`<iAJdc{mr!bg&6OtnZfH`g$Os>W|+K
zlpT$3FfF)tJ_%ZO*p*}l_1}~d;2$Lgc@y!bsOrKh-Yv?9ZqJ_YGMkk#F9nJ7iKf1O
zn^hg(oXK}3udU3TujLQZ%fo?|pI$k27jFCM*t6WeY2X`><+t}We6d41R3>1Snt+Qn
z{JXih__hKP$2L*rlXyAcg@6<n=!S|9o4oYQ%_|zZAfEexkHM3`UVIXQM(Q6$GEvdh
zO2Mfs3>>lcP=m&M%UrXv*_DP+;M29^qZ(u>s=2wjsK)(cEdLIJKF0!X5Tr7ASrvlI
zKaYS)9*zY@2bLt15;=TpSC`2@inlokF<7WetUEKKP4Oy!QSaxErVfe%PfM}^I5M5o
z24xhx=j4q`|Ge;rhdaZu!WOsAROk*50Qz%w=TB2t8MHEp=|qf6mSJ{lzBNV4toX3B
zw6yQO3ZJD3t+<m%&=;Nnqu~x&areI`zI|8W4Q*}I9V|}LWnn$+fye`k%n2wK5hyiy
zGZUlJG^5kOpWr;O##v~CxBBI)ieQj%R%iAQ9+;4Mz!7%9CE`h)ot;va<;BxnK$5Ru
z_PUGgg=DP~z&kCp`)kUCSi48a0B`Is0>Uzl^^uaWo)RFGQz4J{mO5*K;SaF{7I!`e
z?g2Y1NKJY+yU<JCeWZT_k}X~$F#@DL`_;v9;-jw=_>TV~f@ndzy0OL~1HpB|cX_Iv
zPUn$_fCht<1FFv^zM6s*1vEfFQOJR}(7>3*DSf;&`PF%dT$J*Tov<-#|DIE~u0SD)
zwyN@3e@Nsnj6qpya_(S%v4g%*j>%JKO?K+49)}t@^{ND58j>H2i;F$F;<%GY^B``}
zcuFH~<mv85yCn;uC0m1DpK455g$8tLl`#jbpB%C`6Ji0<#HoVe-h1)W1>^{~{}^_F
zSjgb7gp~%pF3nETFqokmkRck>IEQE^jziULg-JlnE2>I6aBH(`(~J%0Qdh>t1(u;q
zbP>?8ITmC&Yfim!A?$isf5Nz`k*0@sH2}+y0Q4&|oX$gW9^o>00mzPw77fx8D;g4@
z6F@dgZL&qP1l$Xx014o(F9Bs3j!g6TdB#RBu>p0z`zU;jJuCd|)y~RbA?bR+0|eh{
zujwA-QZ8dMxC3ky8KY2sFM^{(WdPAl90rd|qzrkRmpV}q%sL+|WZM)ck9Hj(!zUs<
zlt6c7%=!4vgL@XO*v1RcKcfv9&NLYDSZJ9??+M6!MC~)8c7pb~T_o86G1*=1?~61&
z0olq|&drDoM^6kvjd1=va2NKl`9YM#?_<X-wko~l!uzWu1W!r@axhlO-L97aP?!-f
z0tR~UH^Z0qXoErmjr&vBMj{t1EGzR%>;5X#>7R7a1Hmotq`bT>M6GGbO~V=B6fJzl
z(>P780Sl!I=H})n0@hY6AOKxeW?(4sq+tL{jOmKi)zxkLRZrpPWqI?m&a`N0^C+wm
zcdCf}di3K{_*xcVw0Uc#zWlh0%nsXG$y?OrML^6^HYO>}K>U<72z;ns$_3lHI>52A
z#MyO(`Oqd>#7Ohjl6faGz>BLgupnv8(dea(xG5o?V8Cl_dAfIjP2{n=A&?L0CM5u$
zmkK*+(OkgV24#gk3@GRtav!IbmU`t+LR8z-g^2pDS6j$1t$9n<;n=!;?v4BF5KGG+
zRw57la?j4(Ct=Z7-~j6eWvF*!zh$|HcW+fP%w~U5zG=~+{t5yOavj6V8Lj0(S#v0-
z&>=@)$)H$)o1P9m6c7*aO29y}<m|^<Qt$(n%~yye%tt}cSV#ebQ!iXS8;HpMX%3oP
z0h!l7J9g$!-vtUF;~8&OY~#IH;s~hP7fJg(TF!uA#2AW6tYO8;JN6{4vw0f~hF8DX
z`q%|L38;MA{Oiyesk6zn>A4DZ|JCAkYxHQCMEwKCC!v}h!iOkNyi`-yC2g#8S@V`q
z1;eqjRf~E*JoPFn`nkW?;WKR6mBbgijnuFA*aJU=CxI8)DgV0HKn21d{&{qOQ%Kis
zh?3R}-LJOsvaK22;k4~xsIrI&!d4^<Xb-5$$7An*x4mcjq$B*0rj{ZcFC2~3;J{Eu
z4>$SVW_x@g*eB7giT;#jsJQrzb;j^?dwb<Nu@2e6{TAXGT0o?r$Mpj?!oPR&AZcoD
zvlCi65Y36xy=ANu{HULe&W6Tw@Xhb0#WyY^#1Aw;6@EZUi^A<^9m31J9=z8-incQ0
zNdfDtyMJ&Hc9m8j2$6jkI(=`-mXAFz-ASfWqRV(ph%!V|Mti>x5~MNK`x$AVnBfcI
zEdQBLz$mE&06yobqpmk8V09dJq%aE`WE$#l8*}|z7jXu#-3e{rDq|uv1D)nvsXIz~
zcPV~=bCV%o76Ge;5FGOQM*{=@H#^#2--3-*a5f_iH=dwL{n9Ek;1jnD-tn9C#42jM
ztbwD0L-!=XceQe1H4uP-Es!|Aen-sQ-Q7vU0DA{Uydkxr&;u&?1m10i%p8Sv>1xg&
zpi{4a%0t=-2dN=w?hs3k3>>BUMW_|qqCWe3X+fy^>J1k#fT)5kf)6Z;|GZf7GCNxn
zk`1C$fG|jdLNXmk;H|S+^W?NNap)CkTfp;xc0F|sCBPe1viw;M;6#D&?3^<CatL-h
zszqZtgIjQ2%Q+S@NTo4-ee0?^_#{w9b|yd%3QOnBrH`7OP{%=TrlBo84)7yM5QJ_;
zb#3;2#pCr466&9fW&)-its4NwTmPDB3C=BB*t&!#)dUN(ZZ|?390Nix6VPZtujjxj
zdjNIsc%;zw1JJsCJqc<H5OyPolj>bSlnScVs9i&}@K>JpiXu!D-feBJju~leMFnwf
zNGo1J8qv=4&~v2%Q&aI+1-dZAdjcl#2xZp@jlf6)S<{AF*GKm#mE5i1I*@?$41}?u
zs6>VO4IkuI(P2-39*DP497Z7dUxxApYt;%ii<B}rz|c0z7+;u@x{unnnT0$)Z#?+M
z;N$M?D&VU@Qlztl3EO62v}w9}Q`te}-#{`G*e~Xn@`?UsEFgOqO7i`s<>k3NU5}3k
zf`%!;y~V*~C_{O)?@cUM*$!M3Dh<8NsHiAZPVs0{P?@DKRA`PYZ-g>M_zlg+vxOtY
zgU+G23*?G6NQ21rV>eLg&8d+McnZ|Ua23nn74_3-h!MROG~EHaX#8Mx(2xg5Tr<iB
z@bEZ;LWJTN4h`J&JnI!KqtaydvR>J80KYvk$a^O+;6_e>K4TbFQz%Mk^#XRW=d(y?
zhYxra5cw7~!wn*=&maac3E|gh>i~|*zzf(4)z8JcChe^TLI(+=mD8^YRg;#i8~+|X
z09SyzAB!l;KoI+i7A<VOhkP3lx3Yf;Qs<9MO<O`PYXC*Nr@sF$m9*l?B=Ao#Lx_Du
zy7l>*uRtE@6;K-+V3S5T7Lj&Pm&2sZOH-k;LJwW>*xALU8Ey8&k{4SwivX$o2Uu<$
zfTqnB80kG~J|03+CSjmRr<eQC%sas41^NP_ZrdQcg4^bAEW1nZ>mD<-`>IN)f_S4o
z;s~Auf|}63(JmYzV^w(y-&#RndC`#VD;x{<ntuu`Ko8-8`dJHM@~N94G{qQ$RDjoz
z(Z%smvepALN=;3$$qlCJpSRXQ8_m$x)MD<hGf;!N6!>`eu0QbQYEa!_V_q_n0zd_7
z05%5bsKUB}Y~%dqRD8G_NV=#Fqwm1@8|`Z!2nII{VsXI6)!N=3fmkr$Xg6)yp#}Xu
z2?06Nat{ZP8G1&Wl5sdFv9f>nHpBmNC4mP@5*20SslYk_3k1DXZ#5Cv+TQwpP}uC|
zzSQhda<jSowb4JS1EK&y`OBXN3vYdXhe93(5)V6S_lk%VK_*3dB_f$YREz=k54r-G
z^S}Ljs)Z_55<&h&j1hdJ$3{jM-()g)_Y$;cH#@Qbcw+I%>i+*}WG}w;ZN*EV><o3@
zw+x!rKl#uAv*Si!rHwPJGH0A?e|gED()=mdaR;Iawg<ZR(hv;0#y5Z5-DCwY0-|vq
ztR`THf&#S$BI=QAe1h%nCNt2LWUQ-Ik+u#S9us8!$2AdOE(csb+EWS$f!Iq1hEMwM
zaf22NS}u*WRvFkXy42a{GnvaxFqet`4>3#r%Z&DAH#9PkV$Mr_HKLe9bag^MCB=W<
z|B!oGv-b$#7a{}GD|fT~CGf}xfz(AhI&x=N+^*RT)aMyM?^Ygsf&%)SF3YWe*%<IH
zs3%}8Tf%(*ulCM8oa($?<7%e2nYyM@4w*t}H4<smZgR+B8n(Tfk|a`OhNOdI${|EG
z)hdQ&buw}&X=E~lA|V|}C=#VChgFhJ))GZ>SnvIOiS_Qk_IvGp?d#q9s(;L&wSMb&
z_<o=JzCX_s9(h~`AG%X0FB0w=y_NWrAysA5)<Dz&%uqqg;pIkpNNP){XII(4HxZ6n
zq;bE$&0lMd!ZdGA$t*8LT~CQd!l0Bpf*bs(rfnAN4<wBJZ&W2Z$YLHgJtW?zyROCP
zcy~uj%UH+vi;eo<L?&TzH!EcR2jqI)wZO7D*&OaS%c74U(?GiJ!JoRldPTtVQ~T$k
z1m=_&_cs|!KZHhX+dScMIoYA<9ENj&^ZACW;xF7Ky#SC}<5D{x3~nfiE1VJ}k5*DD
z+JJ@0S?+>bLs!(OTt3za>311G9w03*1eW7$ALdN@X6Y=c3!M+q(}AqX7~j8TEa@;7
z`}C5BvJ?O(H9THdykaT!b}a3Hl%xk~ztP)MK_Dthn<9<lh>jvphs;Du3ZUM*U)mN#
zleG!+)6)lW^cj@TlD+~-wt&W<$jJCoC`oBovi0VkAecYTfW6hX<iQMbOKUt@);z9+
zdBc_(kSXl`GI@jl%I(-o6Jlg!<d89B!<QVo`7RCG>;FtJ`d=^|7npfkLQz7(l66Zk
z9k|L&Y)@Kt85~TsEsvqB8N{ikK26>QaE9H`A<NB<^z$Eo{E;8=yRJ7>0R?0Lcfy8D
zdL*iFX_V0Gx;96ErG&iWY;T8L-(|_xr3gn);7&SsR+nB=h6O6vNl5N8W{J|$441NF
zWE()H98brBi~*)o2J-IjJnE#)76Z?QJhHAy8|@x7mU>vB`~%LR0;7Tj`(wZwXf-Ga
z>AMfcg*tabc}z}?Q`h)qzt5Y=$|LFDKrN-tMN(BS(lb{HSkH%nkuH_3WK%s1bF%SG
z9WoHVZG&-^^?zJ5z|n9@+MCM+hLmb>sdrWSg%3<pA1K9r<8S2yqZB-2XW_gjRzO5f
zzHmMcLn@tdEZ9k9JgnDFU0+ylOp~IO7TyQd+C!{Cvc~$#=;%e=PY=I9h=lEe*eDd_
zuFwH2>xXAD7mN8aXb3?bp7SI?Xv2GA3>;lJA<bwLzKybs&w(RG`;MTC9CTbgRQp=k
zRpIz*4U(}%8qhZu=2N{*6}0W6u&&@l>LKr;x&_bcF*u{HZNh6gcmU^PN!*6T@Yby1
zOMU;(RWQf6e2B8AiH9vpiK~Kdqn5=avF<dK=yBL<Nbd6&rfoQj1V?(sAEz`gv9$$0
zhC({?9j}fcFBwC2J$eGL<C(d!J<IDVHfHLzRzc9gDKS%YI6@1qnHzf(FB8qD;~J2;
z&{9Vdwd2g(U=2v$r6ilHEDYvgh43`+L(w&SrsVRV;+7;^*ET;QGIgzI{`PpEvgWNr
zTRMa#`;?Y>aL%Z{;JL74@ejlD8y&s)*Z}r0$G(#N+AzzId|hny5Fb|s-K9jO2cRcC
z6Sf}lgrrJPVi6J@<8}Q@vtsslL5nElQCdUaN6PPeyan6F9U|9nqgEic5m*tQ6b^Pr
z?V$=UMHozGiEYR`TlaXBNpUog0qpl}_@Qtyqi{;pxjR2p#XrNH#waYPV@3G^s5t>!
zpa<e~md{hcOY)DHs`{Hpmz-Gyv@8eOATy}Dop#><^pS&KaWrK+X6QjH=$dDr($@dC
zX)A)pJ`(frh=V#a<K1gIQ1nsjKu_jOcF#&S6r^r^PsN7hZG8XA(&l!Rc=j|}#`Mkh
zhvo)9Lpsx~?vl<KD-D?2*o&9(>kkeXH~Qs+ku0A?465Pfh_*>ZLclFlOiKlEbl>0u
zdqI<)*R~A5PlA9}KFS`uP`{rF)@3}vHjxK|k7A8=aP#nh&)~L$u%E|m-=o?)4RQYc
z>FhPmEB35{{f+l%|1C0zqYC7yHqL<pXIM8hx1wMm<H7Gaw224~6|B3!HgyvnWZ-Ea
zztlcH@c946KOy;x4$22SIiOAt+Wm93uFwj08SA<H_0(kmd>;bYX4{|{ZuOt^+I*-}
zP6^z><B;}>N&{wlg0<qsHtdx1e!3c(wKW@yJpBFl9!)?_qhTl#N{ztXc31$65@UTP
zbDUFxwZ0uQTh>GH(bDbi-m2|<;=~D`8(%HtaA`C)E`#)j=LgmM=={$=&^(e+V-*q+
zIVGnybl5=ZkzJ!*(A$m}as!wF=@k>3hXwz3jGj+gkw`Tf(J-GzRY?B^%6$tyv_~#5
z3HU7>E;f<j6?K#+qJDm&y>>xU4yZEG^TCoK89o=NdM+OeR(7BTxjz{4_r3HB19)G`
zw+GtTYV=zm(#WiV)byuDTn(h)uW`lH&)NL71oDdzD_Ehgy<flFaA;$+jEs^!{Tl)I
z&&}U-FR>2X3Bl<AW^IP*#g(@b7Ay!rNMD60hFoIs#9+tE(U5^04uoW^-@0Sk4$-KQ
zvI06<Dwd;=7}gki0Os;{yekC;9EAm#!$pOXiET1VtO}kUR}#0aU7o);+-|#}APj+A
zY=rI;o|gTkZ#8XG>7?F6pAdioH$m1M(m=v%dwftj(HI-jAHS_&kKJsGQ=`Nd>ap9b
zXci-~6^X{jomS5GV=g!w+-v5BWu%E{PP}+WO|I3dRY%DKz-eVOj%cV6LXYQ@*{jMY
zo!9)D3M;fX_1-Y<z2VVFdjOO$vAG>E$>$fVh6K@qKjO2$$w}!L*qf4IkDTUXSrwqU
zZc%26C^ncLE4VjwZxK>jikNFF&Tk7@>*csy2~tu9J_(Xs2F`iHaMOdP?bPigpNhcN
zIsmfdD@30-(WO6-Ylhg~QqBV|zJ>oZGrrI5ilk3m;Qq4!`W$(qA!Fh>-KvtN5IHF?
z!3a}hxO|+j?7(sgcQ_?^0UbljH}0aE6uZM5@$*0TGMsg1cbVfq=#%$1R4w>!(75$J
zWEyf_V9k=cv&A6I_`N-=zG_X!rRS6&=lur)=uySa@367oNdmM>o^8)mX1PMmP@j9h
zNZg<Lu^x2=U;^dW`er$P0?SwI0%U?(!nQ)20DxMzDR<xWfQ}FQPz~#m<YRpW_iY{T
z&?p~D!t2G3tiTtVSY{d@4n!himHal#8IVRact_XvrzJi1o~Y?ixb;k8)qpT`1q?ZM
zFUWZUg%Jhm4Dt{=uyA&Qj~)VJ1|{{$PvT|MC-L%sS-jjuv`tNuG`bGkheS@m>iek^
z#th<bp-%EUcl*JuVBrpend~h=NW8~dFhJB&X^f`7BV2CE5s5Z#sLb1%V{fTOPf=*8
z6}&Xy6hTPmZUaU-xLSpMU(UL1Cn6yF*1d~cYN*q2xum~uB~Y_~c;<v%9dY{5gYDR#
zxqOP!RC^v44;yg~14Ie`DL^6(CpMN+#>(;<dObkpqe7E<n+ViZ1F@wvHq*9uF5nl$
zcCOIq!|Lkk%q`~1+pR^I%@z~V-_06vz!#G0)B{D>72y5@j&gX<qY?;m7&H4^a*L0b
zHIJ-fYln5j;?#})u)4OQCsr|iKtM_UKrBk@6etU9lhj&Afjl%>PhK25<)NLBcYPpn
zV>kpF0!6Yc93->y@PZ_!_=L019sa}yphq}Bsb))k3)?%V1kFh{gQQa8$=!a47pT{-
z1C7M`EDv1E4s>eUn1GU6Y^xnP6~WrnGjw&d(IL?B%$Y|yhki$zNL7;qYfhTf>;pEI
zRo0h_?w30lA@<)UbB1i4&Q}Nbp}$uRULtoM8j}o-LXsLTGh<dJFdNJu^>Yzsohx_Q
zGq~TPFTLnn`Y2a--5gxFW2<+@3WIUg8Y_L+hh#N2Hl1Me3U->G>+ZF$Y(uKk0?$S>
zh`sjJU2e48-_b~l;bW7CSQqsj91hfegFu{#L7y2MUWEQA-}`(9^y+hKeRu|ghUJA|
zZR$bj0%M(=y0d3o&c%xhnCA|=?TR#dMkUGG<Q^M?FhOTDFf{W*c2v64w!vn>bTz8k
z+5<o$=%b(_%#iQ^Fnp(^UnIZ(!z;h;psHmH&wwvNo9OF7V(a0aaT@-TyO^zYvFreA
zGmz0I!+GJAH@^;*?w!HkYqL`m^4mYzq|hG{*fk~ErEjiRuTP8H$`r>qd6gM^^tq=j
zTc%!A{Ze85Ig=l>qV7)D8f%q4uQceSg2rYmwSPA%wK-z8e(#z2V`ii$wtpXfd)i@s
z|EwSHZM(B9J0O1a<Ga;$&*Hlu_3bdMeJ2hK_Nr)D7X&eI=*U%OBmR$lwJsUefSbA)
zuHNaKobHr@Kn(3phqR9Z8Uc5y7)^Li5HQ|I^2ZrA#%40n)Do8=QW3i0Q3#`1fAHYS
zKjemhtM&%~b9%K_R=glDz{jm2Pwzph!<4ZKTD**mNfQ0bKE8C0g>}_UpLwRQ4u|Ja
z43@zeda3g6U9ShO8#iq_fDp_SKZ9|o34-Z#LV}vFxIvjvQsi1QDb$Uc=TfS-(12db
zOZ4*1E<X@M>7ZpSjBC*D0ilu(XY*`YS|qZaB4jv|v+I4=VjWLfxg{8m3*Kr1rSQPk
z())s3TwKV}&_P289GUS>PYMrI!dqM!r%bOu>#LdBqdm<@KEfB5qv->^K;*r#`aR5Y
zUt3*WT?nnF2+q(a>A~7uUw2beoHSLRzi`;NNtI+}X+&o7cozR130ILY`cM#}tq`UI
zM<*mtf@1y+rWTq(4v~+EHePxBsl*3C)f$20G$glZHFPY*Q%|CJu0mKk>}ekF60)<%
zz=}iDe{e$b_y&YP5y+j~MeMv-uGtYJG_*Qpsr$lEb?5t^0YcnFST08#dJDpuDJU;_
zAXrvsdpIB>E`mz5xZDsQ(vKs?QhX49ni_m`d(mhVfA4jOAFh7XckZ=1EhnO*olqa5
z*DJyzjTMM721{nQ`4*y%S5td}F@pM{cLL7>Lu8nW08(C93Xu;+@OWT&p8wJ{wZYF1
z3}dWvSE9v*{bFQLlR@$eHia-h-K+D9Jaf7Sur!$!I`%ZHh|fJjvz(b}Axdxk3mb4>
z@CvR74S&y=SHKpjlTC6(Go~)`Ki}7KV|te<%28F>kWqEi*&6kVx-H4qE-fpYq2`4B
zS>54WMn*=}e97U(rzYS&DUOk#!`ulJdu46yjx~);ZO1r_6w5~RmUg25`BmGMY(Rv?
zF$TTOqQ|X@<EH9P7(HGk9NjXKsKD`_N593AXi?F-9R0=s9feUH2wlyK-00SsroOA|
zi{|F$6*e{s@LG}KhnQlgT>fF{DN7y>P4Y_(3|eJs?V{TYC_v;C?*e!hxU}ddyKKY|
zRLia9uS6rtXJW=8HgPpICxofR`7OBGzrr9WMxp=c@%G|yrAe}~5Q~eRtwGR7#tJ@j
z(?QMD=hv(0SWJ>9`b(%YL`{v{kcWsldOF9Vi9Qn0LLmZ~iFm}r5RGp4zXOQU+|JG}
z#j(H$Ns~P68BA($M*oBkazq{v@y=-dxBDN`%gT()G`rDgZDVs2QI!%lzF(2_EBoZ*
z=BIfQvov45H+I0~c7A+)FS_-^&n71yfN^~~H4@sBH0Z~tY|corH==_@n>TNM(wf~E
zfz!JP3xgs;((n@)kBv-y60GAFZ==caq;>(C_X#UINESP^Hyp^Dm?-Q7ajx9I$AcZ(
z{-gH7z9eIwZz9yUF6`c^nA&UPE>2BJ2}i44;Sr6Mm){1NyRvJ(WRKy%zx%+S_~|R1
zzE9`j(|z#iKKNgJ4rXJOmkEt)z5KmI@iWOg<9frt&3xt<JXb7|ou!k%!@RfVy(n(C
zl~2D}z2S&6XKO;nhg}aD_gbVgYk_3NqxwlPRiCeu5r$t*krXs#^h-n{9hvg8i_>^N
z_GjlN$>?^BIqsm6uic@V?4_MP$o^Z6WIY-t|2khi$6r7%;@Q;H2((82Rkj7gswcxQ
zr)H_jd-t|}&1U_}SR6<38CvxU8^Zy(s-`f7gWyH#fKkA7cgLg}>Jv-Aa3++e&X6Z>
zNRI5<M3N$O<>RppDjQJE=peo2@gC;*xWZ>WU$un_B$}TSscsSDU;B(zxfZIsviEDq
zjUnG|mfRRw*>v<Iy~jpwaPcPCpC>5Je&z9|cwCw;LNH!sq)tza*UQfvR5rr1)`95c
z@pQd}CQzEEZ&Y|RUNNLDtj?wx&N}QjJ!Jg@&myQT#%PHNva&QV7zDq%@(vN;FdiBn
zPziurbOw^|>c%KFVf7bzhSeZtoDe0n7Q>&2L<|!Wv;kE?IsjH4kIsPB>hr?z@-OH;
zV!qy$JvtDDc@w&F4)S<u1^xPfpLSjk6x7x}LHp$<<2WFDXeet7uV8;;TkKWAo}<T)
z7o+(>5J|ZTU~N57J_a1Gu(ef`l_h152`qgnw}*cJMm@aq6{}Xs%gVwg2)rF@&=nQf
zDXO0WE`GD|Iyl_c*48l>mDnXn#izc6`s>GIS_cqFqqYN3HUXZ4!OJ+%Vnt{WL74--
zh6?1&fs=YSV5H}llfexvjn!YErUpOS88%334Nw|qSm3g<-VFgQMd+|eu$;B6EQUV0
zNsAn2b*AU;7v2`+2l^4J4IL3&-3`_}vF}wq@bjC9R4zNSGVA2n&l@IYKwwan>z(!v
zWAm>_`g_i%q%101$;7CGOW3}N?NA6ek<vy6w!+U=@S~qhQ*vxO76mXzm(1?d<wP8!
z4K4wRbN#=`)1lmzzapHYgG4F<CEV%qId&PQb(B?w{t|R;?QVFlUsHc4dgh9$h`}9O
zZ^)4RyQ4;W9)Jc4gXdcWDvy;wB1U<IYEKy4pfX1Pt3Abo$0HfP&$Y+4`AgAsItN@c
L3)9raKkfcI3}5g;

literal 31283
zcmeFZ2UOEp*FG9wKlL38yfcc7A~I?O=}Hx8&M*!}qzKZaWRxadf)IKb9c2a-G&JcJ
zigYQ0ApvZFp#?>H6$nU6AoQBMPjF`Z?p^n;`(NwcweDT(%UUiJli%;;oPG9w_VYaF
z_osoL_O9*ww_`AvUAVJ98)7h9d@z{JSN^>fe$y56XbOI8y`rQ2GyH_tk`|Q&Kfbwk
z*4z_=5e`9rHbpR}?J$@hFu0$8yy%lS(L;O?&HPQHf24lr@jc)E{AbU<_g5XP`p)V^
zyFtO_D>$Y1-#4V28S0wj%yl2r<5Ln^dLF!eT41An<&e(D(=G)+AKkl0=ve6X(CfG3
zW>#!$fBp64rr~QR<rf02Jg1`<Rw_o0*H)>~M$~UzDKOi~8<jvtR`({}-^>)ZTcf`O
zZ%^yrg8uT<;Qy6hC^Gq8#w&Xc*LPogvkOPoRqDWo93Y&H+szwB_G1I7y<tqgSMhVT
zlKHlE(xvCgJNL<9iG6{@zuxltyIp8}7=d%IcA%dIkM_{e&;PId;<`_feM9XQLCbqR
z#4hPuKRr5J(TaC3Kj1ZVpE90DHI&+0w`OrR?qifHZ`4uEi_?RiAHUTeYQaSku5zV4
zIeJ|qoQ&kg4uAIjIjWa5k+xTHPnfUg{6zQMLI|P#=>-*5yNzd+2=_srRa+x1D*T2Z
z=AnS7esZG6U^R0@lj|EabyFl&Csu|T{xN;~t)G_XCljoQJ&HY4qAc+MCRlO!d7E|J
zrkmRw%75Z~rl&M`kLSl?ml_VO#HyH&C(2)CTPm)1BnFM=suVmrq}=Ue0~ev#6l@ju
z*@Q9a|83jBcjgX(s{CqG2F(SGK$fT7`HPpVsI9|cc(tP=$^LGB3oI+vNet%bLauv|
ze4M826MZ!eq4}pY(q|d0ibu_+%sh#qs08=MclVEY&c4@<Ro&}s@aXDq;d~>}rq=Wr
z3%5AzJ~`X7tcn$uoY=pDpEsw)TF1+hmih``*&BBZ{;}**rpxK=qHC9&9qdc(x_?DW
z&hrwd`wjgG(Zp`>u2Mt!!m6Uk+Gp)L;a!-Aw!Hx}4KB3l>ZPiF5<PA+QJk3AVtBrg
z-J+b<7Cgxz*;GfV@;Tqz2Z1F0y|SBS#VV6j$9FSy-}3Y2i=?U2bd7{4WmECoBJ0BG
zR(j)2A;Ki-JpEhkgUK}q1J#mLi84e5OfcDJ>eV(QslA?WO)O$mYn>#q`#MQss{`&N
zodUDs^IF>sjajhf&QIKf#)U?yXN>tYrFEHZ+V34AhaRw{>NRULlj>wEY26+e%(>wP
z<-YS|mqNV=IWb3}guLb>Gs6r<j(Jh7@GZr&5c<!e1g(7zMW`?PXn0%?%*v`uN>@`_
zGAm^28B><WvTKF6U>*)-b<$nh#`vwZwY0S2BA<z(1&rYgMe2*<l!Oe1f3PbVE!pME
z$rP&--i*PRZQ3G>d1+|RNKV3ze#ou=u{Ni~U(R7Zkx&EUHhFk6_t-W}li%b}l3#Yt
zlgc1Wus|FhpVE~~I?<*ne~4e-SAIsY)nR<jj~SWZ9?xK({Gltq&mqpax^hi#etqx{
zEUu5$!x=Eucbo2Au~a#%Up}Wtww?XvP2obiP%P_~!Su#ne)b*u1fP(HMQ^fvX~(o!
zdPNfY!}S{yQE?hyr*PgK8a3=$)e;`lFcTqu+TKdHYBzA%Z|X8QQLGtr7458G6Ti#L
z{v9)+;|bAJs%yKlwrh{3(8A~U*IkBo$PV|9yWzAhoY8TNgSl*K4W;K-M%5=5EH;^&
ziY1j5tF0!{>FkP@GK+=Y<4a>_?smy%hzSPzF4^q3mFkdU8zxLvk2tT?p5w*9U>@4d
zRb9E=qn4Ffyw>hlF2DTy<&LM?3+X2oXs1*{?!J)5ai*e_<vF)v<tA*_>d3L-uXb8v
zf|ceQo$oT8wblmmO_MLJFJ6g{uzGLm@h6cZ5GHxhqVl*e(TVdGPwop+_w}=+<4<8u
zQ|Ft;-E9Y`Y{kilrTP;MwLRDlQdomrK#F?8JZZZBh_jT4a3qRj*9rPcjQae7Dmj;k
zAG{DxnR{(pmlWxyUYD>L^U%<0W%l><$-%7wQh`*Wvq$&7%|v?Z_?%OxtmyR8^Qcgi
zQg#Ulk+s>EX!xg2YV#Z;I;{iey1>LID98*>m5-aOQ*LDqO(euL$jwC$aoEa@!@6V#
zk@Ma{g`9<Lm>+zn=EQZoLgxAil)=`?4=OeMi|yOlieB@HIi9b$#Xfm%TSQ)N)hYUU
zGv>5VyxfV*gtiww+(Ej`imZn6bb(3tuL2p~doI1mGl+I+m+hFmS5?y+M|0v`dN^rR
zw<}>`qTt5;T%)$8rUAj6o$?UcolTnS?$h&bT4b4?NYWk0ZqhZ|!Z!~UQ!>+M1}42$
z^17c_UD1+RoiTA0`ZwmOS(jdif<?v5eL~Gud3TM)0SUZ?)${qrgdFO6?MJW4ORVo*
z)9pXM7oRn+>^{+Y2`XZ+k|C@MC%I}xq~}DY^?<vD5nj)}Ja+c;^NHtS9~*Rnf?VoG
z{M;V|%sp-$5#Jl>n13{)o3hy6N`1b*oa#~jVzkI~ZJP+@Vc_U|?=Kyk$$6uxJk6Qh
z=>3&;`Cc;|e|fTkOu@5LN}pR5BlR_(S9wzLK~}IG8YTLupE@#{o6=@N_;dEGh*+}z
z%UKaNtbt+G9wRA})~<RJMRvBna!eKuk^pXmUsTtHYRc)>y4%H2f3Q>SdewsCQ@mcD
zz+0!M?$$d6!|KBO7TQx8qLM*M7);&5pSns;5<cueFG6^C89r<<y*@sA$3Z_e+y_dT
z#0dX-tX=iR=f%|~l8vuEz&1UN>sk8PHq#U)d1CUq^wh_}Z10{YI`vUKUVrk-buojh
zwy?EINj|TrSuv(N#psPIVk?2^3XCn7&&TxISvHjU(GfDXa;0W*)wWOHwr8(>?fSqg
z?U^pp9er{lt$E&A;fQbqFASv~#Hw-M4Xg#J%J%jKP^-Gwy1Knd@w3*AszRz>xC)|G
z*N8tSIh9dO@2x3bE8m1U75d7;B_`rJJ0`PeB(=h<{@%DdJtbUSVw`9}>z&MHRJLDa
ziPGDMeQ5_SjGjhG$F;vI&3?`L+Lb~Yu~mJh)HX$5dgPbvMK{_{#)NrYodd^|Vj4+Q
zao_&dYOMv!qzSK!hQ{SijHlRfFLrEvRwv}hVDAiz5e;Uzfk{$U7NPlsRuS7+SrnK2
zVA)%#eDTB1&VWdl?kD~%(&XLqK_vU;Mfq(kvg=BYY2J(L|KSDK)YQ}Am8#p-{KuT>
zTBTwu2O^DSJS_^($5~#r-0Z`?9JZqu8>fE6JY)0fo#{Arclzyd*6OFtn}VYYul43;
zlID5|HOI1hlsk_v1k4+$ET>=aCB3?VrMKhqW#w7fRMy9VPHIyt`oh_cjVpBL-Ip|j
zSoK4bbo$nWSefR-R%Ob4cr(*whks#;iT>_o9@gq>QQb<Ps=5_E-BO$CjH-W7`GZA%
zo1|7r_u&jyp_WF8HJebAdedO)UznWqZ}??<S#$S?X1TgyMV_REhM`<jq0$tZn99_S
zG)Bq!J4{{4t(afauIFab9ZLE{Q#yhYOz=g6TgM%GiHyZ24}Q|8-&_9tWt$PNR7SI2
zb?tR$t4q~c`CX29Bj~QFROYcSge-||(OHdLzv1iLABVxb-2VhHtJyC3UR-)*fe-1W
zSd8Yp4OO-`FCpc8oLQNr&hlrzu+o~IXZFe8U@$rRP=PSpHJ+EUvog)ghC1isJFzw4
z`T9Yfv5pL{w!mR+Ns>DzIJ`fLq>+HlbiR64H`@2!MNuyEUr<+?4$X}}cOzHdC{=$z
zy>^+Ls^jJLcWE`m<5~X&xA-PwqWeWmOA-6LtS7l-dbX>%=yC)rGxx%jHx?~)nwmo4
zbeoENUBZ*@&?<r$26HU8T&d&ucuVUrzy5kr_XU%#l^uI+-xe_(029_h%|Fkp`T7T1
z7M&m6<mx*1=A`Ph5i;PDNsTQnOO4i4X6-a%xOD_CMNBj{!AlJnQ`z*KEju{5>fX}C
z1Fg^a)@^oRUKW;I4al$-Rk!Lrz*HLT7&;;%P27$N{;<5v5@9sumeh~@(i<&uH}H7#
zaHBLaHJTHV=|1x8+fruwi1j;nz7EYb|9lJ9*__J&<uRCNC8On+3rCgcn@dHzCX+4_
zC5Vp|-8n_{Z0>nJ?j=ej1`}oC_2}Ms_$dX=eG3LMt1E+){!cZ9#%$}+_Y|Y;Ouo_s
zvK6XF*bOOt#g+6^>V&$C@OO{aG%r;cMBPREf6KO@Lra4l^hv>^+KKl+UVlhlfN*UP
z#0t&pc<;AlDPS;ZZIrT8n%p4`mB9W+%Gy#aXI8aYF2n6I#YWnc-Pw7^*UnI{+v7Cf
zc%>cRB=`OG%=uSN!qL2|C%8o(@bHrx8yLD93pM?$QkT+bx1xAXu91q#oBC0lR-oLc
zPeTRN$YddG&Dsx{9M6AYntp?QtX)w!uEz}DBzXp{%mLLvIdf9;P;Yi5HOPXc;QULh
z8{I8L^k%qoH?)MUZfMyB+I_Nz@Fa45P`=0LjWS=R-pBc|aiO;Eo}j*=BNo@|6+<>b
zaYiWZ(j));>il`9uJYO>h&nR?V>I3HdYN?rwi>)H9)si$uhX{-|K}I~|Co;~aUv!-
z+W5R7UMfa)Yty-1yftkK7x(!sbf&%Dyko5ZA@Qqehdle<v6y_&6gTOaKj!tu4}Sur
zY6qVjrl@$<2|qbpu0XV=96=F~Bd~{NAj+1pd?0*qNt<GWerNDX>OJ}i{c(N&md~e^
zSRYjU=PWb2PkjDWDS)a}Pzcn;Y1}cNO{eaa`{hL@l=hXN|6}=N5fA#6c8kMDJH>Qh
zbHVF>Lhb)!I}HLU*l?es(?uETM4=_I$15{8SDKiJrBm4z2?_2MD$tvswr_}!&Moad
zwN3cC!PZ^-qq4H>sPhFjHN&rW$Xc@2lsd0IS!HwA-NlbU?+1Npfw%coYg^mM@_4>}
zU|=BSiGuHFS}dmlTKK9f1*7diRLBRdE_VBlBrPp5NXhZ>!+qiONU{7W#T?yyesm49
znNHnT7E*iV9U2p9oLWiLiOK`pL(q0IG&PMCS)U3K-y62{F+(GIFkDsU`Sa&-8i9T?
zFZyq&t<5$st+7cd`9X8p4SD`c%vd8Ut0W6I%2D7@&X1x!5d7SjD(*v?FG^vuN!*ol
zcx581yW68in8L1Dk@Fh+=u*;Qzfa++!R?*<$az-O5?l!|B6h>|aUC`h>VywCF-2i^
zbB2D<>Y&KT;$Q>=NRjABy;I^nWnRvwxC3Fb_CPsZWbx9g^jMQfp^{x~RKRo?z8+|a
zI9#Nt-TLZM+)0lsXxbA4IP^F*U(WV?_p4XmJ-hsdFyG;jf2})-*c*pcVZ9iPQ8Bf!
zNGM<ERmE}Y<rxapg<{9%50xZtbf7t!tkc`xyV+0iiD=Lqsb*5cPoIQe`n0%Z$KIi?
zzIAWZgg5JJs;{u9?oQ6m^6*XX>{pa%v-H%hm`P#}jb-QSW#}cv_@^u{FH2dzxJDTX
ztGbd*T3^D_hZEd;2Odn8bjd9Akw~KFAKd$G^EOc&T(bPNGrBVdR*9P0D&=_GZ}xM2
zvbw56qQ@_Mr#?SSAB3r>o30T_$u!EauYYjDci>O4LKs6nA;!GO{wOXhJKM;@qHEy6
znWL<anMMiWx+xm!xDVz23#+60q}m53Y&f0H`DyNb6&(&KfrO#^+V)9at>&MKg!b=`
zm;R>U-Hk0f<bHh_%%Aj#M?vQFblSb@+F#n^Z9%rJT$^{EUOg0YfPBKHdT2Bw#X9!}
zv6tqHBU_jINfYZIdlCCc+_iKY_wqRt@=#mlx>sn`mFDR-Yk%n@zubAZOS$`A%KB_Z
z2ZV&_;wlWlo5IM+%X8}q<9+MabHauFx!KuuUG9}hsUB9{G4`uU3JO?UZ{R{ti(l(w
zLW!cClWxO+d(C)$WinQ!KX^0Y^*3U+YjbV-j||?Sj|sWcPCU_$Xq@`=W~N4S-QyPX
z1iE%pkk`8##8)@A=yrvLb~EGT9rT*#9W-X*()Us5100Bqlm-Zqcq?U}ss2RY;keNc
zX*y)v>O1>HmhWf;E-MJDE`NT4xG-7y5|}-e$>L@?wC2L9O6Yd5Bz*&cJ8u0{9WG3*
zpGKh=BJ91`)W!n^IX$OrUr~o#I`1FRE^%t;O`YD{GB!RQ>z|jlcBW3YK1EY=$6l$k
zBaKP16z}(hoe8cb=Gg@7U7fS&`m(gC*Myoq+i#_h$Yz$5cvJFu`QYgRn5{_|rjwIX
z;k8e%X$!snqd*~M=TSf17K)#2Yio08e|EXUyUTT^4!>Sl<$TI(Eb6Xu?@6Qd^K=Lf
z3n>o~(&{vpUee{(HP-g*>RtjPS=}O#$KN(pT#QydtpayFaJ}hA+=qehw&Dn|`)uKo
zzKfqiOnp+0lB|*+X@!X5+<>~XYivkVkw@`dc72pAUD_>&lk=aS*eRqWmBL*ev9SE*
zk(rcIk28v-Per_JF`uoqmxdb*$v#R)iKiPvM^|TWeVn`)X}LLtP$giRsvU9A)$rlL
zYcDrtW@geM3Szt5sM20uUXZ1H2k$C#%GrUkIKo(FhLW`Kdx-K`cAuuaef>Mq`UuH>
zd3IEHe61126>CYrB?m2!=1^H>9i9wzQ&kZZ6$WvqhQ8Z{8{8HTRHSCD%1B})esw^w
zE-^nhw_#}{B_%$-g)2SQnrqRh%BCg4Mxjg;))6}FVii>@A*T)Z-{g1mo|q_NG{SC3
zUSU^iq8SKVLT^{1Z$tOEn9kS3$9Lf5@mMQwbGNe5bhX9)o4(w&g`)Sr2-(3;M$;p$
z=I73BO8%`%T`rNpFsdncU(-71_xVvh3|)1y*yuK&a6F_4nWB0vc5k^S<lYfrUh7~t
z%XtjEy=ZKlhOZ3j02ba8$lnr24Rqg!oIRmoy({`5(?HcL<SwY%zPr~AGiayx@kF_i
zY3wc**Pmk3^y)y-k15@>z;Vr{Z#!QL>nFv9Sks;&q@l6!UaauP*dzpaZjE)?_5?BY
zHWjzN*6fL2_E-RXBz%5Ez9RB<dP1yf^`&er1~hpO1O4XmhI%OIoxhasicHdzgy!X`
zrSQB~D;g9=B55D(<rwfTqs@`Z=~$7avy|p2y$G-G#;rGFiYd2>$l^A2Nsr}P@=|h2
zttdtq0|4f~w$M%-r7YLlcDLIOuQ2tB>}SVQcB&s%;XF-EH=;CSFvUA|*dL8t&+itO
zFk-t3&@jh*eltEXUokYIsp~`q2_q(_OMc%BiH;MCEm;g;_t?Jnq*I%PQi`TU>8&C<
zr=nlDNGmBc8m_Orp`YPKul9W&8mT<%mv7oRf~rlxzqiOfBP9FDEk%eN^`A^y{nh8g
z6Yaril6t{)Mxv%HWwA0+^%q5%ZwgZ9ma%<{-i7RkAJj?rpWIfgSQlCM=iFqA3gXEh
z7j5?Z>RQ~fiN2gvmvkdMwL-?%?jn3kjPQN?e_buVqDA<$1;@=Lb59Z*q>KzqV*?ht
z?p8VJr0eSh&9do4(!B$xif4VjJ5^K`M(E0V`finqD#Ps`G<OF_xB7S4-t<D3wf&@q
zc&CG?CAoMnBSp{JS4D-{*V?_m*p%zB#AK1z`$$T8J7}U4#h!%=mDne4czFg&DrS6J
zWSjgL8a*$tuiQ4b)z3gyay#(8tYC*=Nsx<0yA7kd(4W<<xNshnsB3LTruJiZaL{96
zgl`|+AKaQS9aN~tP%LG^j>xh6&95e9zvl0`SrVC9dRg1UO0%=cKN%0Mf9Nu9!9BWn
z(Ny`M1;O7JpI_d(Sg&cFJUh|xI#Bb)$m7W@{zhX{X=?fP)~wgtD)ai68j=e>&q){b
zOzaD@#v6z3nuGl>zGr*dQZn4SQ^%bvuPHoNbv=uf7`e{uo5}YN6iae;suSLU=>!Vy
z8KKoy_GDmZ?OptFP2b_s54PsX-?rs-jO3<UG%vQ^{pr!Yv9Ym>rluKLSy_($qie?M
zZ6*tj?jw7~fo0+PGx$Fk<S<rmdCpK*f*z<Q)71LZ7TY=;#@`_MqEi~~Fn=k-fwH^U
zqUy?t9fC6TuW#+B6He2DAv(J*S=M^$SFHbJ9gyu|=zICkZ-#HZHeAt3%D%MR9+H@S
zsZo7y4CCCmz`x=uvtCD(UruSm;;cyvX#oyzzwPkPe#5zYlPj5!Tj~|J`(LM5fJ<q$
z?s9E=ao}fXop+o0&Q+{8q*feqym~ftmxxQ7Sk=J5<n7x60WD&?s?tg{U!<?!^jA{A
zGSn09ypU0uyp!r5!Jlk;e$2v?dwY1nij%#5GpC_s+`>m!$5l4TOJxC{Vesx3#TLh_
zSI+{<gff-jDqCJTt74*GU^eS-F4_?KX>4z8|IyNH(`ziP(9gw>Q?1>;EhjT9mb=F+
zsk4+A*3u(}q|~&(#oQ|)ji~XaPdSy1)l2mErX!!NDiiTNfqtSDa~H`&zx_!Zz8_mN
zB<~-nif<q3YkuwtYRZ=2d97DFq8dFJq7DIG)9Y2S*e)09#0#N({U9Q4a9+PrcfO<4
z!=FW~zEb#nE30DFyt3*Fq5A~gpDm*?pFxe3$$lYfInsD{3;UtIXrN}H)55WvH<dm^
z59R*Jzv-OL-Y7|<ou%`Z7Cjl|c9MI;To>^_)?9Qs8|t`9jQAL>s#{ZFt3s1IbKO~I
zEX~15cIi$=-m+U++q}86ePMp4r?Ihd%+1TW&+3L(hC_`+>T)BvN|ojAKK2d%vR-pL
z2+Xw+W?n{KecZM+iES=K%kAq&|BW$8xFL8f_ozc(@<Q_vM^cEiY;CCJG5sRVKPOt;
zN4YCtX+*R<o;Kq>16Vj_qFjHRtq-yu6nDd-bd9SK*t=EBotHF9)4wOJH&Eq#JCEq@
zW`yaS|4DXlY$txqGOWh*THa-q1<Uu8nB=ymdv;v4D`^5@xB?AaU}5@5u4aN}{wJ|K
z_0B6vbI&u2Wfmx{;fi*F<}JpGOS3PV3H{zROtL?tagzH&p=!^}e0|%@8w<U6xe-=m
z_gzcuSvJ*%^5kEq&&=#T^=z27d``~IWiY2rags%iQTLwNwcT3cHQ7%0XxU&U5lH8V
zDAnp(Ntp}FK?WB~%rs;>^YE5COU?GT$r>Nezw&MQ&{FH(?rfu{`G1Z`TlE(C<rG;T
zy!F$eTNl&HvQ8P_8iSgak!28_^^;dtc9zjG`1R=Gwvm|b^<OQ+MzXRJ@?6UphJ!Id
z)#R?0Th(j)bsrSm{v0v3TQc9vwD$9V9~cu*eOWarN?2*_=RE(<TxMD+HxBE&+Ok3z
zU~4985tjohJeqEeUg*lM?>mEedLF7w!!lzgQhy@y`dBfi&3Xz?UmGNg8<AzK_Q(34
zG;qFd`j;1Asz24ePXG^--0eLxu$GloZ4;px19bfCv#kf3Zu~UZwVo|lKKiS`dixcD
zJwjso$^w;hjuZp;M+fkn?gFY+N^eieC1?90)7Me~Xj`2>G3PktPchsdI!76#OUeat
zHd6*Z*5owoz+|6(wIgTW^XGJ$@7-HtcaM-6L2gWTLD$Rb{Ak5blxhbt`ayqDp;=wl
z_bo5DO`Go?J%Mo!`Dsw7XGaUp*vo`s)^|C>wdhlsYN=hJxBG?n7LHZ3ot)L><=gd1
zCMnBT0z<2AD<~+qDJy);c+`IHHh+m=Qoj3BeXR@oXzx--84k-6rG<Y26=xEEL(okA
z_JMdP#!FBp(<*iNG!KtbSn<B|Pw$77MyQ<{EodCLL44~4T-sSfW8-K@r~D#k!}d1n
z99-+D#had=zWLn?jk#SSQI)yJZYM;n{OYn}FowpcjbTV9tu#3VOzq#Rp1(x7<-67s
z%US(MvJL9#PVc)3VNg-R_3cO{Rw}KAV@}sRf+TonY-~uiPbXY8eY(7e>v@55^OGbo
z;-&^=)l*v}GSPduutu;|5l>neq$=VMUK!XypB<(9jpthK1(nF)>>&vJOg-IG#>0l=
zxtzDD1xs_>KuJdVv{}*MHdk(3fGc;V@Kc#Zl=wc(MGfxH{5dKx%X*7#)svciN$35%
zn0fdY*Ou_Rc1h!B-tNA8i$I<zv-pr}Ve^X04hpNeQ|_LI2M$Ms^?Y>q%*_Oi=zMDm
zzKB2!(bc@)m$Le6<38_P0!%B&fd-=GPEMSK9tmgHmwnyul}!b<o98lyG)9MpS`#9d
zX2>(7m{o>*mxOhlo7KRZmT)z{>G{iAksc)1?ut)k)%#v%_s<J&vwfqhY;iZ(xz$f`
zy?q<zoO>&6fza{7({u2MpxtFHtLEulZ9RuGgr~v|JnQigPc^9i9U7xWoJehIy1_fv
z_{G+cv)!$fnK7;+twd?5l}MfmidS1X&snq)32HdJu<puCRwTM{UVhiMvpaBlUXLo(
z8#H*gdhR#nP9G*K#dd_%Q{;OqP-GM49J-&I|3oR}`h>om4XYSB*HihCva59+DnSwX
z8J@bX`~IqCw3~c-c`tGbE~;}K=)Y}?9#-Mk{20J}_lC?2^r=_3Ap^=%P{CkM1z!nW
z;s|ib4(Y4qqh`jBB(x`&-IqQ*<ep`-0_z^|VLD}<r~c(^{YXM(CFE`wo0fFvMTG)`
z;`sB-+U?uJ<MT43x4*>|n8hCez;*i2t%=gfXX$I|EBth0?)=3H|IhAGRE)t-_EZ02
z7OaKSU#gR}xfP3AO`RX^?4vhxUDg&C9I*McM3;{I$s!Eq`hUbGrthjvkF_5tt%>W%
z<{n~(Z^j7UfIOd4RFQI^@;ebX?<O(RfS{>O^h(lkd7t;&o*t$DB6PH^``(^QW<M@Q
zMUrE!UAik)8i%qoN;QC@X}F%3*}z??PvU{9Pykh&(Jo~b*6*zmc!KY(n2ND+!3bfG
z$$?2mPz95qjME709+<peswgvv;I+Y_J+#xlNdZnDD4t!tGoa!)=S%mB6!&pzSZlIj
zr6UZTWd9OZ7(DR)_u%RuI=oL4*nK8y=_N}00`==gmS>ly#w|W99LcS8>+v$3+=t%k
zya}nd@REqSA$axq?c2Ba2+Bwic?^Mu0q5v`pd#*@;sX^?XKCSH9=$X(l2KBk6u2_q
za9hajfpkgx;c~w@39p?GcO?P^!VE_jd->ft!>Ni@P90^w(|IODbM=sH<tis698X%C
z^8_-8>6x-(>D}u)vL|N2apAQ2u7hePp~rPMloAzQ<2F-f$XF|?x3uSf;93ulI5!{N
zf6%A@x1G>xtu4>h0RdoP3?d?QVv+c@G1H~R8qMTQo3{=@=j2>xc3NnO6I<E+V!Aqt
zv(;RPcKndK|J=D0jlkL@5Qs3JUv{SUC!O+g;!~e(0yL}PJ5f+W#@^hzs}>Ppt4vuE
zJxsRZ0kPLt4Ec$=DW3As?w(IjlFp}b&aKavRx*_<mvqBV4*UVN%<lE?-|SQKxrS)I
zKX;3joKbst7a6CloT#Fu-4|KMn2P=;PU`_g4Mi8TtSl)zL>ojgoD_G@&CO9hKMpIL
zt`-_$QQ3_L{U`226Dv70+)#(-F1H1YTlCegudq{s)|S)sQ#3|<%l(0mYR<JN%_;5m
z>$nmoUC7W_8Ir^=opS>Mr#?Yh)~5OnJ=vcXhsd|_+|tH<4vF_Eo+BkK<6BI$@~z9E
zt4?k)n_73Zqzp!BTUl)%S(7sM93F;wkRk3zvwhP7n4gR!@7g$r<T&5}>{l135i8*N
z?obFaHf3M0)Jr<W4BpK5PK#eX$|2cTwC13{i%iAh0P^`{@7{e?>rDth<59@LnAZ^u
zJ(*=MIp!Qe`;z?L)}s+>iDU8=@mV1v^@^tW==HUgSXBYIYtyUT){q=mU=R?qx;!`D
zX5t&a^ZVmJpmrJAtJr~Tf7c+=uPxWdK8hc4rz&c$bTBw!_;tA%H_C~Yp0bQAXIILT
z=%0Iv9UJCX_`koW!={KK1~2YA6y8!Lm$tBhAZa&IiE%D!P^tqPmIY@7liLPt+v<>*
zCLJiPxmDm+8-M?1x14(~jZh=;C-82}$6+$H5OC=05A9A_{N0NP*9(4L?Ib`uUDHY>
zvJU{!!VJGa;I8#?tyuqtn_;9Se0fwXP`12q+O4m4U%4+p9?=(Q(y)~E<#A7#)C^!j
z(!zP9XH#BOaep9(F6N1tyAP?jI$=*>07Y@X-1+|g<u|vZ<#>G4X{ur2Eo=%9pg>p1
zfQJc~C3zsW?5e%kz3M=ps0=@_Auyv&V<f)3--vvJ2;3<mXURio*?=U8g_szIIKcsr
z8GknrB8BrtWc@gf0QYwQ(l+f6v{Numtf<uf^3n7J4us*zvp0Noi12Yi^SAbK357wT
zXCvLTf)foeV<=QI@7NhE`W&e7*ez_YHmg`HitreV37JW4&E*T^p=u#_sjuJDUSJc6
zwQwmolbzoIw!}!RzS{K9;)i(|saW->xDOR5CoHstl7W+nDxRIEzI>^R<Qof-DXX^R
zq0EeAr#VF)Eos~T^0%)?u0H*CcO<q(gcOkNFtiSf>{x~@qL^TQQ}}q37Joylg?kcg
zTKkv5n-Gx=8gk!2J6b8k6?xK+#->85%Sa@g`+G==M=6fFx-6iUQFum*>elwZx@iP}
z2)X=V`f+6az%e6VU=!sXKcuFQ|8jSKbh5gC%;E3QoljGZ$opuxLb_<eP?aFQe$%!C
z_ps28y?YX&(RqCx+;@%=#Fg36NTe0K+{`a!4E1bdKIMG`JAS>nOAhieBJ6n)#86)t
z`co|V-RjazQPo}xTXy+E<2Pc7_ipl?I)<!qc6+{jI&};CFi@z0jp_^CiT*5#^8gs#
z;3fPMeZ#Vad01@Xyp=6X?7n}cpxwlrj13V9@=56BNed6lFF{TKb}%zL+c{eb_hI-J
zk9j@qZ};cMgl>O+N#19qKCamWiaET@aM{LFtsiq*nY@4IZ2e2kdip%8zZ~ds4t%*N
zRo;hBpD}Z8H9NL9Y@d>g8Q2$-Mw>HYRe9_D@>0I-7I!ev{Stm44(So@kmf?dQ`CG3
zWLt>ow9t((&5O$EUt1l~OmWUH_fE@K9lRrd?jG;r57)$TuT~hO^`gw~9-vEk+n6qv
zQV$E5fCBjU-Fj!qh+o~O?st8!l-WJtyO|Ic4M8iDrlDtf!=Kc{ZSQr?qln@-w`2nj
zA+`_NH}cTCBN6paFFfHS^sR~lJF8CyG2l4LeX>E*VKhiSx;DH(HY><7?7j|^K8}po
z(1z_@Pr^@0;1I!=ZlkEEsKTbr5ZW(?Fp)3<KEre0f>#1kaSs~%>I_2+r_>iPV_&lZ
zd2wR(Ff>z|Yoq!`pz46DHE+*--XJ@Z$a`prY9<5Qp{!Z~yB7Q}KzG~ab+}O{$yh|s
z{vF1!v*6!dSo)oP^6|p#u2OgM(5r1Cmc#KfR%KY+`f^+4S0Q_2;g2KXXUVG!?y7##
zpt}G3d5H91A`T`27*JMKcI_<20`~fE8(Tahk&NA^5#Xbz!i-Xzdbt_5zpYapauXTb
zx70|4otA=yaH@k!+JL1);kNjCvJtB2<E}46V$ozH3P0U5EF2#&U4>}yc?jiMR*$Nw
zm6tKOdO#3;f8L%}#vyjL8<q|+!3n~Bkb*p0%^i;7>>Cr~U{Bf4PjukOvVo&IDljA+
zfsJ{#$`ofdL_j$|;PLj`x}5A4;MC<%{L27nIsf4JAq&^%#(L^=A9wCocp9~_b2_8R
zdud_7q1%P;+OKf-5zi3)Z2!5EpWhsg6w`0Ci?xaYeakV<6Ti~gCPx^0mwpu0R}AKA
zIaz?_InvkwQb{aWP1R6-$pgvGuG-RXG))D>h$2?m_%LW<*7o8&q8SsDkKN*4U{(*g
z#uK=4Q5@p?LB?~7^Uyxv@iRHKqCx`+Nfm38?y-me$&o~)BSD=(dL9a3glGQv_DIY=
z;!K0G14tlfg~&soO4iRXH1S_A0h)f?{vF+np!uTFZpsXy2Ifv4S&RWB;g8~EC_`k!
zpLwP0DiP3LxRA08jsR7%csV{lfGLS1!%`G|9h~e+ZV)-=+A!-wnb&v>*7wsb!LK8o
z3g1T**XuKQ*94D5Oe|engLD7uTLgi2phrerJrso8vSBDVXA22+r+CwP-4<e69~KEA
z1dyd%Z|@Q+c^yZCNQIH=5IP+*j^Y~CX6paGqSHB#=D&9Prx1~3P;l&r-reu8PjD^d
zuCF+N%5%0Y&uRv++3<bsgLF_n$e<BoaeA6-v#9!hU2Am7fR%<3j4P-)m;fS_J>bH)
z<Cj6+Nr3^Ya93w^1bB-H+nWB}t}uiVQPs+OF%XJSXk8+^CEFA&AkT;mj_bOM#?cC2
z*3z(WA1&--YwK`oq+TsfEGg<r!Gn$>2Zv9uZ-M@->+0&7DZ<OuO>WV&ek;RA24^D<
zs<#du;U%EXT*XUMGLEo-`!s_BfDkQ&y^kOzGnBW02KUdEeVzn;?Y_ur|F#a;KMBH>
zGmWavMGlOBf!!(Z9{zZ<5FpEl4u07XlB7W`ILzTSd5&&B)eGVT0N^RPnhf297?=ea
z&(#k)vJD7BR|5P&WNs+$U_+Bzl<ijb>}azmxV6P_CB1&*me=2((H#MZ^M0{EE2|rJ
zZJe^Zb+XUEpJ;=lmH<?@|CSmwS3mKi#E}6xP;H^xQw54)LZg+>r&~C3v!1H-&29X8
zu`*VLF#M5t=R6#FbvlCd-%=Ae-y^Ytme+pF%G#?r0EM9L?Vo@XQdYW0;;rPbHa*1B
zUw90D9hxg|ck5$uN<G(`_eO}(^;9Q6PFOCmdLps1<+Fxg=hW=%Y$12+meqL>4)w^W
zUjd!Hp-~Z(PP$wTZ!NBxLoq>M0mlKcSV`C!sy~v{v2+j`E9({AY16-uZ2^#c9bIb^
z4w?zrrQU^#Sct$>kgOdvSIpC4zOhg>+vVB-hpFJmhqP`-{R1>^EX<^E1N`R3qO<eL
zBtTu(fiMZ0eIjzy(bB(^#9i05VV8~P9SB<dwA8~6ip8od4MpnwReLhUt8|^6ooArP
z+2)te*Mfk~OWLC#nW1jO-X`MI*wc?KEwHOJxvMVETTH$lu`W5cq*Efla@`NIOeA11
zH6gbbm&4WP+h@mGvO8Wt5F+&vbUDvL5dNb7MyrK4lyOLBM6Ji?rbl(cC>KDUHLT`^
z+|}4~g<ILR1{+R+ocKaPFjT@TXaD~LmaM1oP^V#+WM8+^)zyVje5iKhUA;2{iYsW4
zdRA6F5>^)AEm)iHlFdl=a|+j3{`~P{6!tKB-{~rg@tCU*Ac9eG1hQ92mG#1=*PmTZ
z(B^G4Gs$S^J>~4q0W54OCM}O<jJT6lEJStVqmb%9@b*s%fFpzjczVBQv#v4|!h+|P
z5aBM?pcW8dJ7uD@F9=}cu`}n<ZG%Hi$aDN<M0T=lz^5H(fvO`!P}iua+E?n{hlEpY
zVXupjoGK1L8@;r?P?<74STiyx!u8=(8M+rRDAe~3d=+H)LK0}t@ty1sHvD<xFp}BP
zxL5wrPuO=BJ5r!7#fTGQ=jr-Ea!7a;)k}QPw>qq>TjLL5CYkr*FpdCYjy^2Pn~pP*
zxL5w5;EUn-G#2|!?rMTH1As`cpL5Y5d<zFSfbjXfE3iB8dKh>N&<;c!7A^mWtZ4)@
z$ZD%oZyLghlbwn<Il$9O=Xoi#Ndx+lNk9T*TM&YC_u~NyVMjstE*gZU8j!{;Z|n~#
zuxhA|lZS|vD^}UnvX^dL?h9OS^<Hf;#p{x}jNWo>N2`P0|Kc{Dd~@e<;(v<w3!E^)
zz7zlPr^qAX$Nps<d@p#5f}rKY2B#Z}Jy#87pKkGr%;{)1ev(medoJ{`!nuz}4i|p^
zovXpNW9A7DPwbXGZhlyhQc+J^39;M0$@bXlof*zj;+xC8eQZj=w)K_d5jA%hagj(f
zFBe?A{nPf|{J=?P92w0WnwU$ME_tiYqmx8v-(V?IuXk9axOcxefrySsENLm0Wd8HU
zs(a_A|8d5D_VfKp?n)&?vCT3P96=*l*?Gtd@&Ok9fbY<KM>n^|eA3!Df?%lPlCfeb
z0R(VIFpBd$AkLm!zn=n7`{<7cXb~`zmkw=lC!dd#I|_-5G7u_R2>AnQf@b?<dj#Z;
zB~9)KH??e8lW&i^B$2z6!c{?#HY}8fq?+6W2xqj&z=|+Ab+QapPymrD)j(SjE(97J
z<Qxjb7xFA1e6`5Ob_dm*29;&0dql(Z;?HOZ!QQ#gcGJ2J1kS#v497bY8Wh`X$6Iqd
ziXDx>0)t8iXn9VhGYEh@e!`o_V&_|eG*uT9qe$qkXV<!Uz1G7k0@hP*9fd~$6kJ4x
ztdSIwhNMZ(J76Z_VIMg_x42J*V5@g(H1)9!x-;ft;!itzk&OlY-X!<Z^EWfb*(N-;
z2bqHEWT;Y`x9)t`L+nd(>nRi1w^7d@?=hK@0pw4b79_o2rWj@DAwDP^vSSXuQe)Sy
zUFL<hIPwtkp`hk0*4j@TQi+VjMn`x?Jhi{)g@+Al^Lc<7z<SB`$xg2fk2FJZq&Oap
zwIm<ai5V(x)~iDd9Jr1SU)hMCsIkL)lT<wIP4caifFoeO*B0rqqa<Kq@laiy0bn--
z{-=6KxieF5_yxZt_klfzQ@!BHz>yQ&%EVh*{Cgj#r;DNnflREs%$th&oO?GGC?m-_
z<iu;Qzj00WT%E2F!=c*$^G1swYIOEhS9f<JHWOCE{~k$mlnRr4c%yZPiQL|+AH<2o
zk)fL|YVd@~jK-qu_3cKtBB6I^3;Xl?Cy`iFOOG?POhUqN0NcM$hYYTgoIf_&E`baf
z0zw?8u7AD<xl47om|i_V|MIQ^n3+Bx;yih|MW*cY<G=44e|WF@#IrY?cAE%0C?Ck<
zmad<YEYDDrN1YiY)0DEZG6h9NrkLgujs@S)rTy6<9EV6+Am)KN9bQ+Tz~Fg@;D#F#
zV$mTxbE>y1up*Hvp2HK~s{*E$_fg~ZbOQvzFeUewAbKS}K3*<xc{b5+EK{|d+Z3V^
zg(3!u=cT;L^<~%<|3r-U4PpX}!}8jzEtLKnI}BxF0k5!ADHN4$A_3Q+IEMBfL}S?q
z;{2>E2|%OmE0m+?m#3>bGj1#Z&eWk;_Tr&e2M1Pr{`f{nWCcq2XanhBm|0v|h@}U~
z-r@aT!uS!saHHS$m>>NwHlX8qQ(of5%zw}jpeZV{$BD(fw>}d#-SD0k8?9XAvA8Re
zHU3UlulnHUd2hI?KcC4<zl|Cz6H*Y}Bxc7e9oh0I`Noq+P&adE961g!#SZX!RQ-IR
zgPQ>uriR#CSc=G)n3%xrA3uB`BQ_a4SEJB?#z>iyeP@Ozo8!@EG0|63w)cV=&9*93
z0Fg)>6w7zy8{Z+Q?%15pyd`M;4w%BO;>OPB&NIy^>k6nJM9|^Kn_HQ1YO%JhV*WWw
zF0eLD6>oV|!PB#fD`hV)hX}_Y%!MpJoHo*0n=e#i&$$-U5%Bi81(IsTGln358qlp^
zm0WshDu4$ZU0muBsMSqSntV14yi+_d6(a!1JKKDCLl$sO+0pB&vHU*5p4Cl3v<{S{
z8EBdrII`Dxs~vdM{1Zt{KW?)FdXLwX0NuvUH~aG8N&#9IfMXu%Tzyw)1{y#f%!39j
zg=t#3Y;5g6`8>y>R0bW}4F_a_AoAvIeD%nnguIln>sL026NnJ@7Pb<KB@LhoMqqZX
zU$Efa=Rni^sLqCCXL0b0pS{=!WIcEVdw)6E+oXq#@bKys{{pL2zZ_1!nz3<C5HksS
z0eZ*7B#G4iK;%}1@$S{&rNsWzCPr@`r!0TWFqDXf*zQux02h;^qhm+*20yf|kpHc~
zE=kk8^Ex_pB42nQY9U^4${R|(MX8(Xnx#ZE8LNHJ*ETD=kE-S7=!i2iG9nKF67Lq-
zXu<jZy^_{%MIqQZXHv^ut6VLtJR1~Oj_`u|<)+J(duYhA8K00qT%=91)sy;l$#3{I
z)wFK!1g}t1vE~38KIXuG=5=@ghc&eXXjmunVl8mPRLvOPH~Cc%{&yt=0x5s;+T#tT
zVJ#o>l4{?Q-nVr;Q1>iJ9e<D4|G(U3cu35L+XVe>Au#Y!bxieeb<VpZeIdNchMB})
z-oi%Fij2Kr9X18*-Zbp%D*hX9vX4FGRf6>i#=6wqyLZn>@jANG+a-Ls(<0o0=iTbi
zMap|mv<JY^5uDd^NfPg4^RbYxkU<ScKwUj&h<?U}_aE0@N)DYxHbTG+bEYKYw;bL#
zPJMWz6p9p%4558^<7@t~MSzC*1F|i&318h;41g&XF{7boymS=2=i(IKeuZPm-&wKT
zl0OXjQ}PP$y9aN!*8MiD0`wjV=sxnmprbsvPu1H+V}Zh8>aP#!8v<?y-MFVw`1r4<
zzbXXo(5gb9T+-A;{y8Y{T?(a^GXE4S!WmvV!Bj~O{TCYwfYsYV-WPF!0#RyH6MjZ&
zWA?o*-}8Roq$fbQLG))5R6wt204r}2$XYpO1&5tmbLfcu#>Snn`2`hRfD6T+kMTYv
zkk)jAINdII+wCtez(5s(0sDo7h^j~g!I8He^!ee*FLx!rF!5)^Cqi#mg$ZUrBc021
z3*i%%gXS@*+H0cy1QaOb9fdy6_~HhIi|Qp6@*8RjadN=e(_or9pzxz2)7jZ6T&<+2
z2%cwg9Qm8wqP%t3U{{`=ytRi0u*RS@&nf|mydAh6N8dlu&dMg18$(Bz4ix?lS=w=F
zX=zmLzM5C|+CpCnma;UQK(v@xf0UhHaSA5~mhiueCBzMn(13Vt9acd-0y*{3qeoxW
z-_pE1865231wR(57;uCgUcsRk&%W}9lqrUTr2SzSB}W_?`VE1v@I|<YrUp*Vt^0*i
z%DZxi3X2imG(U5IZv#!A9%@&Zb+H)^jN~{?QV`mvD&COmk;fbLmmk673Tg7pUz4DC
zOX`wQB$As@V_;)rg9JAyl(Umv$}rnF0@QN*_m9s&mUc-}2Eh(^pv-^1$!;2sP=BIk
z2pw_<7A+DB(`8<K^}_fHh&<0@s8(gy>Bt}rR?5d4i2*Q--jti?+i$gPA)A|*xR@g;
z6jm_p(W65N%I<olZav6So(MwWzEjt)0!lHk-5}gUjjq0;>Vn)R2h=DChm&jVysXq8
zJXNuWM%oFqN+a^BqfHA93u)pcjtm{0D!=CHHAHd^!++v{V35E5xe3KfOAEC^+NL31
z5dgh52N+=oW^FW(a-Qh%0zg{<ts3r|LgMX`{+mM)CR2HUJR$gXA)5i)od~4vMCtQ{
zgammI2o;o-*`<mLEf5d!m1|v<x)e*dz!K=Wq^!ymA~e^1++KXT{qf_(V&8N&eoX@K
z>L{P5NuFh#ShCMyq|YD=KByo4fh)qiaNv)5U12MY13D+yGH8i8+LGPI$3XsLM8%;d
zc`M(62PpMIr6navTa>J3EJ(!GnO6n{#z<SlAa()(eplP#3L0;exDf=LCoJbMdfdU0
z2sIvS4p%j`hO6_Mk;{$1eGtGbn_($4f>;i;G*B$CNMS?nUhI+xcaUT|H`YSBgXi@o
zi<n~*|HTHS1Tj2G$@!O_O8_A38(l)0hpDyiQGLy|>yVx1{VKtMuEFKVatqsLl-#->
zfV?R+Gs0lhMoS_yFIo_XBrg-x(w1Ahn<fX&cc7}A0y3iKt1W`X8f-|pdC<Gu0ZNoN
zUvS9c08pbi@|m!`c6Bk*=m86escuz4r5w>G3Q9`$z~%XZ-5Du7YvkQ%X0PJ6tGrrS
z>EmiZ+}o2={TEnhH(T_nLpR5Ev$RTraI|Yq*F}y#5CUfa>)Qg#IP-I78d9+kU`7Wm
zs;jFJhN+8{WuUNw{4!Kpxi;zF=<eu92L%J^g%u+&5KDuTgZh%x+tSvCx-}db31PVm
zL0RiWEaGasDZ}>;X_2wq<$UgPvOh24qO<Sqrg3NWxlx9hQI)H8mGztX)t<c(`~x&w
zK=c7X7rUx|T0!^^kqeMVVz_ISTq5vj#pS&-(6BnKQD7a5Qr0H?!}|*DbSl>8@*Myx
z!Wj?3)}7zqgO0}qT41zokZuD|QvsMieGtQmR6S2m&l2dyx)vbcBfb$7KUh+k8gxj+
z{Kd7)>yv%!(r_^yGCCH6BWo^2YWgzzo@j;eu_+564b4C_SW@J`5>N|c>-BU+Q{R6x
z7pgqO6pq{tl(cFTHQO}5k;ANu#G)x7K9IlG{}Ri#c6gTGNvL`gsjm?S>dXjWphhr#
zTcn7<dhdhhS)hFbVrx3&3}jqiPif@9%42cp$Ooi2Hle;`fC#%em$5#=T}@$ASg6g$
z(la#XGJC?sW9yb`O)CoPh(LC@uT`RvVLCH@%DW1$v*ZK*69v?OWGu9ky`QXV!i6KT
z`4!7f-`@8w%>ALK$MF;J%Br|RmyW`~P6r4R9GO4xvldmjXtK8lBAMk;4nQ0^$IWRz
z9;b+`)s$0{Jm&`x6M9nz(dmG~H={j6Jw`>i5Bd1YB)y`N;$ke_O`pQjhooD%(!(ax
z{AF2rdaG`AGoZ_`o2_9P=%m#IT5Np$!a93#E}xXd4qBih^o*!+paDOr=SQ3^tdso}
zfSwh4&<G6BE+P(nw_yG}C`uK9EA}AonHO9+0}?<4*0=Qd+AIE`dsvv!Zaq(wXs|dq
z&=B@V@y*Gw2G=tzmJ3OL5;lIH-(lsY2+llry^ou4;mI$P$K9JA{<xG-xs<@8DqBHb
zt5Jt0p|EbDlDo$C>~bw-!sqlsCO|A4BVbrwRp2w}X<6ARL~IfL*5ObM6V7Fk7!)+X
zm)r2rxdK14gpPKgw|GLhA(^-hPYUF<xvJ5RaOg-BmTvBlWc+igunQ@E9wN!HIjNjZ
z#XoKB?xuASsmrv<<y#F2$_|hasH=v>>FMcM9Ow;1A8Y6aA@2=ApqwY)IjDXLFpjGg
zFC(0MCSW)fG^&HIRj!!6i<S<dGXnG!&y(X3Beo9vu-Iz?ISzm;P>fKYHzV7E6agvW
zM=L8Ubf5%^eU5ecDG-R=nrWy-2w^ldG$7-J#(Y5p6Oe7A%m~sNK0k?QgcEIX(UPXD
zivyxOR}6Nky+)&TIS2MD(ASMngJudM0;OW{LB0uaJQrTuZvRwan+t*z%n_qGO(z!W
zO=%1jWx9jF=?GV&e#TXrSUH#L&IC9wb@&4x<Xt(aby!q_y;baCmC$g|k*Wnthm}gA
z!{XE@92ry<)+}swFSM<p^mE=v1QpLMq0STe8KR8z=%p2N(o7jNjAxOki&p^)zzmoU
zhR}f_gp^3DHKck(Ou=;~kFy@$<g10lL4{Y}Kep->w1@(YP!zX5nzD|V7ngEFz+1bP
zAaf<6Ru4G_#5AV=4D5BD>VJiftw1@Mc_OlQ4E6$qJ(6?A*{h}iwpA^K3vzc>%zn^W
z@M+_X=i09lL!q}_Vc2p3i&L$;bEt)IaizZBsG44*AC7{>$_tGT=-@8Z^8EaK27HxV
zsKP6n@6oX=6wv;t@S-z-xx_Iyc{s|2PBZ1EKv^$>9fYOBDZxI^No0hoTr1_4F!rG?
zAb38b=|-va`u9wh5^BFu0SGzhcMKV53ZdIVy#yWEgg(6q8G%DDz3%>VpWh!rGRqA7
z6CwOMz=t2e>9SKq19k9SpZYEX7Rl9QvvhhomwAoBQ%hjr{_g?q>y@mP5aJrE5p~xH
zxWdD9K3TIb0JT0GNG9s<!h<G+I{seIZg7r$D!;^QyG|Y#01eBUcJhdJIKaO_w;C7f
z4c-Hhpn!$m;Z-7o3)@%b^Zo~t{|_YpUx4HW(3b@)aiCepBFr7u&dXx`!KT1(Ld}Pr
zb(9XlZvymP0q^|W%T2lWWX*-3haG1mD7i}f-Ty*iKo+y*8OUOxUtb0~L^Y80abytL
zH34|Jhqd#lxc*?1cfvHC7%3d|5(E(ThlTnUe?ipF`D8T}FyUQ-Ga7xKERgI&Hyl|8
z9^1uzzy}fw^xm5yWQi&5Ce$4qiK%1$)75?9F(3NeO6>1>c3JCho&qvreEd%_JaOb#
z)xYS@jP0^MrohiaXRQvPNEd)4GBoiCpBh3b^^BlZ6cC@bP^Qk*pv;Paj<hGu0E!f-
zhRR^j=+I-IOEcWRfB%Q3rhDir!m8f&fcF!N8c${a&5}O5K?H1%d`lLZDTG?_eK3J#
z-jnfQXG||BIDuID7+HKWLQ$ZHH-a-F5%dhfFCLA@Kv~OJl;k|`Ui%PR5k*=IAtA&|
zRGFV^!!wWt4IQQx#0PjFZ6g(s#hJIfg?xJKR`)A^5T%at)G9DoI+rrtc+d%TQ*urp
zpJM5nt8X;-eAT>a%(vU%$cW_C9n6LY%E0MJEbbr(rT<Lw_FiNV#SuPU(0=aQ2*i65
z79CssZ|k(cPQiB2d_5CJTzbkBad56I;_C|_wGM|kPq@J2Ght*bG-u*p^xG<s1=PV8
zgle1ua&=a&k5{rhRuPl}&yFZEk1h^`g?$wkr7j(0P*qMsXS-CSxl#mI$G<buSC2+!
zQgeV>sF{H-%7IQkN`Z5v1n|8MquGO#>hT0hB6k6bj2?PcA_mib0mvM0R<}x;VdmTn
zI-QBUlo2z4k&3&BOoDHvEidCf-2*YUA-3_NZT{AVV7doUGemap|B;&cUmzC0JY3`d
zf}PBaf2@2AI?1x@j_3?1xDXLt3?02LQb~F4Cb=tb|2O~#F3)K5I57wZ@a+i7qUYZ=
zx3|COIx$!ip-)Eg7=j%4H>gF4bC*pOq@dUUo$n}3R0RP7Naaa|JV0XN(dUp?45ZM!
zf0%;py$H#HqJ$&Ay18wH8H;ZO3Ky|p0Hj#QOENRX!9a~8BR3KBx{_b5a^UAjA*c|e
z^Rz(D;1C&RRX(1Jj-CUGv;=1ZFsE?Bq512@1~bh%mJXYQ)^%lp-zspfg@=#!eig(r
zX3u`W!KsB(C5Fn_(<fvsNKt>&DZs+;+=>JWt<6&zG?2<c-ojr@g(QJazvBvR>0~8_
zqm$D)^6Yq9{y$X84I+;X1A_DU($6d)OS?|-xYA&CNr+i!649A-DSU+*5H<CHkeUJS
z))gS-4SW`!E0d)e01m-9De>1NWQyCqkB<jyj{GQiT7YmvEjl{02g(4<HjWIOopRV-
z>2tvPjMLyP89-nD^Gy<g*tH2)5k>Cot^5ZQv1qzrKV9l$#4Bb$??wzLa`u5cV$T-~
z&cGfZ2osKkpFE6=ScP_Va<Ijae;TF;ZCC(2F<?#}q0Kbbp)<T-MmYRrV}6jbg?KAV
z88IB<u)slN2hw<)lIx{!yLcgX;2z{?q=y63f%RQ*x((>|a0&0fPCgVMY-0y++I)%B
z3wq|aUnu!0_yG7ms#{Y8L125kmv^UaG`JI-PalHZqDw~J`lU&ClEqtiVi>%(75~KQ
zw*p}i4Hcb??^~=^CcqOJ?12A8N7%uv!A<tmzVrQsuLV13;-b!_2`BeJN~17iEap%C
zpEIq{42PQW#u7XU*|+uMo<b?!y}LX%2-(yBhyvRybuL4J;PDs0vZ?`5jk-5vvHwe@
zKo6V*>%Rbc+#$#Y1A8t_fzqnk0KYGCYRMAjQ8M5LbVrBFExkMJP=P?*;dR(B@4jel
zX;4LgCs_)sTnuYvpq3q3;vfgx0Wd|13n-;zEV$xb^iF>Iod)<D83r$W5gpG#g+|*F
zQUaEaOx@7^r8&3dsos^0K8KnLq$z`-g+-+ob(zqi3x3NZmmfaf!#U~<UKFII%ENv^
z1)VBIjO00lQ-b^e#NI-9+`Qb+>wjM<HpqFXL1|@v42zE+KMujUd>xLP3a{CInG8Lh
z1`rxJ@@~D8j~YC~HP;t;t<cxS1W<k?7LF&n@^jc?fo(bFMqjP#3NmEy3`HfrjakOo
zWGsAZ*}wJ|loDacSY+hTJ_}@Db0V_S<H*o|e9hRRw(=aX*o1{1q61h_d||5+RZA{P
zFws*nkc5dvVBq54_XUjyR7o>9w@|QC_>=-X-h+Y6UBht>_Rz1XtS=5Ck1F)Fz>I0q
z!Fn9IOGpXKr7ZS(qrk?aWL<$tPQW5bhUPoNp)dkvU^(D%1D<l=`soR;1P214$E!`?
zpM87gjS?KyI0s@7(siM{d1Wk!EH{Oqqx(!qMs-lpQqYe3lFy=E1B;8su|wv?56JZe
zVlx~(NrJ8vwL9SGv3lRd3(1Q)Af2*b42BmM7Y~6&fsl}F3Qh*Fy?4b*^S19Iapn4%
zHwbL3wsDv7Ml!L)r42@WV8ZBC7f>o#dIV{85Yd6q5<*uEsR&0-3dQsG&3wuiZ5E4!
z1!Um_jFk$?JG_xCV{HYGuFEQ4=8ffa7hDk8Zqk9pU6JI4Ajbc43Ut<xv3x=2DFVWl
z2G4lufCs3MhY$x!9zu?OuQ?5`cf8w8RD-UI`vAF%=ix)0YpVmA0RU?iVEULC-_V2L
zeS$P9qN-bEFKukj@1Z;(ZX@ig{qL4<DdPao&OlmeGXOmSNfTQ{Ndr+!=JBq8QVCGY
z_QK;&(FT7E8~nfZYj=<$fQ;X^AD>=0Q~jkJx15ot`AiQ$;AB9mhGYx8esS}QaADQ-
zixPI=4?>Ub4Graui-&%d-Fp*x4^=_A9(V^8A!WCEh#pXvsBhXK+K_t%CwKYv@30Jc
zUzM!|Fn-j6`i6EddK87a-|ShhxxCUWcp?M{wVvRPNl!~V3Ss>9&M5~gXe1Wc)ET$1
z9?k0@5{n@>vH9P-0NRIN;X!{0!gs2Tz@uc&kpYK%jV4=>Nv@X3v1R}of3prJ$>3#h
zACf+B&a+s6V>+E`82}M@B-_^r5nNhgILKIS!@DWS(>l(JHVn_40hT}1uM|P=0>|s+
zfuYv-^qi5Y$pfJa5x<kAeaS$ao{v}fM+Qm1cgEQXJrO4rWEaH9fw9vNyum=Y!EA$*
z4>qaPrQ&0tKIYCK%L`cK|49)9u;zus`4YsVB8v+-6{ZX7vQwZY<os4bDa5MzvtNIN
zX8;-gt$Mxtrq>l6m8wrtRfY;t2jCSQuR%}^4mu&T+A_doa6ibR-UHH4=p7zzhqadW
zDMhS5vN?j_UoP^cy#8l9b>q2Q!N}p02h=U}Ksx9_8vLZiH(*CWNet1I$WHCMzBB^+
zGz=bOfF|7oKKlqp4FIB{4^Hf`XY(MxAu9ne5%&<dMvM`9ZW0Q}uX>k-aXi-#J>#R5
z0ho|Sn82BhQ-}mbB@40@+KlEE0lqxLJLEn@mxpc~Pyk*++y_jlotIeav0M(Glf(`X
z1LO8vnVJeX{k2rBTyY=pwJeZLIz~p#ZZ*6am=f7exw!ymLU81Or8;~g`gRLaE)lQ-
z-C(LfhujjeOL*=$>7ONnLINgE92w>dnu`Cey)zGLD$n+KXzjM!b{WSN8ye>s8dp%j
z1tE!wEg}jio3h0f1O$`-5dsNlyOfI9Fp3KhdQ=cl0oe^n=vD+}iO8m`wjfcquqHqt
znR9M>l$rVCy{b2_YNlRWMOj5I$<4jL<(%*Ne9y(MJmesllM^=*2T-KRtbm%$a6-@q
z6sourU!^JfBEfIqU0>a91WDA+DBOJ~Od=%8%@wUFXL9yH^(^X8*Q<<Ak=UF9x}ra(
z3Pd!ba}x-V*~NV*vqD+s8U%D=Uq!p`$4>)@#6`#+*Hz*xkn|$W!GNAWZD_xMkft(t
zx@mvU&~yZ71n?$eSei>wkHkvLNQoKsp-4#7+X}-I-00sy%O1#9K2VB})@Ln61i$TB
zv@Fk6sTha*FViMyt3tI~mzS$B{3XrQxHnDGA{Fyfpj06(GR1RZpcK0PR&<^a<NOo&
z@ZaYr|4dJKo^MY0KUqglyHuwCzK;GY*HP`_NGQ0dr{OaK{y(j$?2}BG^8^wO8BCT(
z2<&WAEeU+eT7ByB6D<%JV;*Y-OO0#!@|e=kwN*;X3Wyr8N>&VYgnQ5f?%X?KE^XWN
zxdiyzCivA_REDthh)Dq;XQi@y;t-8hQ$1X?AWJ7d!%CVJbcy=Nu%zl9sKhp7=KaZ5
zp(zvI6Ini_50@PX!nK{W+xll?@Y#@`q5WzC*n*-^RNDpm*nzda&CZ{)E5V^|guG78
zq+Wv+s~uMt>@^2_VLQfrt=ZEN)0_}QH@QO8i7CR`NIm%0iNvYj)i~{3*$eTP31|Z=
z52)l1g$r;%xbW!^8)hS1F@)=F=yYLmr#f-AM|hClA=6DLK<@w}n;;}Q7^3WG?gs|Y
zc^M|e@k{T%(Ioa<+hub<{VEP<iKsRx@voS05um<+CYh19^M%)-u8@6hCGie3PMi*T
zRu74B|CO{5ZFCuPq0#{>^J?|g2e#nG4q8PbvbBsU(-utRJmt_}F8HjR3A@~ND8l4;
z_&~k(<ArHEPl%bM^A51r1IEBRcbK>`uYmFNg`(ugATm$2Z70NiTsSeDMm`7BCj>4_
zbh(oI-F3uY?~}objA{aqK~aa!Zf45GtWKKKG$MG?IPwwo>8>I^l$91wyppcONF44b
zKbSx+3KcdbbEl(E!xh|Pzo;M-jS@s4dqO&FxZ9I$>r?f1fWScq^g84ju{mt~CUlYD
zkT;ZNA$}h@s8T(TI7mysYKGp1ZZJFUb1m-dCLDk}0)|-3642ZTr6d-dD}RTe0poUb
zWk7E3{XFkfaA3J{o)F(|{i3dzlQQ(DK+HCSn>)@=LP7<sTJ9y6|77&xcC^$?m03*s
zg}w9yNE^>JsdPZbKo3Dry{!1R*d1S0zChAZ({hlQhUrQy5sHmlmLAwk-1nSJ@z;LD
zrw(uYc;){$z~8^*O*Xul+6+>TJLCd@c2~&-GEk1=K>A*c!7KQJ8E86yDlpsvkE$`6
zTOfX!H>Z-?BQ@WQR|9#hZr}#M>_}*f9BBUX7O3+lI_u_Aoj?q{s_SwCX$G{9dwK>c
zTl$L2h~W^j$}Z$*C*7D#$bActRRcrYLnsBxD5ighp146=lMF>Tq(rg)kN6Kh+M^BZ
z;G>0y3B8M&w@mPc4|?{19?&8zif`-F2<#=G&ryusgQyfb7?c-=PeKFYJ1nfFX8+Go
z5qT_u)~sKpag5oUQ-o0u_>|ip@On@LgLNzxG$=6?h|9pP-jDftS}h9@?PxNrLQNlB
zpuH1@)#Q;PJ$8oVSG@aSleQLd-|#z9QXM(Ez!iX{NFLvcqA~gd5XM2HJyZ|TOo%yE
z*Xv9MIEvfR%!SE2Kx&j8I>`Qr@0+96)a_S>bkXUr=eOhK$Cz8`uewi#yZb(uNFkw>
zu5U-bEyi|zMf2T($-Xa|38oW-V_<TWK?NDJNt$wi*3udbinnD{8&qYthlY;p79U;C
zhHy3C>b26^!omX1ub^rm;+u7e0K#AnEDPSL8WgQz%T6@BRawjdlW%w@NAlq}ZLi+8
z*9SsE^ZuV^h91d1nA61Lf!xZE;F<AwkCB9gnOiTB6Ej4j^|>Sv@?%H@TC{^?f<~9P
z8{U}17l7f%+c6VVzaZIUNqlwnumnGvRBw4J50R!UMZx`&rZ^Nhf^rK%ZW*+cAU!#2
z@z&eO&^-h=p7Nju%HjNqF;}=QNT_!Yj)7WX1tDrC6y?VsY2iEeDt7n!xXW#FB2D_Y
z-IVaXeSDffyv~W{H9Vld0&Ow}?ilC>U50@Tc3;65Yw9eM+<{#X@TWFQ5dac-rq>B^
zCCh4Rb2m^AcR*o)P`(_5t176IQz~H3d=AMAAT#I_03B;U*}!X9l-lw5Qmc?xnwdJ;
zfqeUE(v!;V3s-Lp*`{$TJuvawhCOE&7;i1E{Bp}JkF7hmbf3F*X4xI;<wN?%vonq-
zzO}JAYG66-%(jd5HzE$ER&L2y`Nxyz)i(LZ@IxyfEk5bS+;21{V^h<ljVw{VuuqdN
zij~8_It+wX!^J$Lq@vm`o6h*4Bv9&vUk&Xj`y6(s*1W8(y-J4J_j)Z2RA3pl?L>1u
zE3^snGcY9Rxfay0VD%OOnqvSPp9VCP9N$BHf5T<rG8Z)dl2Pe=2zhdRu^w_nP;h$&
zy(WNtxr*6wU0p161Hs0%SQiKJi@w3w*e0zGRY_;RXWHgbFqn7_v?s(i*Sj|I$cdQR
z+FIx1MRU=t*e>D<YvkAj6_Jc->M!L=f1pH>BEES+_tBE>?rslX-<G%Is8Q>Hn<CJc
zSUwhBDCA9tJjIg5Vqr1C1VCY|(x7*%HIha}ISsT?c?d=*>(Lw319J2j*xW@hnFyR7
z4}mO)HGawmM-Iu?E3Z(_rk#^o0x9l*o);*Ebf-&vyuHg|He(wTc3ewCHy*q3U@trg
z1^$I?{;B}~C!PMIius<Zt!ipMXS=Ghv*gc(g#Ki`BwR&&TYsj9$8Y>;^6zi(`!)D|
z9Q=pQg}Z;ULxVwrqp!bD360qpb;qhD?vm26b|AR4w{q@-h3i=t8x78~-$xsL=w;}S
zgeBXF)WW*|3s(ojbg@0N&Fg|yMM|Chq37XMe#7kI_W61Wa?-bbSGrdFiBiPf$A*F_
zUhD2v+!E0)4+v>1y=PmN-_dD~nq_KcN109<{T|nsBi7676yMPwjSpXJ8aEv|%ixxU
zAEen2-HuD-HOFadcIPY3iVxd}pO$fh=$^@Tl&?vnWkk=ah1Fx6>sNk!99a67s+hI5
zjXR?f9c>h2y_t+6a}JkFm7g-{t=6=&`Yen+3s<&<FD|5+L>JSCcL`2(*^Lz5&9D<C
zmt~pKV$9`QF~(ZFjE<$G7kaq3ZJab(bb)3Llt9VF^D(v(gDKqDBzmw3T&cUkQ%Qr@
zaYcL)=E@GW3)EWu-O9&i4wO57ty?8GgFE)}er4KOsfAHz6CY~LceP^JFlr0}zbfCU
zl*wC%f|*4Yqy))7E}jc_58<jB7^)`}EHm(FOsnH6Yt>8pGiV#2wDU`;^15D#`vDQF
zR5#QhyE<osSTyfa>FByzXwpm_FE;9W480NV?G@_lwD=dY*bevYOPg6;(T`>F9aPQ}
z`n$r6*}?qTr0qI*`(7+}X^P`3kMR|kXr25Gxn2x+k+>t&g!8Umoaxs+u=izUe2_uU
zwuT=>GWK%PoU+B37cH=TkSEJ3$T_~V#(=ah>DUz9212DJhlglmKdHU2^{VL4cX2AH
zb!mwe+bi|Ojls4X)(NkW_l^!WU0>~CKX<m;z9c{5uZoVOc#n%-B|0)@DBkG!_1`w4
zq*0XpoUXFHBx`ltXTmn^AHF<`!im)4a{^3B`*ixx;r=_?b<OJxU(DjH9<j*p{o%B+
zw*`}LtS9K-F^|T!5qWW`@1Mj;$4S=O#G6n%J9<JyvPYdarwp=M)MI#^O-`XLdyAdp
zL!D-Rn?9YCz8uCzm|DQAvr7F|GcDJlX$#BV?~>b7!oR38F&rtmnBp*49-kU*vTNwP
zB1o8LGM;Y0sK^)?4QL9r+9qvD5~c<U_G{e@R*SeB^!9c&b-kcN-S7s*Db+RI0!BcX
zTDIug;)n2)60f!@!L$6?D=S<6UwpSlSncjqjzl`!x9oRnwiik10}6WcRXD4P`Ii)q
zTu~Eav%Yk314ZqN?0Tv^F-S|fcy6FJvGY#qh@-Kul5g^O`nNsa&r7SZ5j|a3#}^y<
zdZny!A8CBBvVzrmzmC!H*g-7imRm)IbQaS^9%k%-R0%2U=ex-r80${+(OR%3@IvdF
zPJ5rB>jAz6H+q7ZqP|y@obh`<B-US5Uc7(j>aNFuq7P>Dhee%U>&6}k8EOejRHxvX
z`zz~^ggLq4^i1dUB8ES&^Cn~9)O?vnq{bsBX`{EZmhj-<ykg$~{%cV-$$DNS>Z}Q(
zOZjfE<N})aFXH|>ou+0+h$5?#F+xfD&YV5kHM2jdF4yh3TGh)&Lvp0B_tw3ZtQ>J>
zW~4AM)i^+T*|EE6=A7m}(U)I@x%`MTa3hyLFr(4?bz@U@Q6Em^D_)#vU%j}S<k!Qs
zqhoCjGM|9vWU~<Mq#Z_SB9>?J;Ufd-{yywZWtuXWo2Ew6Rt>7a^9Ixaj~)u=`-<V}
zJ4S=1+_D$#`_*3>8cvD0d9CWwn5W`+MVjBHhL*&u%$A4-W9}<uUc9h@v%%M_98Ttj
zF;4a2Q&LZLu9<FwqWIg8V}tDL0gW~I!>77SDAEr#Ol6aplSio{C2Z^Y$H<DfJV!RJ
zYx1{oWrcdf{0i%Cn*rff*cXPNX{G3N>BCflM>9f?q>rlyyauTz`I)44+^WbTCHL*S
zB!dIP=OD9sfGe51Y9#}X!~z)kt7HAODM_{|sXPPaXtfN=eTK~5&#cKfO%l`~snP|l
z#Zn;4^y(i>;iLvzYIh`yjc2;V&#xO>Mi&lt;^P0?LRyWjy-XFumt&RVtqh5f!xQh#
zir6>wNLf+98Qtuod_CqI^zI6#7OC-*N#pvri4kXe3o0)bTBFA+#^2_T+~m2rSH8S3
z@FQPIUVNs@eke<=um0(pAZf0q`kH&%xGDMYvfH4!k{d*_c0-1w_v?t04UYi+iK#0v
z(A3n#{2l9Ahk$Ok6D|CMsjK=Gcg{_Xkq=e`4-9p_s``4zp%tW8y2QXmdl7Uk*fTVZ
zln`lh>Tdz9WKq_EqVw349Rw_vFl|>_9d=)92XuIcySsRXGZC_PE%~-`Rf)+^was<t
z;Q&RFKO=7i#f@!{opRHV1aE^*X(IGpp<^5TD+|K1Vc?SrrN^d778yWM5;BgZ<_|Ho
zfEYbg@DwNu$T4CDHMqy0sv93AQo!`qFAGlXfE64IIF9!r0fMR;0tQr<gUH2$=P{}c
zDvi%QlHzh6GYQ$J!Rn3X4~Vb>bhn&90op;U=ordJ@CNQz9b>FQ_3YU$$H~o`I>6~#
z0UaUY$@x%u!eX*(91|0C!X|)rY2cOA(<4+TKkLMlWZ-?!)h{s;Ku?0InwlYjc!g8W
z3Sg-$qp)HQVwU=gPseBIK?j===)He~90WF3WCdju8am+6^h_zn<Q#zIDItsTBVKSm
zlPp>b`P+&;=?UUtXs%NVmD=2q*HWp}-`^kn5<qAn3T1rw)A6#`BKBHvz?ny(e+LB3
zh7JRw+61Mb_*QVX9eQ$Zma(_$4`}cK4yiTxI0Y#E>zMlpM%u@q%~kVT1N038avn{-
z4tXH91Beq4u>E`E@LV-ix)IBY@(&1j2VRwqAYUlcAwL&I2HzV}Bc9og+Qzt~h6$2u
z;1QfPix-?g9|n3nuqx<OKm}7xJdX(-(xSlvWVagQF4Frc;kn^hZ50EPy|(2W1GgPW
zt!3rp^t9RG0UJ3%TTCYz0U3e=$6LSzMeIBeU$I`m>|J{e*!@*-yBgkRgOAY=W&_&L
zzE8yi!W74*_@2V=1FB8V32$I#-Vzx48~{$IzsFt#Xo3bS^I_;mfei|3qKUs_|KY?p
zwq|B#2zA5)Qir5M4dZVw;g=*$eTPrEgn$2r31bIGEnSblU|uX~5YN#3d&?~kHt#$6
ESBxke<^TWy

diff --git a/docs/reference/read_model.html b/docs/reference/read_model.html
index 913e8c840..04b22a941 100644
--- a/docs/reference/read_model.html
+++ b/docs/reference/read_model.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/region.html b/docs/reference/region.html
index 78017f629..675338335 100644
--- a/docs/reference/region.html
+++ b/docs/reference/region.html
@@ -18,7 +18,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -156,7 +156,7 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-in"><span></span></span>
 <span class="r-in"><span><span class="va">real_map</span> <span class="op">&lt;-</span> <span class="fu"><a href="world.html">world</a></span><span class="op">(</span>xrange <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="op">-</span><span class="fl">15</span>, <span class="fl">40</span><span class="op">)</span>, yrange <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">30</span>, <span class="fl">60</span><span class="op">)</span>, crs <span class="op">=</span> <span class="st">"EPSG:3035"</span><span class="op">)</span></span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> Reading layer `ne_110m_land' from data source </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpFhhY5X/naturalearth/ne_110m_land.shp' </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpqG5m0G/naturalearth/ne_110m_land.shp' </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>   using driver `ESRI Shapefile'</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> Simple feature collection with 127 features and 3 fields</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> Geometry type: POLYGON</span>
diff --git a/docs/reference/reproject.html b/docs/reference/reproject.html
index 247854bdb..dbdebb20b 100644
--- a/docs/reference/reproject.html
+++ b/docs/reference/reproject.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/resize.html b/docs/reference/resize.html
index 797335847..e8550af3b 100644
--- a/docs/reference/resize.html
+++ b/docs/reference/resize.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/schedule_sampling.html b/docs/reference/schedule_sampling.html
index d3bebeecd..b66d6281d 100644
--- a/docs/reference/schedule_sampling.html
+++ b/docs/reference/schedule_sampling.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/set_dispersal.html b/docs/reference/set_dispersal.html
index c9eb77cb9..3cb3b7e40 100644
--- a/docs/reference/set_dispersal.html
+++ b/docs/reference/set_dispersal.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/set_range.html b/docs/reference/set_range.html
index b986a01ca..64f927d35 100644
--- a/docs/reference/set_range.html
+++ b/docs/reference/set_range.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/setup_env.html b/docs/reference/setup_env.html
index 39c4a24b2..8efae561a 100644
--- a/docs/reference/setup_env.html
+++ b/docs/reference/setup_env.html
@@ -14,7 +14,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/shrink_range.html b/docs/reference/shrink_range.html
index 43b2f193c..c7ea8bda3 100644
--- a/docs/reference/shrink_range.html
+++ b/docs/reference/shrink_range.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/slendr.html b/docs/reference/slendr.html
index 5c8b589b9..7d2ecf260 100644
--- a/docs/reference/slendr.html
+++ b/docs/reference/slendr.html
@@ -40,7 +40,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/slim.html b/docs/reference/slim.html
index 5f8b4e742..cfb4200ed 100644
--- a/docs/reference/slim.html
+++ b/docs/reference/slim.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/subtract.html b/docs/reference/subtract.html
index 28bb8373b..a1d4c8745 100644
--- a/docs/reference/subtract.html
+++ b/docs/reference/subtract.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/summary.slendr_nodes.html b/docs/reference/summary.slendr_nodes.html
index 2903fcfe1..ed297212a 100644
--- a/docs/reference/summary.slendr_nodes.html
+++ b/docs/reference/summary.slendr_nodes.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_afs.html b/docs/reference/ts_afs.html
index 7ecb010a8..69580c190 100644
--- a/docs/reference/ts_afs.html
+++ b/docs/reference/ts_afs.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_ancestors.html b/docs/reference/ts_ancestors.html
index c22a67fe5..36b9fae3b 100644
--- a/docs/reference/ts_ancestors.html
+++ b/docs/reference/ts_ancestors.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_coalesced.html b/docs/reference/ts_coalesced.html
index ef61b9c15..332c57155 100644
--- a/docs/reference/ts_coalesced.html
+++ b/docs/reference/ts_coalesced.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_descendants.html b/docs/reference/ts_descendants.html
index a6de13b2d..41b27e944 100644
--- a/docs/reference/ts_descendants.html
+++ b/docs/reference/ts_descendants.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_divergence.html b/docs/reference/ts_divergence.html
index b6bbcdecd..c8588696a 100644
--- a/docs/reference/ts_divergence.html
+++ b/docs/reference/ts_divergence.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_diversity.html b/docs/reference/ts_diversity.html
index b071ef2ac..cc32d8832 100644
--- a/docs/reference/ts_diversity.html
+++ b/docs/reference/ts_diversity.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_draw.html b/docs/reference/ts_draw.html
index 3c6637331..24646c301 100644
--- a/docs/reference/ts_draw.html
+++ b/docs/reference/ts_draw.html
@@ -16,7 +16,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_edges.html b/docs/reference/ts_edges.html
index 29d920f22..a1cdf1d75 100644
--- a/docs/reference/ts_edges.html
+++ b/docs/reference/ts_edges.html
@@ -14,7 +14,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_eigenstrat.html b/docs/reference/ts_eigenstrat.html
index add549297..361c0129c 100644
--- a/docs/reference/ts_eigenstrat.html
+++ b/docs/reference/ts_eigenstrat.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_f4ratio.html b/docs/reference/ts_f4ratio.html
index 82eb73360..f7afbea1a 100644
--- a/docs/reference/ts_f4ratio.html
+++ b/docs/reference/ts_f4ratio.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_fst.html b/docs/reference/ts_fst.html
index 0697f2b16..40d4256a6 100644
--- a/docs/reference/ts_fst.html
+++ b/docs/reference/ts_fst.html
@@ -14,7 +14,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_genotypes.html b/docs/reference/ts_genotypes.html
index 88d94d272..da181b773 100644
--- a/docs/reference/ts_genotypes.html
+++ b/docs/reference/ts_genotypes.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_ibd.html b/docs/reference/ts_ibd.html
index 42d2ba9e4..8179fc0b2 100644
--- a/docs/reference/ts_ibd.html
+++ b/docs/reference/ts_ibd.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_load.html b/docs/reference/ts_load.html
index 2cf0a7677..3e27957ec 100644
--- a/docs/reference/ts_load.html
+++ b/docs/reference/ts_load.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_metadata.html b/docs/reference/ts_metadata.html
index d55c28856..66bc58af4 100644
--- a/docs/reference/ts_metadata.html
+++ b/docs/reference/ts_metadata.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_mutate.html b/docs/reference/ts_mutate.html
index a53399544..edea31af6 100644
--- a/docs/reference/ts_mutate.html
+++ b/docs/reference/ts_mutate.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_names.html b/docs/reference/ts_names.html
index 4a5181369..8c66b5e66 100644
--- a/docs/reference/ts_names.html
+++ b/docs/reference/ts_names.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_nodes.html b/docs/reference/ts_nodes.html
index b204b0398..b954386ef 100644
--- a/docs/reference/ts_nodes.html
+++ b/docs/reference/ts_nodes.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_phylo.html b/docs/reference/ts_phylo.html
index 53e6eeaeb..5023dcc12 100644
--- a/docs/reference/ts_phylo.html
+++ b/docs/reference/ts_phylo.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_recapitate.html b/docs/reference/ts_recapitate.html
index 96a0c71b5..32794f761 100644
--- a/docs/reference/ts_recapitate.html
+++ b/docs/reference/ts_recapitate.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_samples.html b/docs/reference/ts_samples.html
index 769896043..f75f61255 100644
--- a/docs/reference/ts_samples.html
+++ b/docs/reference/ts_samples.html
@@ -14,7 +14,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_save.html b/docs/reference/ts_save.html
index 78f0a3780..28a17e237 100644
--- a/docs/reference/ts_save.html
+++ b/docs/reference/ts_save.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_segregating.html b/docs/reference/ts_segregating.html
index 203eaf4ac..49c0129b5 100644
--- a/docs/reference/ts_segregating.html
+++ b/docs/reference/ts_segregating.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_simplify.html b/docs/reference/ts_simplify.html
index a6e49e0f2..cddded357 100644
--- a/docs/reference/ts_simplify.html
+++ b/docs/reference/ts_simplify.html
@@ -14,7 +14,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_table.html b/docs/reference/ts_table.html
index a4b150a6b..77b9c9cf7 100644
--- a/docs/reference/ts_table.html
+++ b/docs/reference/ts_table.html
@@ -14,7 +14,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_tajima.html b/docs/reference/ts_tajima.html
index 61cdc31f3..2752db4c8 100644
--- a/docs/reference/ts_tajima.html
+++ b/docs/reference/ts_tajima.html
@@ -14,7 +14,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_tree.html b/docs/reference/ts_tree.html
index 5b7f86467..f71df37ee 100644
--- a/docs/reference/ts_tree.html
+++ b/docs/reference/ts_tree.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/ts_vcf.html b/docs/reference/ts_vcf.html
index fd33e7370..157e948c4 100644
--- a/docs/reference/ts_vcf.html
+++ b/docs/reference/ts_vcf.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/docs/reference/world.html b/docs/reference/world.html
index 4b1f52886..ae4c42843 100644
--- a/docs/reference/world.html
+++ b/docs/reference/world.html
@@ -14,7 +14,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">slendr</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">0.9.1</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -167,7 +167,7 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-in"><span></span></span>
 <span class="r-in"><span><span class="va">real_map</span> <span class="op">&lt;-</span> <span class="fu">world</span><span class="op">(</span>xrange <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="op">-</span><span class="fl">15</span>, <span class="fl">40</span><span class="op">)</span>, yrange <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">30</span>, <span class="fl">60</span><span class="op">)</span>, crs <span class="op">=</span> <span class="st">"EPSG:3035"</span><span class="op">)</span></span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> Reading layer `ne_110m_land' from data source </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpFhhY5X/naturalearth/ne_110m_land.shp' </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpqG5m0G/naturalearth/ne_110m_land.shp' </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>   using driver `ESRI Shapefile'</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> Simple feature collection with 127 features and 3 fields</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> Geometry type: POLYGON</span>
diff --git a/docs/search.json b/docs/search.json
index daa194275..56a4e2082 100644
--- a/docs/search.json
+++ b/docs/search.json
@@ -1 +1 @@
-[{"path":"https://www.slendr.net/LICENSE.html","id":null,"dir":"","previous_headings":"","what":"MIT License","title":"MIT License","text":"Copyright (c) 2022 Martin Petr Permission hereby granted, free charge, person obtaining copy software associated documentation files (“Software”), deal Software without restriction, including without limitation rights use, copy, modify, merge, publish, distribute, sublicense, /sell copies Software, permit persons Software furnished , subject following conditions: copyright notice permission notice shall included copies substantial portions Software. SOFTWARE PROVIDED “”, WITHOUT WARRANTY KIND, EXPRESS IMPLIED, INCLUDING LIMITED WARRANTIES MERCHANTABILITY, FITNESS PARTICULAR PURPOSE NONINFRINGEMENT. EVENT SHALL AUTHORS COPYRIGHT HOLDERS LIABLE CLAIM, DAMAGES LIABILITY, WHETHER ACTION CONTRACT, TORT OTHERWISE, ARISING , CONNECTION SOFTWARE USE DEALINGS SOFTWARE.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"quick-installation","dir":"Articles","previous_headings":"","what":"Quick installation","title":"Installation instructions","text":"slendr available CRAN R package repository. , can install simply executing install.packages(\"slendr\") R console. want (need) get development version, can install directly GitHub executing devtools::install_github(\"bodkan/slendr\") via R package devtools (can gen devtools running install.packages(\"devtools\")). fact, decide try slendr, please make sure update regularly keep eye changelog regular basis! can find information latest bugfixes potential breaking changes. install slendr, calling library(slendr) check software dependencies available. , R package provide brief helpful guide resolve potential issues. rest vignette talks necessary software dependencies bit detail. Please note slendr extensively tested macOS Linux moment. However, experimental support runnig SLiM msprime simulations Windows analyzing tree-sequence outputs using tskit interface platform well. Feedback issues using slendr Windows highly appreciated.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"software-dependencies","dir":"Articles","previous_headings":"","what":"Software dependencies","title":"Installation instructions","text":"slendr relies three main software dependencies: geospatial data analysis R package sf (encoding spatial slendr models analysing spatial tree-sequence data), forward population genetic simulator SLiM (forward simulations), Python modules tskit, msprime, pyslim (coalescent simulations tree-sequence analysis), also pandas used internally simulation back ends. three widely used respective fields , , easily obtainable major operating systems (see information troubleshoot potential problems). Note depending use case, three sets dependencies necessarily needed. ’re going running forward spatial simulations, don’t need SLiM geospatial R packages sf, stars, rnaturalearth. applies also animation spatial models using gganimate R package. slendr install dependencies. need functionality, install respective R packages manually. vignette, briefly explain get slendr’s software dependencies installed. said, note normal circumstances (exception SLiM), manual installation individual dependencies required.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"sf-stars-rnaturalearth","dir":"Articles","previous_headings":"Software dependencies","what":"sf, stars, rnaturalearth","title":"Installation instructions","text":"R package sf heart geospatial data analysis R. available CRAN can installed major platforms executing install.packages(\"sf\") R session. applies stars rnaturalearth. first load slendr via library(slendr), ’re missing three geospatial R packages, notified instructed can easily obtain CRAN using single call install.packages(). said, sf depends number geospatial libraries depending exact setup Linux macOS machine, libraries missing. Luckily, easy install via Homebrew (macOS) via appropriate package manager Linux distribution (Ubuntu, Fedora, etc.). Detailed instructions operating system can found . ’re problems installation three packages, look help .","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"macos","dir":"Articles","previous_headings":"Software dependencies > sf, stars, rnaturalearth","what":"macOS","title":"Installation instructions","text":"One user recently installed slendr fresh macOS system reported needed install libgit2 order able install package devtools devtools::install_github(\"bodkan/slendr\") step described top page. Additionally, install couple C/C++ libraries well (dependencies sf package). end, able successfully install slendr running: Note assumes Homebrew package manager already setup Mac. ’re beginning computational scientist using Mac, strongly encourage install Homebrew. Sooner later need specific Linux/unix program anyway, Homebrew way get (Mac unix machine, without Homebrew poor one default).","code":"brew install libgit2 udunits gdal proj"},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"linux","dir":"Articles","previous_headings":"Software dependencies > sf, stars, rnaturalearth","what":"Linux","title":"Installation instructions","text":"Testing slendr installation fresh, pristine Debian installation dependencies previously installed, run following: ’s unlikely need (might need packages non-Debian distributions), got slendr dependencies running completely clean system. Might good start case trouble Linux machine. Windows special treatment necessary get slendr running. install slendr via install.packages(\"slendr\"), get binary version package dependencies without need compiling sources.","code":"sudo apt-get install libudunits2-dev libssl-dev libgdal-dev libgsl-dev libgit2-dev libfontconfig1-dev libharfbuzz-dev libfribidi-dev"},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"slim","dir":"Articles","previous_headings":"Software dependencies","what":"SLiM","title":"Installation instructions","text":"forward population genetic software SLiM available major software platforms. complete installation instructions can found . Mac, recommend installing SLiM via pkg installer available direct download website. Linux, can either install SLiM via appropriate package manager Linux distribution (see SLiM manual information), can easily compile . install SLiM Windows, please follow instructions section 2.3.1 SLiM manual. Note although SLiM also available conda, comes without SLiMgui! Note slendr requires SLiM 4.0 work earlier version. , running library(slendr) inform potential issues SLiM installation.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"adding-slim-to-the-path","dir":"Articles","previous_headings":"Software dependencies > SLiM","what":"Adding SLiM to the $PATH","title":"Installation instructions","text":"order slendr able find SLiM installation, R must able find slim binary (slim.exe case Windows) called $PATH variable. easiest way verify true call Sys.(\"slim\") (Sys.(\"slim.exe\") Windows) R session. instance, Mac, get : Windows, get (followed section 2.3.1 SLiM manual describing “official” way install SLiM recommend follow well): , hand, might get something like — empty string: means $PATH R configured properly R (slendr) won’t able find SLiM $PATH. add SLiM $PATH ? Probably convenient way editing .Renviron file. precise location file depends operation system can automatically get open text editor using command (might install.packages(\"usethis\") first): , can either add following (note ’s $ line!): Alternatively, already $PATH contents specified R session, can get calling Sys.getenv(\"PATH\") R console, grab entire string get way, append path installed SLiM using appropriate delimiter (: Linux/macOS, ; Windows) string. case, edit might look something like : instance, Mac, .Renviron file contains line (note last item /usr/local/bin matches path slim showed ): testing Windows machine, .Renviron file (, look path C:/msys64/mingw64/bin;$PATH\" matches slim.exe binary shown ): important check R can find SLiM using Sys.() command described . gives positive result, won’t able use slendr’s spatial SLiM simulations. ’re struggling , search advice related .Renviron $PATH online.","code":"> Sys.which(\"slim\")                  slim  \"/usr/local/bin/slim\" > Sys.which(\"slim.exe\")                             slim.exe  \"C:\\\\msys64\\\\mingw64\\\\bin\\\\slim.exe\" > Sys.which(\"slim\") slim    \"\" usethis::edit_r_environ() PATH=\"<path to directory with SLiM binaries>\" PATH=\"<original Sys.getenv(\"PATH\") content><delimiter><path to directory with SLiM binaries>\" PATH=\"/opt/homebrew/bin:/opt/homebrew/sbin:/Users/mp/.my_local/bin:/Users/mp/.my_local/AdmixTools/bin:/Library/TeX/texbin:/opt/homebrew/opt/gnu-sed/libexec/gnubin:/opt/homebrew/opt/coreutils/libexec/gnubin:/usr/local/bin\" PATH=\"C:\\\\msys64\\\\usr\\\\bin;C:\\\\rtools43\\\\x86_64-w64-mingw32.static.posix\\\\bin;C:\\\\rtools43\\\\usr\\\\bin;C:\\\\Program Files (x86)\\\\R\\\\R-4.3.2\\\\bin\\\\x64;C:/msys64/mingw64/bin\""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"fallback-options","dir":"Articles","previous_headings":"Software dependencies > SLiM > Adding SLiM to the $PATH","what":"Fallback options","title":"Installation instructions","text":"don’t want deal editing .Renviron file, able set path SLiM using command Sys.setenv(PATH = \"<path directory SLiM binaries>\"). beginning slendr R scripts though. Alternatively, slim() function argument slim_path= can specify full path slim slim.exe binaries directly (directory ’s case $PATH! full path slim/slim.exe binary ).","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"python","dir":"Articles","previous_headings":"Software dependencies","what":"Python","title":"Installation instructions","text":"order able run coalescent simulations process tree-sequence files, slendr needs Python modules tskit, msprime, pyslim (also needs pandas library). Setting isolated Python environment specific version Python packages (important avoid clashes among different Python programs needed system) can bit hassle users. especially true R users might use Python daily work. order make sure R package appropriate version Python available, correct versions Python module dependencies, slendr provides dedicated function setup_env() automatically downloads completely separate Python distribution installs required versions tskit, msprime, pyslim modules dedicated virtual environment. Moreover, Python installation virtual environment entirely isolated Python configurations already present user’s system, avoiding potential conflicts versions Python Python modules required slendr. Next time call library(slendr), need activate environment automatically calling init_env(). ’re comfortable Python don’t need worry beyond calling setup_env() init_env(), interaction Python necessary working slendr R.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"failing-conda","dir":"Articles","previous_headings":"Software dependencies > Python","what":"Failing conda?","title":"Installation instructions","text":"order support Windows, slendr uses conda download Python interpreter explained . Given fact, run setup_env(), slendr tries leverage conda present install Python dependencies (msprime, tskit, pyslim, pandas) via conda . Unfortunately, conda can break frustratingly many random reasons completely trips setup_env(). run issue, fallback option install Python dependencies msprime, tskit, pyslim, pandas via pip , unlike conda, works practically every time. setup_env() fails conda-related reason, first clear broken environment calling clear_env(), restart R session, call setup_env(pip = TRUE) instead default setup_env(). Note might require install GSL numerical library, ’s trivial issue macOS (brew install gsl) Linux (Ubuntu, instance, sudo apt-get install libgsl-dev).","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"information-for-python-experts","dir":"Articles","previous_headings":"Software dependencies > Python","what":"Information for Python experts","title":"Installation instructions","text":"case wondering slendr accomplish : slendr’s Python interface implemented using R package reticulate. embeds Python interpreter inside R session, enabling high-performance interoperability languages without need user intervention.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"docker","dir":"Articles","previous_headings":"","what":"Docker","title":"Installation instructions","text":"currently official Docker image slendr one soon. current plan use geospatial image published Rocker project (already contains pre-compiled R, RStudio, necessary R package dependencies sf ) extend slendr SLiM. Stay tuned!","code":""},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"motivation","dir":"Articles","previous_headings":"","what":"Motivation","title":"Introduction and basic tutorial","text":"motivation starting project create programmable simulation framework add explicit spatial dimension population genetics models. Specifically, original idea able take models one —  representing simplified view history anatomically modern humans (AMH) West Eurasia last ~50 thousand years (comprehensive overview can found review Lazaridis)—design tool makes possible simulate models explicit geographical context capture processes similar following figure (taken study Haak et al. 2015):  reason probably clear. lot studying history humans species focused reconstructing population movements, expansions gene flow events, happen geographic context. fact, geographic component often interested (.e., “ancestors population come ?”, “route fast migrate?”, etc.). However, goes beyond just simulating demographic history. instance, selection pressure driving adaptation can often spatially heterogeneous: members population occupying one part continent exposed different environmental pressure individuals elsewhere, allele frequency distributions shaped adaptation process reflect spatial heterogeneity accordingly. framework enables simulation explicitly spatial genomic data situations allow us build realistic models test specific hypotheses, goals simply possible using non-spatial simulation methods. R package slendr introduced vignette presents framework. Internally, package two independent tightly interconnected units: R interface provides set functional primitives (“mini-language” sorts) encoding various features spatio-temporal models: population migrations, expansions gene flow, happening real geographic landscape defined freely available cartographic data. Populations represented simple R objects easily visualized spatial boundaries, making possible build complex models interactively set small simple building blocks. SLiM simulation back end represented built-generic SLiM script designed read spatio-temporal model configuration parameters objects established step 1. , tailor simulation run user-defined model. Alternatively, slendr also supports executing standard population genetics models random-mating setting. means models need explicit geographic map can simulated either built-SLiM back end script, efficient msprime back end also provided package. important design objective make integration parts 1. 2. appear completely seamless. Even extremely complex models, model building execution (.e., simulation) can performed without leaving convenience R interface RStudio. simulation complexities happen automatically hood knowledge SLiM required. fact, motto slendr package “Write complex spatiotemporal population genetics models simple R script.”","code":""},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"geospatial-data","dir":"Articles","previous_headings":"","what":"Geospatial data","title":"Introduction and basic tutorial","text":"Geospatial analysis deep complex topic, dozens libraries programs designed deal fact Earth three-dimensional object forced plot geographical objects (, case, simulate data) two-dimensional plane. Luckily, technical issues Coordinate Reference Systems, transformations manipulation geometric objects (shifting population boundaries, expansions, etc.) pretty much solved now. Unfortunately, dealing issues practice quite challenging requires non-trivial degree domain expertise. Programming even simple task geospatial data analysis also often requires lot code. R package designed provide collection primitives (“mini-language” sorts) programming population dynamics (splits, movement, gene flow, expansion spatial boundaries) across space time without explicitly deal challenges inherent geospatial analyses.","code":""},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"installation-and-setup","dir":"Articles","previous_headings":"","what":"Installation and setup","title":"Introduction and basic tutorial","text":"slendr R package available CRAN can installed simply running install.packages(\"slendr\"). need run latest development version (instance, need latest bugfixes), can get via R package devtools executing devtools::install_github(\"bodkan/slendr\") R terminal. can find detailed installation instructions vignette. get slendr installed, just need load : dependencies (SLiM necessary Python modules) missing, get informative message proceed.","code":"library(slendr)  # activate the internal Python environment needed for simulation and # tree-sequence processing init_env() #> The interface to all required Python modules has been activated."},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"defining-the-overall-world-map","dir":"Articles","previous_headings":"","what":"Defining the overall world map","title":"Introduction and basic tutorial","text":"anything else, need define section map world provide context downstream spatio-temporal manipulation population ranges. principle, source geospatial data can manipulated using simple features (sf) infrastructure used. now slendr package implicitly uses Natural Earth project data (vectorized form!), internally loads using rnaturalearth interface. first slendr function look map(). function load map entire world vectorized format zoom specified section world. Note call , specify coordinates zoom geographical Coordinate Reference System (CRS), longitude/latitude, also specified want perform downstream manipulation spatial population maps projected CRS (Lambert Azimuthal Equal-Area projection) appropriate representing wider European continent used tutorial. course, different CRS projections used based part world want simulate. Describing intricacies coordinate reference systems beyond scope tutorial, ’re interested learning encourage read freely available textbook dedicated topic. approach slendr: let user specify everything easy--understand longitude/latitude geographical CRS (can read map, making easy define spatial boundaries trajectories movement), internal data structures final exported spatial maps internally handled projected CRS, important make sure distances proportions distorted. Internally, map object currently normal sf class object without additional components. unlike slendr objects described , also sf objects carry additional internal components. Note summary object says “projected CRS: ETRS89-extended / LAEA Europe”. means world map indeed transformed projected CRS specified .","code":"map <- world(   xrange = c(-13, 70), # min-max longitude   yrange = c(18, 65),  # min-max latitude   crs = \"EPSG:3035\"    # coordinate reference system (CRS) for West Eurasia ) map #> slendr 'map' object  #> -------------------  #> map: internal coordinate reference system EPSG 3035  #> spatial limits (in degrees longitude and latitude): #>   - vertical -13 ... 70 #>   - horizontal 18 ... 65"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"plotting-geographic-features-and-population-ranges","dir":"Articles","previous_headings":"","what":"Plotting geographic features and population ranges","title":"Introduction and basic tutorial","text":"slendr package implements plotting function called plot_map(). order make easier convenient iteratively build complex models. function can intelligently decide (based given input arguments) right way present data user, helps define models quickly without relying lower-level mechanisms sf package. see examples plot_map() action .","code":""},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"defining-smaller-geographic-regions","dir":"Articles","previous_headings":"","what":"Defining smaller geographic regions","title":"Introduction and basic tutorial","text":"addition overall spatial map context, can also define smaller geographic boundaries. mostly useful whenever want restrict population’s movement (spatial population expansion) smaller region map intuitive geographic meaning (.e., Anatolia, West Eurasia, etc.). Note objects created population boundaries (yet anyway)! simply labels generic geographic boundaries can used later. attached population point. , object returned region() function actually sf object, carries additional annotation name region (“Anatolia”): However, object also carries additional class annotations purpose internal slendr machinery: Furthermore, note region() calls specified map object defined beginning. object added hidden attribute slendr object represents context geospatial transformations, expansions, plots. can use generic plot_map() function plot geographic regions context defined section world map:  Note map object longer explicitly specified. needed, class objects provided plot_map() function must carry “map” attribute. fact, object must carry map context — slendr complains whenever case. can check component really , although hidden, using built-attr function verify map object created beginning:","code":"africa <- region(   \"Africa\", map,   polygon = list(c(-18, 20), c(38, 20), c(30, 33),                  c(20, 33), c(10, 38), c(-6, 35)) ) europe <- region(   \"Europe\", map,   polygon = list(     c(-8, 35), c(-5, 36), c(10, 38), c(20, 35), c(25, 35),     c(33, 45), c(20, 58), c(-5, 60), c(-15, 50)   ) ) anatolia <- region(   \"Anatolia\", map,   polygon = list(c(28, 35), c(40, 35), c(42, 40),                  c(30, 43), c(27, 40), c(25, 38)) ) anatolia #> slendr 'region' object  #> ----------------------  #> name: Anatolia  #>  #> map: internal coordinate reference system EPSG 3035 class(anatolia) #> [1] \"slendr\"        \"slendr_region\" \"sf\"            \"data.frame\" plot_map(africa, europe, anatolia, title = \"Geographic regions\") all(attr(europe, \"map\") == map) #> [1] TRUE all(attr(anatolia, \"map\") == map) #> [1] TRUE"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"defining-spatial-population-boundaries","dir":"Articles","previous_headings":"","what":"Defining spatial population boundaries","title":"Introduction and basic tutorial","text":"One aims slendr package formalize specification spatial population boundaries changes time. core function population(), takes population name, time want enforce population’s boundary, effective population size population time, map object described . also specify existing population specified population split (explicitly say ’s ancestral population). specifying actual spatial boundaries, several options.","code":""},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"polygon-population-ranges","dir":"Articles","previous_headings":"Defining spatial population boundaries","what":"Polygon population ranges","title":"Introduction and basic tutorial","text":"can define detailed population boundaries using polygon geometry object region object created region() function , using polygon = argument population(). , reminder, note coordinates described context geographic CRS. First, let’s create African ancestors modern humans. restrict spatial boundary African population africa region defined :","code":"afr <- population(\"AFR\", time = 52000, N = 3000, map = map, polygon = africa)  plot_map(afr)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"circular-population-ranges","dir":"Articles","previous_headings":"Defining spatial population boundaries","what":"Circular population ranges","title":"Introduction and basic tutorial","text":"want simulate abstract simple population boundary, can specify circular range center radius arguments instead polygon. distance units slendr package specified coordinate system given “world creation”. instance, EPSG 3035 (’re using ) specifies distances meters. define location population non-Africans right split African ancestors: call plot_map() function returned object, option either plot population range “raw” form “intersected” form, case raw boundary intersected “background” landscape (removing large bodies water, etc.). intersected form ultimately exported serialized format (see ) loaded spatial map SLiM. plot_map() function renders intersected population ranges default.","code":"ooa <- population(   \"OOA\", parent = afr, time = 51000, N = 500, remove = 25000,   center = c(33, 30), radius = 400e3 ) plot_map(ooa, intersect = TRUE, title = \"'Intersected' population range\")"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"population-movement-across-a-landscape","dir":"Articles","previous_headings":"","what":"Population movement across a landscape","title":"Introduction and basic tutorial","text":"describe directional population movement, can use function move(). accepts coordinates destination points along way (trajectory) duration migration, automatically generates number intermediate spatial maps along trajectory movement produce reasonable degree spatial continuity (number can also specified manually). can inspect object returned move() function see now contains just first YAM population range 7000 years ago, also ranges intermediate locations: Checking result visually , see:  Let’s create population Eastern Hunter Gatherers (EHG), split first non-Africans 28000 years ago: ’re , let’s also create population Western Hunter Gatherers (WHG). people living region eventually became present day Europeans receiving gene flow groups time (see ), call “EUR” simplify modeling code little bit:","code":"ooa <- ooa %>% move(   trajectory = list(c(40, 30), c(50, 30), c(60, 40)),   start = 50000, end = 40000 ) ooa #> slendr 'population' object  #> --------------------------  #> name: OOA  #> habitat: terrestrial #>  #> number of spatial maps: 28  #> map: internal coordinate reference system EPSG 3035  #> scheduled removal at time  25000  #>  #> population history overview: #>   - time 51000: split from AFR (N = 500) #>   - time 50000-40000: movement across a landscape plot_map(ooa, title = \"Intermediate migration maps\") ehg <- population(   \"EHG\", parent = ooa, time = 28000, N = 1000, remove = 6000,   polygon = list(     c(26, 55), c(38, 53), c(48, 53), c(60, 53),     c(60, 60), c(48, 63), c(38, 63), c(26, 60)) ) eur <- population( # European population   name = \"EUR\", parent = ehg, time = 25000, N = 2000,   polygon = europe )"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"spatial-population-expansion","dir":"Articles","previous_headings":"","what":"Spatial population expansion","title":"Introduction and basic tutorial","text":"can simulate expanding range population using function expand_range(), accepts parameters specifying many kilometers boundary expand (argument), long expansion take (duration argument) many intermediate spatial map snapshots exported representing expansion (snapshots argument). instance, let’s represent expansion Anatolian farmers, also split OOA population 28000 years ago time split EHG population. Note use optional parameter, polygon, restricts expansion Europe, instead around Anatolia: Note , principle, specify entire spatio-temporal history population single pipeline using pipe operator %>%. , can inspect object returned expand_range() function see contains spatial maps (“snapshots”) expansion process across time: can () check results visually:  visually see really going behind scenes, can also plot raw, non-intersected form expansion : can see population Anatolian farmers point invades spatial boundary EUR population. , doesn’t imply gene flow. section gene flow , see slendr implements gene flow overlapping (non-overlapping) populations. Let’s add couple populations migrations move implementing gene flow . Yamnaya steppe herders:","code":"ana <- population( # Anatolian farmers   name = \"ANA\", time = 28000, N = 3000, parent = ooa, remove = 4000,   center = c(34, 38), radius = 500e3, polygon = anatolia ) %>%   expand_range( # expand the range by 2.500 km     by = 2500e3, start = 10000, end = 7000,     polygon = join(europe, anatolia)   ) ana #> slendr 'population' object  #> --------------------------  #> name: ANA  #> habitat: terrestrial #>  #> number of spatial maps: 16  #> map: internal coordinate reference system EPSG 3035  #> scheduled removal at time  4000  #>  #> population history overview: #>   - time 28000: split from OOA (N = 3000) #>   - time 10000-7000: range expansion plot_map(ana, title = \"Anatolian expansion into Europe\") plot_map(ana, title = \"Anatolian expansion into Europe (not intersected)\", intersect = FALSE) yam <- population( # Yamnaya steppe population   name = \"YAM\", time = 7000, N = 500, parent = ehg, remove = 2500,   polygon = list(c(26, 50), c(38, 49), c(48, 50),                  c(48, 56), c(38, 59), c(26, 56)) ) %>%   move(     trajectory = c(15, 50),     start = 5000, end = 3000, snapshots = 8   )  plot_map(yam)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"plotting-multiple-slendr-objects","dir":"Articles","previous_headings":"","what":"Plotting multiple slendr objects","title":"Introduction and basic tutorial","text":"addition plotting individual population ranges, generic function plot_map() can handle combination population ranges, can also partition individual facets. useful visual inspection specified model, looking potential issues export individual spatio-temporal maps. Obviously, lot multi-dimensional information:  see better way explore slendr model interactively.","code":"plot_map(afr, ooa, ehg, eur, ana, yam)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"defining-gene-flow-events","dir":"Articles","previous_headings":"","what":"Defining gene flow events","title":"Introduction and basic tutorial","text":"way slendr implements gene flow events calling gene_flow() function. function straightforward interface, shown . One thing note default, populations gene flow events must overlapping spatial ranges order simulate gene flow. probably rather obvious, populations can’t mix space-time don’t overlap given point space-time. example, look spatial boundaries plotted , ’ll see European African populations don’t overlap population ranges. try instruct slendr simulate geneflow , get error: error message instructs us visually verify case, can done slendr’s plot_map() function optional parameter pop_facets = F (set TRUE default). Many models include multiple gene flow events, can collect simple R list: gene_flow() function simply returns data frame collecting geneflow parameters compile_model() step :","code":"gf <- gene_flow(from = eur, to = afr, rate = 0.1, start = 20000, end = 15000) Error: No overlap between population ranges of EUR and AFR at time 20000.    Please check the spatial maps of both populations by running   `plot_map(eur, afr)` and adjust them accordingly. Alternatively, in case   this makes sense for your model, you can add `overlap = F` which   will instruct slendr to simulate gene flow without spatial overlap   between populations. gf <- list(   gene_flow(from = ana, to = yam, rate = 0.5, start = 6500, end = 6400, overlap = FALSE),   gene_flow(from = ana, to = eur, rate = 0.5, start = 8000, end = 6000),   gene_flow(from = yam, to = eur, rate = 0.75, start = 4000, end = 3000) ) gf #> [[1]] #>   from_name to_name tstart tend rate overlap #> 1       ANA     YAM   6500 6400  0.5   FALSE #>  #> [[2]] #>   from_name to_name tstart tend rate overlap #> 1       ANA     EUR   8000 6000  0.5    TRUE #>  #> [[3]] #>   from_name to_name tstart tend rate overlap #> 1       YAM     EUR   4000 3000 0.75    TRUE"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"compile-the-whole-model-and-load-it-in-slim","dir":"Articles","previous_headings":"","what":"Compile the whole model and load it in SLiM","title":"Introduction and basic tutorial","text":"crucial function slendr compile_model(). takes population ranges defined across space time, together list gene flow events (optional, since models won’t include gene flow), proceeds converting vectorized spatial ranges raster bitmaps. Furthermore, compiles information split times, \\(N_e\\) values, gene flow directions, times, rates series tables. saved automatically dedicated directory format understood back end SLiM script provided slendr (). files model directory look like? Ideally, user never worry ; fact, whole purpose slendr let work much higher level abstraction without worrying low-level details. said, might find useful see things look like behind curtain… First , can inspect contents directory see , indeed, contain defined spatial maps (now PNG files, SLiM requires). also contains series tab-delimited configuration tables. tables contain summaries model parameters defined graphically , namely: table population splits: table geneflow events: finally, table populations whose spatial maps updated throughout simulation, well times updates (table rather large, ’re showing ). object returned compile_model() function (called model ) binds information together. fact, easier debugging sanity checking, carries locations tables (well important information) inside , elements list: model$splits, model$geneflows, etc. case ’d want separate model specification running different scripts, slendr includes function read_model() just purpose:","code":"model_dir <- paste0(tempfile(), \"_tutorial-model\")  model <- compile_model(   populations = list(afr, ooa, ehg, eur, ana, yam), # populations defined above   gene_flow = gf, # gene-flow events defined above   generation_time = 30,   resolution = 10e3, # resolution in meters per pixel   competition = 130e3, mating = 100e3, # spatial interaction in SLiM   dispersal = 70e3, # how far will offspring end up from their parents   path = model_dir ) list.files(model_dir, pattern = \"*.jpg\") #> character(0) read.table(file.path(model_dir, \"populations.tsv\"), header = TRUE) #>   pop            parent    N tsplit_gen tsplit_orig tremove_gen tremove_orig #> 1 AFR __pop_is_ancestor 3000          1       52000          -1           -1 #> 2 OOA               AFR  500         34       51000         901        25000 #> 3 EHG               OOA 1000        801       28000        1534         6000 #> 4 ANA               OOA 3000        801       28000        1601         4000 #> 5 EUR               EHG 2000        901       25000          -1           -1 #> 6 YAM               EHG  500       1501        7000        1651         2500 #>   pop_id parent_id #> 1      0        -1 #> 2      1         0 #> 3      2         1 #> 4      3         1 #> 5      4         2 #> 6      5         2 read.table(file.path(model_dir, \"geneflow.tsv\"), header = TRUE) #>   from  to rate overlap tstart_gen tstart_orig tend_gen tend_orig from_id to_id #> 1  ANA YAM 0.50       0       1518        6500     1521      6400       3     5 #> 2  ANA EUR 0.50       1       1468        8000     1534      6000       3     4 #> 3  YAM EUR 0.75       1       1601        4000     1634      3000       5     4 loaded_model <- read_model(model_dir)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"visualize-the-entire-history-of-splits-and-gene-flow","dir":"Articles","previous_headings":"","what":"Visualize the entire history of splits and gene flow","title":"Introduction and basic tutorial","text":"code snippets , defined simple history European populations last 50000 years. history includes population splits gene flow events, well demographic changes. slendr tries make formal specification spatio-temporal population dynamics concise possible, hard really visualize everything happen SLiM side simulation starts just code alone. purpose, package includes function named plot_model() takes information relationships populations (.e., population gene flow objects defined ) plots form -called admixture graph (see discussion admixture graph concept). One important thing note unlike traditional admixture graphs node/population present , full slendr graph single population can participate many gene flow events course history. visualized assigning color population, different nodes color represent snapshots time demographic event affecting population happens.","code":"plot_model(model, proportions = TRUE)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"interactive-exploration-of-spatio-temporal-models","dir":"Articles","previous_headings":"","what":"Interactive exploration of spatio-temporal models","title":"Introduction and basic tutorial","text":"slightly fancier way visualize models implemented function explore_model(). function accepts compiled model parameter spawns R shiny-based browser app makes possible click time snapshots interactively visualize spatial maps time point.","code":"explore_model(model)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"running-the-simulation","dir":"Articles","previous_headings":"","what":"Running the simulation","title":"Introduction and basic tutorial","text":"way feed entire serialized model SLiM slim() function, understands format model directory created compile_model() function generates SLiM script (using back end skeleton script part package can found calling system.file(\"scripts/script.slim\", package = \"slendr\"), case ’d like peek internals). output slendr simulation tree-sequence file (produced SLiM simulation behind scenes). tree sequence automatically loaded R returned user: lots slendr allows , terms simulation , also terms analyzing tree-sequence data. list provides list additional resources might want look .","code":"ts <- slim(model, sequence_length = 100000, recombination_rate = 1e-8) ts #> ╔═══════════════════════╗ #> ║TreeSequence           ║ #> ╠═══════════════╤═══════╣ #> ║Trees          │    133║ #> ╟───────────────┼───────╢ #> ║Sequence Length│ 100000║ #> ╟───────────────┼───────╢ #> ║Time Units     │  ticks║ #> ╟───────────────┼───────╢ #> ║Sample Nodes   │  10000║ #> ╟───────────────┼───────╢ #> ║Total Size     │2.7 MiB║ #> ╚═══════════════╧═══════╝ #> ╔═══════════╤═════╤═════════╤════════════╗ #> ║Table      │Rows │Size     │Has Metadata║ #> ╠═══════════╪═════╪═════════╪════════════╣ #> ║Edges      │18682│583.8 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Individuals│12838│  1.2 MiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Migrations │    0│  8 Bytes│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Mutations  │    0│  1.2 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Nodes      │18332│681.0 KiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Populations│    6│  2.6 KiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Provenances│    1│ 34.7 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Sites      │    0│ 16 Bytes│          No║ #> ╚═══════════╧═════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"more-information","dir":"Articles","previous_headings":"","what":"More information","title":"Introduction and basic tutorial","text":"vignette described basic features slendr package already quite long. much slendr demonstrated . instance: can tweak parameters influencing dispersal dynamics (“clumpy” populations , far offspring can migrate parents, etc.) change dynamics evolve time. See vignette information. can use slendr program non-spatial models, means standard, Wright-Fisher demographic model can simulated lines R code , instance, plugged Approximate Bayesian Computation pipeline analyses leveraging readily available R packages. can find vignette much detailed example vignette SLiM msprime back ends. can build complex spatial models still abstract (assuming real geographic location), including traditional simulations demes lattice structure. complete example shown vignette. SLiM saves data .trees tree-sequence file format, thanks R package reticulate interfacing Python code incredible power tskit pyslim process simulated data massive scale right fingertips, within convenient environment R. See much detailed example vignette vignette extensive tutorial feature.","code":""},{"path":"https://www.slendr.net/articles/vignette-02-grid-model.html","id":"simple-two-dimensional-grid-model","dir":"Articles","previous_headings":"","what":"Simple two-dimensional grid model","title":"Demes on a regular spatial grid","text":"First, let’s load slendr R package create two-dimensional abstract world map: Next, define helper function ) create single slendr population object, b) place population appropriate coordinate lattice world map based numeric identifier population (runs 1 \\(n \\times n\\) , n total number demes along one side regular grid): defined population construction function, let’s build model. Let’s say want create regular grid n × n populations, N individuals population: Let’s plot whole spatial population configuration, make sure set things correctly:  far, way model specified, population stuck circular “island”. can change programming gene flow events using slendr function gene_flow(). , let’s first program simple helper function generate gene flow events according neighborhood relationships two-dimensional grid, allowing population exchange migrants neighbors (making sure coordinates population stay within grid using simple modulo arithmetic population index ). Let’s test function. gene flow events population lower left corner grid (, first population series)? everything works, population allowed exchange migrants neighbor right (population number 2) neighbor . Looks right! Let’s generate entire set continuous gene flow events: total number individual gene flow events : Finally, can compile whole model: familiar SLiM manual recognize model described section 5.3.3. Finally, can run simulation using slim() function.","code":"library(slendr)  init_env() #> The interface to all required Python modules has been activated. map <- world(   xrange = c(0, 1000),   yrange = c(0, 1000),   landscape = \"blank\" ) create_pop <- function(i, n_side, map, N, radius) {   # get dimensions of the world map   dim <- c(diff(attr(map, \"xrange\")), diff(attr(map, \"yrange\")))    # position of the i-th population on the two-dimensional lattice grid   coords <- c((i - 1) %% n_side, (i - 1) %/% n_side)   center <- coords / n_side * dim + dim / (2 * n_side)    pop <- tryCatch({     population(       name = sprintf(\"pop%d\", i),       N = N,       time = 1,       map = map,       center = center + c(attr(map, \"xrange\")[1], attr(map, \"yrange\")[1]),       radius = radius     )   }, error = function(e) NULL)    pop } n <- 5  populations <-   seq(1, n * n) %>%   lapply(create_pop, n_side = n, map = map, N = 100, radius = 40) do.call(plot_map, populations) + ggplot2::theme(legend.position = \"none\") set_geneflow <- function(i, n_side, rate, start, end, populations) {   pop <- populations[[i]]    # get the position of the i-th population on the n*n grid   coords <- c((i - 1) %% n_side, (i - 1) %/% n_side)     # get coordinates of the i-th population's neighbors on the grid   neighbor_pos <- list(     c(coords[1] - 1, coords[2]), c(coords[1] + 1, coords[2]),     c(coords[1], coords[2] + 1), c(coords[1], coords[2] - 1)   )    # generate geneflow events for population coordinates inside the grid   geneflows <- lapply(neighbor_pos, function(pos) {     if (any(pos < 0 | pos >= n_side)) return(NULL)     neighbor <- populations[[pos[2] * n_side + pos[1] + 1]]     if (is.null(neighbor)) return(NULL)      rbind(       gene_flow(from = pop, to = neighbor, rate = rate, start = start, end = end, overlap = FALSE),       gene_flow(from = neighbor, to = pop, rate = rate, start = start, end = end, overlap = FALSE)     )   }) %>%     do.call(rbind, .)    geneflows } set_geneflow(1, n, rate = 0.1, start = 2, end = 1000, populations) #>   from_name to_name tstart tend rate overlap #> 1      pop1    pop2      2 1000  0.1   FALSE #> 2      pop2    pop1      2 1000  0.1   FALSE #> 3      pop1    pop6      2 1000  0.1   FALSE #> 4      pop6    pop1      2 1000  0.1   FALSE geneflows <-   seq(1, n * n) %>%   lapply(set_geneflow, n, rate = 0.05, start = 2, end = 1000, populations) %>%   do.call(rbind, .) %>%   unique # filter out duplicate events due to symmetries nrow(geneflows) #> [1] 80 model <- compile_model(   populations = populations, gene_flow = geneflows,   generation_time = 1, resolution = 10,   competition = 10, mating = 10, dispersal = 10,   simulation_length = 1000 ) ts <- slim(model, sequence_length = 10000, recombination_rate = 0) # simulate a single 10kb locus ts #> ╔═══════════════════════╗ #> ║TreeSequence           ║ #> ╠═══════════════╤═══════╣ #> ║Trees          │      1║ #> ╟───────────────┼───────╢ #> ║Sequence Length│  10000║ #> ╟───────────────┼───────╢ #> ║Time Units     │  ticks║ #> ╟───────────────┼───────╢ #> ║Sample Nodes   │   5000║ #> ╟───────────────┼───────╢ #> ║Total Size     │1.3 MiB║ #> ╚═══════════════╧═══════╝ #> ╔═══════════╤════╤═════════╤════════════╗ #> ║Table      │Rows│Size     │Has Metadata║ #> ╠═══════════╪════╪═════════╪════════════╣ #> ║Edges      │9065│283.3 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Individuals│6316│618.4 KiB│         Yes║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Migrations │   0│  8 Bytes│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Mutations  │   0│  1.2 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Nodes      │9090│338.0 KiB│         Yes║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Populations│  25│  5.7 KiB│         Yes║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Provenances│   1│ 34.6 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Sites      │   0│ 16 Bytes│          No║ #> ╚═══════════╧════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-02-grid-model.html","id":"population-grid-on-a-real-geographic-landscape","dir":"Articles","previous_headings":"","what":"Population grid on a real geographic landscape","title":"Demes on a regular spatial grid","text":"can take things one step . wanted similar thing (.e. simulate regularly spaced demes) real geographic context? Let’s zoom interesting part world create grid demes using helper function create_pop defined (population boundary 300 km diameter): course, lay regular grid across map world, population boundaries fall outside African continent. solve issue, go list populations filter least 50% area land, using another helper function: Let’s plot layout population grid real geographic background:  Next, probably set scenario gene flow subpopulations; perhaps interested studying selected allele spreads continent based factors interest. , simulate data spatial model, first compile_model() run SLiM via slim() function. Given process described example , won’t repeating .","code":"map <- world(   xrange = c(-25, 55),   yrange = c(-32, 35),   crs = 4326 )  n <- 20  populations <-   seq(1, n * n) %>%   lapply(create_pop, n_side = n, map = map, N = 100, radius = 1.5) continent <- region(   map = map, polygon = list(     c(-10, 35), c(-20, 20), c(-15, 8), c(-10, 5), c(0, 2),     c(20, -40), c(35, -32), c(50, -25), c(55, -10), c(50, 0),     c(53, 13), c(45, 10), c(37, 20), c(32, 30), c(16, 38), c(0, 38)   ) )  check_area <- function(pop, map, continent) {   if (is.null(pop)) return(NULL)    # total population area   pop_area <- area(pop)$area   # population area overlapping a map   map_area <- area(overlap(pop, map))   # population area overlapping African continent   continent_area <- area(overlap(pop, continent))    # at least 50% of population's boundary be on land, and it must fall   # on to the African continent itself   if (continent_area == 0 || (map_area / pop_area) < 0.5)     return(NULL)   else     return(pop) }  filtered <- lapply(populations, check_area, map, continent) %>%   Filter(Negate(is.null), .) do.call(plot_map, filtered) + ggplot2::theme(legend.position = \"none\")"},{"path":"https://www.slendr.net/articles/vignette-02-grid-model.html","id":"more-customized-spatial-model","dir":"Articles","previous_headings":"","what":"More customized spatial model","title":"Demes on a regular spatial grid","text":"want introduce spatiality model need manual control position subpopulation, can course customize spatial layout much detail. Consider followin map South America:  Let’s lay demes scattered, irregular fashion. illustration purposes, define population’s geographic range circle radius 200 km. Importantly, note introduced one ancestral population population associated location world map! want simulate data coalescent backend, formally encode population genealogies eventually coalesce. words, models specified run msprime() back end contain isolated demes formally descend single ancestor run models msprime, get error infinite coalescent times. formally introduce ancestral population : Furthermore, let’s say fairly good idea complex interaction/gene-flow network deme, can encode like (instead forcing regular arrangement demes gene-flow interactions previous examples): Now can compile model&nbsp`twist, however: try simulate data non-spatial way using msprime, skip serialization model data disk, skipping spatial interaction dispersal parameters. words, care simulating demographic model traditional, Wright-Fisher, random-mating demes gene flow , dispersal within individual deme. ? Mostly just demonstrate possible! However, certainly situations “half-spatial” model useful – instance, situation within-deme individual dynamics interest, location subpopulation . can make sure demographic model specified correctly quickly visualizing :  can also plot spatial organization demes map glory, including gene flow events (one arrow per unidirectional gene flow). Note get warning message stating (non-spatial) ancestral population won’t visualized map:  Finally, can simulate tree sequence model! Notice simulating data using msprime(), effectively dropping continuous space dimension model. situation, map serves visual aid, making easier set complex “spatial” position demes map.","code":"xrange <- c(-90, -20) yrange <- c(-58, 15)  map <- world(xrange = xrange, yrange = yrange, crs = \"EPSG:31970\") #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpI0CWYj/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84 plot_map(map) # non-spatial ancestral population p_anc <- population(\"p_anc\", N = 1000, time = 1)  # spatial populations p1 <- population(\"p1\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-75, 0), radius = 200e3) p2 <- population(\"p2\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-60, 5), radius = 200e3) p3 <- population(\"p3\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-65, -5), radius = 200e3) p4 <- population(\"p4\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-60, -20), radius = 200e3) p5 <- population(\"p5\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-65, -35), radius = 200e3) p6 <- population(\"p6\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-69, -42), radius = 200e3) p7 <- population(\"p7\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-51, -10), radius = 200e3) p8 <- population(\"p8\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-45, -15), radius = 200e3) p9 <- population(\"p9\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-71, -12), radius = 200e3) p10 <- population(\"p10\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-55, -25), radius = 200e3) gf <- list(   gene_flow(p1, p2, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p2, p1, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p1, p3, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p3, p1, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p2, p3, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p3, p2, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p2, p7, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p7, p2, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p3, p7, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p7, p3, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p7, p8, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p8, p7, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p4, p7, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p7, p4, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p4, p5, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p5, p4, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p5, p6, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p6, p5, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p3, p4, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p4, p3, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p1, p9, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p9, p1, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p3, p9, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p9, p3, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p4, p9, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p9, p4, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p10, p4, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p4, p10, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p10, p8, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p8, p10, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p10, p5, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p5, p10, 0.1, 1000, 2000, overlap = FALSE) ) model <- compile_model(   populations = list(p_anc, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10), gene_flow = gf,   generation_time = 1, simulation_length = 5000,   serialize = FALSE ) #> Warning: Model containing a mix of spatial and non-spatial populations will be compiled. #> Although this is definitely supported, make sure this is really what you want. #> Warning: Spatial models must be serialized to disk for SLiM to simulate data from. #> Compiled like this, your model can only be simulated with msprime. plot_model(model) plot_map(model, gene_flow = TRUE) #> Warning: All gene-flow event will be visualized at once. If you wish to visualize #> gene flows at a particular point in time, use the `time` argument. #> Warning: Non-spatial populations in your model won't be visualized ts <- msprime(model, sequence_length = 10e6, recombination_rate = 1e-8, random_seed = 42)  ts #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │      34057║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│   10000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │      22000║ #> ╟───────────────┼───────────╢ #> ║Total Size     │    9.4 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │182742│  5.6 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│ 11000│300.8 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │     0│ 16 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │ 74344│  2.0 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│    11│605 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     1│ 10.5 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │     0│ 16 Bytes│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"setting-up-python-environment","dir":"Articles","previous_headings":"","what":"Setting up Python environment","title":"Tree-sequence processing and statistics","text":"First, order able interface tskit pyslim using reticulate package (run simulations using msprime, ), need working Python environment required Python modules pyslim, tskit msprime already installed. setting Python environments can quite hassle, slendr provides single function setup_env() make things easier. call without arguments, slendr automatically download, install, setup completely separate Python environment (based “miniconda” distribution) just slendr activate background. important stress setup_env() interfere way Python installations might already computer. Python installation environment entirely isolated used just purpose slendr workflows. Python environment set , can activate calling: can use another built-function check_env() make sure slendr installed configured correct environment us: Now ’re good go ready simulate analyse tree sequence outputs R!","code":"setup_env() init_env() #> The interface to all required Python modules has been activated. check_env() #> Summary of the currently active Python environment: #>  #> Python binary: /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.0_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python  #> Python version: 3.12.1 | packaged by conda-forge | (main, Dec 23 2023, 08:01:35) [Clang 16.0.6 ]  #>  #> slendr requirements: #>  - tskit: version 0.5.6 ✓  #>  - msprime: version 1.3.0 ✓  #>  - pyslim: version 1.0.4 ✓  #>  - tspop: present ✓"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"model-of-neanderthal-introgression-into-eurasians","dir":"Articles","previous_headings":"Setting up Python environment","what":"Model of Neanderthal introgression into Eurasians","title":"Tree-sequence processing and statistics","text":"First, let’s set simple non-spatial model Neanderthal introgression using slendr. essentially procedure shown another vignette introducing non-spatial slendr models. different spatial model, except left map argument calling population(). ’s toy model visualized “demographic graph” sorts (.e., tree-like structure specifying population splits additional edges representing gene flow events). particularly illuminating simple example, ’s always worth keeping mind graph embedded within every slendr model can always invoked make sure model ’re setting correct:","code":"library(ggplot2) library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union set.seed(314159)  # create the ancestor of everyone and a chimpanzee outgroup # (we set both N = 1 to reduce the computational time for this model) chimp <- population(\"CH\", time = 6.5e6, N = 1000)  # two populations of anatomically modern humans: Africans and Europeans afr <- population(\"AFR\", parent = chimp, time = 6e6, N = 10000) eur <- population(\"EUR\", parent = afr, time = 70e3, N = 5000)  # Neanderthal population splitting at 600 ky ago from modern humans # (becomes extinct by 40 ky ago) nea <- population(\"NEA\", parent = afr, time = 600e3, N = 1000, remove = 40e3)  # 3% Neanderthal introgression into Europeans between 55-50 ky ago gf <- gene_flow(from = nea, to = eur, rate = 0.03, start = 55000, end = 45000)  model <- compile_model(   populations = list(chimp, nea, afr, eur), gene_flow = gf,   generation_time = 30,   path = paste0(tempfile(), \"_introgression\") ) cowplot::plot_grid(   plot_model(model, sizes = FALSE),   plot_model(model, sizes = FALSE, log = TRUE),   nrow = 1 )"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"scheduling-of-sampling-events","dir":"Articles","previous_headings":"","what":"Scheduling of sampling events","title":"Tree-sequence processing and statistics","text":"Now defined model, sample data ? Ideally, like schedule sampling events given time, sampling defined number individuals given population. slendr provides function schedule_sampling() serves define sampling schedule automatically enforces populations already (.e. appearance simulation) still (removed simulation) present sampled . example, want sample two Neanderthal individuals (older one Altai Neanderthal published Pruefer et al. 2014, younger one Vindija Neanderthal published Pruefer et al., 2017). two genomes need estimate Neanderthal ancestry proportion using -called \\(f_4\\)-ratio statistic (, also see Petr et al., PNAS 2019): can see, schedule_sampling() function simply accepts vector times remembering schedule, list pairs (<slendr population>, <number individuals>) encoding populations many individuals remembered time points given times vector. Next, want sample present-day individuals: outgroup representing chimpanzee, couple Africans Europeans: can see , schedule_sampling() function returns plain old data frame simple structure three columns: time, population name, number individuals. means can define sampling events using whatever input data might already available (radiocarbon-dated ancient DNA samples Excel sheet publication). instance, lot interest estimate trajectory Neanderthal ancestry Europe time using ancient DNA data anatomically modern human individuals (also called early modern humans, EMH) across last couple tens thousands years. can simulate something close available EMH ancient DNA data set last 50 thousand years running : samples single ancient European individuals randomly chosen times 40 10 ky ago. One nice feature schedule_sampling() function schedules sampling events population, population present simulation given time. makes possible simply take wide time range sampling, specify populations sizes samples, let function generate sampling events populations present time. reason stricter control sampling required, behavior can switched setting strict = TRUE like : Now already model object ready, can simulate data , sampling individuals according sampling schedule. Although use slim() function shown previous vignettes, case run simulation msprime() coalescent back end. , model non-spatial using coalescent simulator much efficient forward simulation. Switching msprime SLiM back ends slendr demonstrated much detail dedicated vignette. simulation back end utilized msprime() function (well slim() function) produces tree-sequence output immediately loaded ready downstream analysis. Note bind individual sampling schedule data frames using rbind function provided base R (show , sampling schedule really just data frame can manipulate ).","code":"nea_samples <- schedule_sampling(model, times = c(70000, 40000), list(nea, 1)) nea_samples #> # A tibble: 2 × 7 #>    time pop       n y_orig x_orig y     x     #>   <int> <chr> <int> <lgl>  <lgl>  <lgl> <lgl> #> 1 40000 NEA       1 NA     NA     NA    NA    #> 2 70000 NEA       1 NA     NA     NA    NA present_samples <- schedule_sampling(model, times = 0, list(chimp, 1), list(afr, 5), list(eur, 10)) present_samples #> # A tibble: 3 × 7 #>    time pop       n y_orig x_orig y     x     #>   <int> <chr> <int> <lgl>  <lgl>  <lgl> <lgl> #> 1     0 CH        1 NA     NA     NA    NA    #> 2     0 AFR       5 NA     NA     NA    NA    #> 3     0 EUR      10 NA     NA     NA    NA emh_samples <- schedule_sampling(model, times = runif(n = 40, min = 10000, max = 40000), list(eur, 1)) emh_samples #> # A tibble: 40 × 7 #>     time pop       n y_orig x_orig y     x     #>    <int> <chr> <int> <lgl>  <lgl>  <lgl> <lgl> #>  1 10320 EUR       1 NA     NA     NA    NA    #>  2 11188 EUR       1 NA     NA     NA    NA    #>  3 11396 EUR       1 NA     NA     NA    NA    #>  4 11529 EUR       1 NA     NA     NA    NA    #>  5 11927 EUR       1 NA     NA     NA    NA    #>  6 12675 EUR       1 NA     NA     NA    NA    #>  7 13689 EUR       1 NA     NA     NA    NA    #>  8 13744 EUR       1 NA     NA     NA    NA    #>  9 14775 EUR       1 NA     NA     NA    NA    #> 10 16362 EUR       1 NA     NA     NA    NA    #> # ℹ 30 more rows # this attempts to sample a Neanderthal individual at a point when Neanderthals # are already extinct, resulting in an error schedule_sampling(model, times = 10000, list(nea, 1), strict = TRUE) Error: Cannot schedule sampling for 'NEA' at time 10000 because the population will not be present in the simulation at that point. Consider running this function with `strict = FALSE` which will automatically retain only valid sampling events. ts <- msprime(   model, sequence_length = 100e6, recombination_rate = 1e-8,   samples = rbind(nea_samples, present_samples, emh_samples),   random_seed = 314159, verbose = TRUE ) #> -------------------------------------------------- #> msprime command: #>  #> /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.0_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python /private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpQbon7e/filee17860409253_introgression/script.py --seed 314159 --model /private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpQbon7e/filee17860409253_introgression --output /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//RtmpQbon7e/filee1784cccb6b3.trees --sequence-length 100000000 --recombination-rate 1e-08 --sampling-schedule /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//RtmpQbon7e/filee17859cbb4d3 --verbose   #> -------------------------------------------------- #>  #> Tree sequence was saved to: #>  /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//RtmpQbon7e/filee1784cccb6b3.trees  #> Loading the tree-sequence file... ts #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │     240041║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        116║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   40.0 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │916588│ 28.0 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│    58│  1.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │     0│ 16 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │189654│  5.1 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     4│338 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     1│  7.0 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │     0│ 16 Bytes│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"r-interface-for-tskit-and-pyslim","dir":"Articles","previous_headings":"","what":"R interface for tskit and pyslim","title":"Tree-sequence processing and statistics","text":"Tree-sequences one revolutionary developments population genetics last couple decades number reasons. One possibility store extremely large data sets succinctly encoding entire evolutionary history sample individuals series correlated tree genealogies along genome. Going much detail topic clearly beyond scope tutorial, especially everything explain much better elsewhere. Instead, demonstrate rest vignette can access manipulate tree-sequence outputs generated slendr models perform various statistics using Python modules tskit pyslim directly slendr, without leave R! key magical R package reticulate creates seamless binding Python modules R. means even don’t know Python, slendr allows quite lot tree-sequences R. course, proficient Python user, needs said tree-sequence file generated slendr & SLiM, can easily perform every conceivable analysis directly using tskit. intention show can continue working tree-sequence files R even run entire slendr simulation.","code":""},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"loading-and-processing-tree-sequence-output-files","dir":"Articles","previous_headings":"","what":"Loading and processing tree-sequence output files","title":"Tree-sequence processing and statistics","text":"default, msprime() slim() run produce tree-sequence object (saved temporary file) immediately load . Thus, use-cases, explicit loading simulated tree sequence needed. said, sake completeness, let’s run simulation specify custom path tree-sequence file . case tree-sequence output saved custom location disk, can load tree sequence using slendr function ts_load(). ’re dealing tree sequence produced SLiM back end (case ), can also instruct function simplify tree-sequence individuals explicitly sampled (recall sampling schedule set schedule_sampling() function ). Note provide model object generated compile_model() order model annotation information simulated tree-sequence data (, loading): surprisingly, get output got printed tree sequence returned msprime() function. shows normal circumstances, loading output manually via ts_load() needed. try simplify msprime-generated tree sequence, get warning. tree sequence already simplified form, definition coming coalescent simulator. default, simplification trims tree sequence remembered individuals (.e. explicitly scheduled sampling), true every msprime tree sequence. Alternatively, can also narrow simplification defined set individuals using simplify_to = argument. Internally, simplification implemented dedicated function ts_simplify() can always call explicitly, like : Similarly, slendr provides function ts_recapitate() performs [recapitation]https://tskit.dev/pyslim/docs/latest/tutorial.html#recapitation). , needed msprime tree sequence, fully coalesced (recapitated) definition. current tree sequence object, simply get warning informing us ’re attempting something effect: can make sure tree sequence fully coalesced calling another slendr function ts_coalesced(). useful dealing slim()-produced tree sequences: might noticed simulate mutations SLiM run. computational efficiency. Luckily, tree-sequence contains complete history sample individuals makes easy sprinkle mutations genealogies simulation . can add mutations given rate running: processed simulated tree sequence, can calculate basic statistics simulated data. However, , first like note everything rest vignette (.e. whenever call function prefix ts_*() slendr), interfacing tskit Python module hood. goal capture analyses one might want perform tree-sequences R wrap neat interface indistinguishable R function—, , reason reticulate created first place (making various Python data science modules appear regular R packages).","code":"output_file <- tempfile()  ts <- msprime(   model, sequence_length = 100e6, recombination_rate = 1e-8,   samples = rbind(nea_samples, present_samples, emh_samples),   output = output_file, random_seed = 314159 )  output_file #> [1] \"/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//RtmpQbon7e/filee17821d8acf8\" ts <- ts_load(output_file, model) ts #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │     240041║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        116║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   40.0 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │916588│ 28.0 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│    58│  1.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │     0│ 16 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │189654│  5.1 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     4│338 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     1│  7.0 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │     0│ 16 Bytes│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝ ts_simplify(ts) #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │     238681║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        116║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   36.7 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │864772│ 26.4 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│    58│  1.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │     0│ 16 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │139203│  3.7 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     4│338 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     2│  7.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │     0│ 16 Bytes│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝ ts_small <- ts_simplify(ts, simplify_to = c(\"CH_1\", \"NEA_1\", \"NEA_2\", \"AFR_1\", \"AFR_2\", \"EUR_20\", \"EUR_50\")) ts_small #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │     132496║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │         14║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   19.0 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │441206│ 13.5 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│     7│220 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │     0│ 16 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │ 79249│  2.1 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     4│338 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     2│  7.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │     0│ 16 Bytes│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝ ts <- ts_recapitate(ts, recombination_rate = 1e-8, Ne = 10000) ts_coalesced(ts) #> [1] TRUE ts <- ts_mutate(ts, mutation_rate = 1e-8, random_seed = 314159) ts #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │     240041║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        116║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   76.8 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │916588│ 28.0 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│    58│  1.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │623249│ 22.0 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │189654│  5.1 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     4│338 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     2│  7.8 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │621331│ 14.8 MiB│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"visualisation-of-trees-and-tree-sequences","dir":"Articles","previous_headings":"","what":"Visualisation of trees and tree-sequences","title":"Tree-sequence processing and statistics","text":"Now introduce function ts_phylo() can used extract one tree tree-sequence (either -th tree sequence, tree overlapping -th position simulated genome, depending value mode argument) convert phylo class, standard format phylogenetic trees R world. phylo format, see packages ape, phangorn, phytools. type tree object R console, can verify got ordinary phylo object: means whole R phylogenetic ecosystem disposal analyze trees. instance can use powerful package ggtree plot tree just extracted:","code":"# extract the 42nd tree in the tree sequence tree <- ts_phylo(ts_small, 42 - 1) #> Starting checking the validity of tree... #> Found number of tips: n = 14  #> Found number of nodes: m = 13  #> Done. tree #>  #> Phylogenetic tree with 14 tips and 13 internal nodes. #>  #> Tip labels: #>   13 (EUR_50), 12 (EUR_50), 11 (CH_1), 10 (CH_1), 9 (AFR_2), 8 (AFR_2), ... #> Node labels: #>   77386, 70, 2699, 5175, 5199, 6642, ... #>  #> Rooted; includes branch lengths. library(ggtree) #> ggtree v3.10.0 For help: https://yulab-smu.top/treedata-book/ #>  #> If you use the ggtree package suite in published research, please cite #> the appropriate paper(s): #>  #> Guangchuang Yu, David Smith, Huachen Zhu, Yi Guan, Tommy Tsan-Yuk Lam. #> ggtree: an R package for visualization and annotation of phylogenetic #> trees with their covariates and other associated data. Methods in #> Ecology and Evolution. 2017, 8(1):28-36. doi:10.1111/2041-210X.12628 #>  #> Guangchuang Yu.  Data Integration, Manipulation and Visualization of #> Phylogenetic Trees (1st edition). Chapman and Hall/CRC. 2022, #> doi:10.1201/9781003279242 #>  #> G Yu. Data Integration, Manipulation and Visualization of Phylogenetic #> Trees (1st ed.). Chapman and Hall/CRC. 2022. ISBN: 9781032233574 ggtree(tree) +   geom_point2(aes(subset = !isTip)) + # points for internal nodes   geom_tiplab() + # sample labels for tips   hexpand(0.1)    # make more space for the tip labels library(ape) plot(tree) nodelabels()"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"accessing-tskit-functionality-directly","dir":"Articles","previous_headings":"","what":"Accessing tskit functionality directly","title":"Tree-sequence processing and statistics","text":"mentioned previous section, goal vignette show use slendr perform main tree-sequence operations using convenient R interface tskit. However, always keep mind restricted subset tskit functionality slendr translated R (.e. functions prefix ts_). Thanks incredible R package reticulate, can access Python methods object variables directly, using $ operator. example, instead calling function ts_coalesced() tree-sequence , check trees coalesced running following snippet instead (note inefficient ’re operation first one hundred trees): believe makes sense use R interface whenever possible (even makes many operations little bit convenient). However, functionality slendr missing, can always resort accessing Python objects directly just demonstrated. can verify methods attributes Python tree-sequence object still accessible R: fact, recognize elements output examples involving ts_ functions vignette! short, blackbox—slendr provides slightly convenient layer tskit R users.","code":"# iterate over all trees in the tree-sequence and check if each # has only one root (i.e. is fully coalesced) - note that Python # lists are 0-based, which is something we need to take care of all(sapply(seq_len(ts$num_trees)[1:100],            function(i) ts$at_index(i - 1)$num_roots == 1)) names(ts) #>   [1] \"alignments\"                      \"allele_frequency_spectrum\"       #>   [3] \"as_fasta\"                        \"as_nexus\"                        #>   [5] \"as_vcf\"                          \"aslist\"                          #>   [7] \"at\"                              \"at_index\"                        #>   [9] \"breakpoints\"                     \"check_index\"                     #>  [11] \"coalescence_time_distribution\"   \"coiterate\"                       #>  [13] \"count_topologies\"                \"decapitate\"                      #>  [15] \"delete_intervals\"                \"delete_sites\"                    #>  [17] \"diffs\"                           \"discrete_genome\"                 #>  [19] \"discrete_time\"                   \"divergence\"                      #>  [21] \"divergence_matrix\"               \"diversity\"                       #>  [23] \"draw_svg\"                        \"draw_text\"                       #>  [25] \"dump\"                            \"dump_tables\"                     #>  [27] \"dump_text\"                       \"edge\"                            #>  [29] \"edge_diffs\"                      \"edges\"                           #>  [31] \"edges_child\"                     \"edges_left\"                      #>  [33] \"edges_parent\"                    \"edges_right\"                     #>  [35] \"edgesets\"                        \"equals\"                          #>  [37] \"f2\"                              \"f3\"                              #>  [39] \"f4\"                              \"file_uuid\"                       #>  [41] \"first\"                           \"Fst\"                             #>  [43] \"genealogical_nearest_neighbours\" \"general_stat\"                    #>  [45] \"genetic_relatedness\"             \"genetic_relatedness_weighted\"    #>  [47] \"genotype_matrix\"                 \"get_ll_tree_sequence\"            #>  [49] \"get_num_mutations\"               \"get_num_nodes\"                   #>  [51] \"get_num_records\"                 \"get_num_sites\"                   #>  [53] \"get_num_trees\"                   \"get_pairwise_diversity\"          #>  [55] \"get_population\"                  \"get_sample_size\"                 #>  [57] \"get_samples\"                     \"get_sequence_length\"             #>  [59] \"get_time\"                        \"haplotypes\"                      #>  [61] \"has_reference_sequence\"          \"ibd_segments\"                    #>  [63] \"impute_unknown_mutations_time\"   \"indexes_edge_insertion_order\"    #>  [65] \"indexes_edge_removal_order\"      \"individual\"                      #>  [67] \"individual_locations\"            \"individual_populations\"          #>  [69] \"individual_times\"                \"individuals\"                     #>  [71] \"individuals_flags\"               \"individuals_location\"            #>  [73] \"individuals_population\"          \"individuals_time\"                #>  [75] \"kc_distance\"                     \"keep_intervals\"                  #>  [77] \"last\"                            \"ll_tree_sequence\"                #>  [79] \"load\"                            \"load_tables\"                     #>  [81] \"ltrim\"                           \"max_root_time\"                   #>  [83] \"max_time\"                        \"mean_descendants\"                #>  [85] \"metadata\"                        \"metadata_schema\"                 #>  [87] \"migration\"                       \"migrations\"                      #>  [89] \"migrations_dest\"                 \"migrations_left\"                 #>  [91] \"migrations_node\"                 \"migrations_right\"                #>  [93] \"migrations_source\"               \"migrations_time\"                 #>  [95] \"min_time\"                        \"mutation\"                        #>  [97] \"mutations\"                       \"mutations_node\"                  #>  [99] \"mutations_parent\"                \"mutations_site\"                  #> [101] \"mutations_time\"                  \"nbytes\"                          #> [103] \"newick_trees\"                    \"node\"                            #> [105] \"nodes\"                           \"nodes_flags\"                     #> [107] \"nodes_individual\"                \"nodes_population\"                #> [109] \"nodes_time\"                      \"num_edges\"                       #> [111] \"num_individuals\"                 \"num_migrations\"                  #> [113] \"num_mutations\"                   \"num_nodes\"                       #> [115] \"num_populations\"                 \"num_provenances\"                 #> [117] \"num_samples\"                     \"num_sites\"                       #> [119] \"num_trees\"                       \"pairwise_diversity\"              #> [121] \"parse_windows\"                   \"population\"                      #> [123] \"populations\"                     \"provenance\"                      #> [125] \"provenances\"                     \"records\"                         #> [127] \"reference_sequence\"              \"rtrim\"                           #> [129] \"sample_count_stat\"               \"sample_size\"                     #> [131] \"samples\"                         \"segregating_sites\"               #> [133] \"sequence_length\"                 \"simplify\"                        #> [135] \"site\"                            \"sites\"                           #> [137] \"sites_position\"                  \"split_edges\"                     #> [139] \"subset\"                          \"table_metadata_schemas\"          #> [141] \"tables\"                          \"tables_dict\"                     #> [143] \"Tajimas_D\"                       \"time_units\"                      #> [145] \"to_macs\"                         \"to_nexus\"                        #> [147] \"trait_correlation\"               \"trait_covariance\"                #> [149] \"trait_linear_model\"              \"trait_regression\"                #> [151] \"trees\"                           \"trim\"                            #> [153] \"union\"                           \"variants\"                        #> [155] \"write_fasta\"                     \"write_nexus\"                     #> [157] \"write_vcf\"                       \"Y1\"                              #> [159] \"Y2\"                              \"Y3\""},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"calculating-f-statistics","dir":"Articles","previous_headings":"","what":"Calculating f-statistics","title":"Tree-sequence processing and statistics","text":"addition revolutionary breakthrough terms computation efficiency, many statistics often interested population genetics natural consequence direct access tree sequence genealogies, simply genealogies capture true demographic history sample. , can’t go much detail encourage take look paper Ralph et al. duality statistics expressed terms branch lengths traditional summaries based samples genetic variation. instance, functions ts_f2(), ts_f3(), ts_f4() ts_f4ratio() calculate well-known set Patterson’s \\(f\\)-statistics: functions accept mode = argument, specifying whether statistics calculated using mutation site patterns (mode = \"site\", default), branch lengths (mode = \"branch\"), node (mode = \"node\"), well windows argument, similarly “multiway” statistics implemented tskit. See relevant sections official tskit documentation topic. Note previous chunk referred individuals names (numeric IDs nodes tskit Python). allow readability make easier see individuals based specified sampling schedule (names assigned individuals based order sampling). can get overview individuals scheduled sampling (.e. permanently remembered) names helper function ts_samples(): said, like run statistics nodes rather individuals, can simply using integer IDs instead character names function’s interface.","code":"# f2 is a measure of the branch length connecting A and B ts_f2(ts, A = \"EUR_1\", B = \"AFR_1\") #> # A tibble: 1 × 3 #>   A     B            f2 #>   <chr> <chr>     <dbl> #> 1 EUR_1 AFR_1 0.0000448 # f4 is a measure of the drift shared between A and B after their split from C ts_f3(ts, A = \"EUR_1\", B = \"AFR_1\", C = \"CH_1\") #> # A tibble: 1 × 4 #>   A     B     C            f3 #>   <chr> <chr> <chr>     <dbl> #> 1 EUR_1 AFR_1 CH_1  0.0000130 # this value should be very close to zero (no introgression in Africans) ts_f4(ts, \"AFR_1\", \"AFR_2\", \"NEA_1\", \"CH_1\", mode = \"branch\") #> # A tibble: 1 × 5 #>   W     X     Y     Z        f4 #>   <chr> <chr> <chr> <chr> <dbl> #> 1 AFR_1 AFR_2 NEA_1 CH_1  -118. # this value should be significantly negative (many more ABBA sites # compared to BABA site due to the introgression into Europeans) ts_f4(ts, \"AFR_1\", \"EUR_1\", \"NEA_1\", \"CH_1\", mode = \"branch\") #> # A tibble: 1 × 5 #>   W     X     Y     Z        f4 #>   <chr> <chr> <chr> <chr> <dbl> #> 1 AFR_1 EUR_1 NEA_1 CH_1  -741. ts_samples(ts) #> # A tibble: 58 × 3 #>    name   time pop   #>    <chr> <dbl> <chr> #>  1 NEA_1 70000 NEA   #>  2 NEA_2 40000 NEA   #>  3 EUR_1 39171 EUR   #>  4 EUR_2 39134 EUR   #>  5 EUR_3 37739 EUR   #>  6 EUR_4 36531 EUR   #>  7 EUR_5 36362 EUR   #>  8 EUR_6 35945 EUR   #>  9 EUR_7 33900 EUR   #> 10 EUR_8 33854 EUR   #> # ℹ 48 more rows"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"estimating-neanderthal-ancestry-proportions","dir":"Articles","previous_headings":"","what":"Estimating Neanderthal ancestry proportions","title":"Tree-sequence processing and statistics","text":"Let’s try put new tools practice estimate proportion Neanderthal ancestry Africans Europeans simulated data. can using Patterson’s \\(f_4\\)-ratio statistic implemented ts_f4ratio() function slendr (can find information particular version statistic Petr et al., PNAS 2019): now summarise inferred Neanderthal distribution populations, see Neanderthal ancestry Africans (expected model–Africans receive Neanderthal introgression pulse) small proportion Neanderthal ancestry Europeans (consistent 3% introgression pulse simulated ):  exactly specified model configuration , suggesting simulations work . can see quite bit noise ’s simulated small amount sequence. can also plot trajectory Neanderthal ancestry Europe time-window simulated ancient present-day DNA samples:  , result consistent empirical estimates Neanderthal ancestry using ancient DNA data (see Petr et al., PNAS 2019).","code":"# first get a table of simulated African and European individuals in the tree-sequence inds <- ts_samples(ts) %>% dplyr::filter(pop %in% c(\"AFR\", \"EUR\"))  # estimate the amounts of Neanderthal ancestry in these individuals and add # these values to the table inds$ancestry <- ts_f4ratio(ts, X = inds$name, \"NEA_1\", \"NEA_2\", \"AFR_1\", \"CH_1\")$alpha ggplot(inds, aes(pop, ancestry, fill = pop)) +   geom_boxplot() +   geom_jitter() +   labs(y = \"Neanderthal ancestry proportion\", x = \"\") +   theme(legend.position = \"none\") +   coord_cartesian(ylim = c(0, 0.1)) dplyr::filter(inds, pop == \"EUR\") %>%   ggplot(aes(time, ancestry)) +   geom_point() +   geom_smooth(method = \"lm\", linetype = 2, color = \"red\", linewidth = 0.5) +   xlim(40000, 0) + coord_cartesian(ylim = c(0, 0.1)) +   labs(x = \"time [years ago]\", y = \"Neanderthal ancestry proportion\") #> `geom_smooth()` using formula = 'y ~ x'"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"admixtools-analyses","dir":"Articles","previous_headings":"","what":"ADMIXTOOLS analyses","title":"Tree-sequence processing and statistics","text":"case like verify f-statistics results using venerable ADMIXTOOLS software (see linked paper formally introduced statistics first place), can convert tree-sequence data file format called EIGENSTRAT using ts_eigenstrat() function. file conversion internally handled R package admixr returns EIGENSTRAT object ties individual EIGENSTRAT file components together (see tutorial admixr extensive overview). admixr R package running automated ADMIXTOOLS analyses entirely R makes types analyses convenient. Running admixr analysis easy plugging object admixr function. instance, can estimate proportion Neanderthal ancestry couple individuals \\(X\\) like (admixr calls proportion alpha): fact, lets compare values obtained tskit admixr/ADMIXTOOLS individuals:  correspondence two looks good! 🎉 , note large amount variance around expected value 3% ancestry due extremely small amount sequence data simulated .","code":"snps <- ts_eigenstrat(ts, prefix = file.path(tempdir(), \"eigenstrat\", \"data\")) #> 1282 multiallelic sites (0.206% out of 621331 total) detected and removed library(admixr)  f4ratio(data = snps, X = c(\"EUR_1\", \"EUR_2\", \"AFR_2\"),         A = \"NEA_1\", B = \"NEA_2\", C = \"AFR_1\", O = \"CH_1\") #> Rows: 3 Columns: 12 #> ── Column specification ──────────────────────────────────────────────────────── #> Delimiter: \" \" #> chr (8): X1, X2, X3, X4, X5, X6, X7, X8 #> dbl (4): X9, X10, X11, X12 #>  #> ℹ Use `spec()` to retrieve the full column specification for this data. #> ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message. #> # A tibble: 3 × 8 #>   A     B     X     C     O       alpha  stderr Zscore #>   <chr> <chr> <chr> <chr> <chr>   <dbl>   <dbl>  <dbl> #> 1 NEA_1 NEA_2 EUR_1 AFR_1 CH_1  0.0218  0.00506   4.30 #> 2 NEA_1 NEA_2 EUR_2 AFR_1 CH_1  0.0328  0.00788   4.16 #> 3 NEA_1 NEA_2 AFR_2 AFR_1 CH_1  0.00408 0.00208   1.96 europeans <- inds[inds$pop == \"EUR\", ]$name  # tskit result result_ts <- ts_f4ratio(ts, X = europeans, A = \"NEA_1\", B = \"NEA_2\", C = \"AFR_1\", O = \"CH_1\") %>% select(alpha_ts = alpha)  # result obtained by admixr/ADMIXTOOLS result_admixr <- f4ratio(snps, X = europeans, A = \"NEA_1\", B = \"NEA_2\", C = \"AFR_1\", O = \"CH_1\") %>% select(alpha_admixr = alpha) #> Rows: 50 Columns: 12 #> ── Column specification ──────────────────────────────────────────────────────── #> Delimiter: \" \" #> chr (8): X1, X2, X3, X4, X5, X6, X7, X8 #> dbl (4): X9, X10, X11, X12 #>  #> ℹ Use `spec()` to retrieve the full column specification for this data. #> ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message. bind_cols(result_admixr, result_ts) %>%   ggplot(aes(alpha_ts, alpha_admixr)) +   geom_point() +   geom_abline(slope = 1, linetype = 2, color = \"red\", linewidth = 0.5) +   labs(x = \"f4-ratio statistic calculated with admixr/ADMIXTOOLS\",        y = \"f4-ratio statistic calculated with tskit\")"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"vcf-output","dir":"Articles","previous_headings":"","what":"VCF output","title":"Tree-sequence processing and statistics","text":"case need process simulated data software, can use function ts_vcf() save simulated genotypes VCF format: can also specify subset individuals saved VCF:","code":"ts_vcf(ts, path = file.path(tempdir(), \"output.vcf.gz\")) ts_vcf(ts, path = file.path(tempdir(), \"output_subset.vcf.gz\"),        individuals = c(\"CH_1\", \"NEA_1\", \"EUR_1\", \"AFR_1\"))"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"other-statistics","dir":"Articles","previous_headings":"","what":"Other statistics","title":"Tree-sequence processing and statistics","text":"follows brief overview statistics implemented tskit slendr provides easy--use R interface. see, goal functions get result using single function call, making convenient quick interactive exploratory analyses simulated data right R console. continue use simulated Neanderthal introgression tree-sequence data examples.","code":""},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"f_st","dir":"Articles","previous_headings":"Other statistics","what":"\\(F_{st}\\)","title":"Tree-sequence processing and statistics","text":"\\(F_{st}\\) statistic implemented function ts_fst(). single genome-wide \\(F_{st}\\) calculated (.e. window-based calculation), ts_fst() returns simple three-column data frame case non-named list sample sets provided, set names generated automatically: course, much less readable encourage name sample sets appropriately. case two sample sets specified, pairwise statistics computed: many statistics implemented tskit, ts_fst() accepts windows argument, specifying breakpoints windows. case, Fst column resulting data frame called “list-column”, item column vector \\(F_{st}\\) values, one per window. List-columns can little confusing new R users, highly encourage get used allow extremely concise elegant handling structured data within normal data frames (can start introduction). instance, window-based \\(F_st\\) values afr-vs-eur calculation (first row table ):","code":"ts_fst(ts, sample_sets = list(afr = c(\"AFR_1\", \"AFR_2\", \"AFR_3\"), eur = c(\"EUR_1\", \"EUR_2\"))) #> # A tibble: 1 × 3 #>   x     y        Fst #>   <chr> <chr>  <dbl> #> 1 afr   eur   0.0495 ts_fst(ts, sample_sets = list(c(\"AFR_1\", \"AFR_2\", \"AFR_3\"), c(\"EUR_1\", \"EUR_2\"))) #> # A tibble: 1 × 3 #>   x     y        Fst #>   <chr> <chr>  <dbl> #> 1 set_1 set_2 0.0495 ts_fst(ts, sample_sets = list(afr = c(\"AFR_1\", \"AFR_2\", \"AFR_3\"),                               eur = c(\"EUR_1\", \"EUR_2\"),                               nea = c(\"NEA_1\", \"NEA_2\"))) #> # A tibble: 3 × 3 #>   x     y        Fst #>   <chr> <chr>  <dbl> #> 1 afr   eur   0.0495 #> 2 afr   nea   0.555  #> 3 eur   nea   0.541 # define breakpoints between 20 windows breakpoints <- seq(0, ts$sequence_length, length.out = 21)  # calculate window-based Fst statistic win_fst <- ts_fst(   ts, windows = breakpoints,   sample_sets = list(afr = c(\"AFR_1\", \"AFR_2\", \"AFR_3\"),                      eur = c(\"EUR_1\", \"EUR_2\"),                      nea = c(\"NEA_1\", \"NEA_2\")) )  # we get 20 values for each parwise calculation win_fst #> # A tibble: 3 × 3 #>   x     y     Fst          #>   <chr> <chr> <named list> #> 1 afr   eur   <dbl [20]>   #> 2 afr   nea   <dbl [20]>   #> 3 eur   nea   <dbl [20]> win_fst[1, ]$Fst #> $`1` #>  [1] 0.04241163 0.06803338 0.04129167 0.06031210 0.05516414 0.03059425 #>  [7] 0.03320629 0.05755971 0.05271680 0.04346970 0.03659568 0.05432308 #> [13] 0.05913699 0.05905198 0.04119386 0.04389669 0.05086942 0.06181126 #> [19] 0.03929837 0.06373172"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"tajimas-d","dir":"Articles","previous_headings":"Other statistics","what":"Tajima’s \\(D\\)","title":"Tree-sequence processing and statistics","text":"function ts_tajima() nearly interface ts_fst() shown . non-window version calculated, get single genome-wide values sample set (named non-named list character vectors individual names): window-based version, function returns D column list column vectors \\(\\)-th element Tajima’s D value \\(\\)-th window:","code":"ts_tajima(ts, list(afr = c(\"AFR_1\", \"AFR_2\", \"AFR_3\"), eur = c(\"EUR_1\", \"EUR_2\"))) #> # A tibble: 2 × 2 #>   set           D #>   <chr>     <dbl> #> 1 afr   -0.000701 #> 2 eur   -0.0140 ts_tajima(ts, list(afr = c(\"AFR_1\", \"AFR_2\"), eur = c(\"EUR_1\", \"EUR_2\")), windows = breakpoints) #> # A tibble: 2 × 2 #>   set   D            #>   <chr> <named list> #> 1 afr   <dbl [20]>   #> 2 eur   <dbl [20]>"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"diversity","dir":"Articles","previous_headings":"Other statistics","what":"Diversity","title":"Tree-sequence processing and statistics","text":"can calculate diversity within given groups individuals function ts_diversity(). instance, even extremely simplified example, expect highest levels diversity Africans, followed Europeans, Neanderthals “degenerate” single individual outgroup “chimpanzee”. true? Let’s find . First extract individuals populations, creating list character vectors group (functions ts_diversity() expects input): Now can calculate diversity population sort results increasing order diversity: Great! matches expectations. simulated chimp “population” one individual, expect essentially diversity millions years evolution.","code":"# get sampled individuals from all populations sample_sets <- ts_samples(ts) %>%   split(., .$pop) %>%   lapply(function(pop) pop$name)  sample_sets #> $AFR #> [1] \"AFR_1\" \"AFR_2\" \"AFR_3\" \"AFR_4\" \"AFR_5\" #>  #> $CH #> [1] \"CH_1\" #>  #> $EUR #>  [1] \"EUR_1\"  \"EUR_2\"  \"EUR_3\"  \"EUR_4\"  \"EUR_5\"  \"EUR_6\"  \"EUR_7\"  \"EUR_8\"  #>  [9] \"EUR_9\"  \"EUR_10\" \"EUR_11\" \"EUR_12\" \"EUR_13\" \"EUR_14\" \"EUR_15\" \"EUR_16\" #> [17] \"EUR_17\" \"EUR_18\" \"EUR_19\" \"EUR_20\" \"EUR_21\" \"EUR_22\" \"EUR_23\" \"EUR_24\" #> [25] \"EUR_25\" \"EUR_27\" \"EUR_26\" \"EUR_28\" \"EUR_29\" \"EUR_30\" \"EUR_31\" \"EUR_32\" #> [33] \"EUR_33\" \"EUR_34\" \"EUR_35\" \"EUR_36\" \"EUR_37\" \"EUR_38\" \"EUR_39\" \"EUR_40\" #> [41] \"EUR_41\" \"EUR_42\" \"EUR_43\" \"EUR_44\" \"EUR_45\" \"EUR_46\" \"EUR_47\" \"EUR_48\" #> [49] \"EUR_49\" \"EUR_50\" #>  #> $NEA #> [1] \"NEA_1\" \"NEA_2\" ts_diversity(ts, sample_sets) %>% dplyr::arrange(diversity) #> # A tibble: 4 × 2 #>   set   diversity #>   <chr>     <dbl> #> 1 CH    0.0000431 #> 2 NEA   0.0000455 #> 3 EUR   0.000394  #> 4 AFR   0.000400"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"divergence","dir":"Articles","previous_headings":"Other statistics","what":"Divergence","title":"Tree-sequence processing and statistics","text":"can calculate pairwise divergence groups individuals using function ts_divergence(). Given model, expect lowest divergence two modern human groups AFR EUR, Neanderthals two modern humans, three groups (AFR, EUR NEA) equal, much deeper divergence outgroup chimpanzee CH. sorting table based value divergence column, can see results fit expectations.","code":"ts_divergence(ts, sample_sets) %>% arrange(divergence) #> # A tibble: 6 × 3 #>   x     y     divergence #>   <chr> <chr>      <dbl> #> 1 AFR   EUR     0.000449 #> 2 EUR   NEA     0.000757 #> 3 AFR   NEA     0.000780 #> 4 CH    NEA     0.00401  #> 5 CH    EUR     0.00402  #> 6 AFR   CH      0.00402"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"more-information","dir":"Articles","previous_headings":"","what":"More information","title":"Tree-sequence processing and statistics","text":"couple examples statistical functions implemented tskit provide native R interface slendr. can find tree-sequence statistics reference manual project website. statistics tskit library implemented, intend expand selection provided slendr near future. functionality like use project missing slendr, please don’t hesitate let us now creating issue GitHub page. Finally, like see examples tskit interface action, take look vignette describes switching SLiM msprime back ends slendr package.","code":""},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"introduction","dir":"Articles","previous_headings":"","what":"Introduction","title":"Spatially annotated tree sequences","text":"main selling point slendr R package programming complex spatially explicit population genetic models. use SLiM simulation engine, can store simulated data efficiently tree sequence format allows us run large population-scale simulations. previous vignettes, described can specify spatial population dynamics can access tree sequence data calculate population genetic statistics (focusing non-spatial models simplicity). Now ’s time show work simulated tree sequence spatial context.","code":""},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"model-specification","dir":"Articles","previous_headings":"","what":"Model specification","title":"Spatially annotated tree sequences","text":"Let’s first load required R libraries: begin specifying spatial model. use demographic model modern human history West Eurasia, extensively discussed introductory tutorial main landing page. complete model definition script, without comments: sanity check defined demography correctly, can plot graph summarizing population divergences geneflow events calling plot_model(model):  completeness, (slightly busy) overview spatial population ranges defined :","code":"library(slendr)  library(dplyr) library(ggplot2)  init_env()  seed <- 314159 set.seed(seed) # simulated world map map <- world(   xrange = c(-13, 70), # min-max longitude   yrange = c(18, 65),  # min-max latitude   crs = \"EPSG:3035\"    # coordinate reference system (CRS) for West Eurasia )  # couple of broad geographic regions africa <- region(   \"Africa\", map,   polygon = list(c(-18, 20), c(38, 20), c(30, 33),                  c(20, 33), c(10, 38), c(-6, 35)) ) europe <- region(   \"Europe\", map,   polygon = list(     c(-8, 35), c(-5, 36), c(10, 38), c(20, 35), c(25, 35),     c(33, 45), c(20, 58), c(-5, 60), c(-15, 50)   ) ) anatolia <- region(   \"Anatolia\", map,   polygon = list(c(28, 35), c(40, 35), c(42, 40),                  c(30, 43), c(27, 40), c(25, 38)) )  # define population histories  # African ancestral population afr <- population(   \"AFR\", time = 52000, N = 3000,   map = map, polygon = africa )  # population of the first migrants out of Africa ooa <- population(   \"OOA\", parent = afr, time = 51000, N = 500, remove = 25000,   center = c(33, 30), radius = 400e3 ) %>%   move(     trajectory = list(c(40, 30), c(50, 30), c(60, 40)),     start = 50000, end = 40000, snapshots = 20   )  # Eastern hunter-gatherers ehg <- population(   \"EHG\", parent = ooa, time = 28000, N = 1000, remove = 6000,   polygon = list(     c(26, 55), c(38, 53), c(48, 53), c(60, 53),     c(60, 60), c(48, 63), c(38, 63), c(26, 60)) )  # European population eur <- population(name = \"EUR\", parent = ehg, time = 25000, N = 2000, polygon = europe)  # Anatolian farmers ana <- population(   name = \"ANA\", time = 28000, N = 3000, parent = ooa, remove = 4000,   center = c(34, 38), radius = 500e3, polygon = anatolia ) %>%   expand_range(     by = 2500e3, start = 10000, end = 7000,     polygon = join(europe, anatolia), snapshots = 20   ) # expand the range by 2.500 km  # Yamnaya steppe population yam <- population(   name = \"YAM\", time = 7000, N = 500, parent = ehg, remove = 2500,   polygon = list(c(26, 50), c(38, 49), c(48, 50),                  c(48, 56), c(38, 59), c(26, 56)) ) %>%   move(trajectory = list(c(15, 50)), start = 5000, end = 3000, snapshots = 10)  # geneflow events gf <- list(   gene_flow(from = ana, to = yam, rate = 0.5, start = 6000, end = 5000, overlap = FALSE),   gene_flow(from = ana, to = eur, rate = 0.5, start = 8000, end = 6000),   gene_flow(from = yam, to = eur, rate = 0.75, start = 4000, end = 3000) )  # compile the spatial model model <- compile_model(   populations = list(afr, ooa, ehg, eur, ana, yam),   gene_flow = gf,   generation_time = 30, resolution = 10e3,   competition = 150e3, mating = 120e3, dispersal = 90e3 ) plot_model(model) plot_map(afr, ooa, ehg, eur, ana, yam)"},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"scheduling-sampling-events-and-simulation","dir":"Articles","previous_headings":"","what":"Scheduling sampling events and simulation","title":"Spatially annotated tree sequences","text":"Now schedule sampling single individual population every two thousand years, starting 40 thousand years ago way present (feature discussed basic tree sequence overview): Finally, can simulate data model process output tree sequence (recapitate simplify ):","code":"# one ancient individual every two thousand years ancient <- schedule_sampling(model,                     times = seq(40000, 1, by = -500),                     list(ooa, 1), list(ehg, 1), list(eur, 1),                     list(ana, 1), list(yam, 1))  # present-day Africans and Europeans present <- schedule_sampling(model, times = 0, list(afr, 5), list(eur, 30))  samples <- rbind(ancient, present) ts <- slim(   model, sequence_length = 100e3, recombination_rate = 1e-8, burnin = 200e3,   samples = samples, method = \"batch\", random_seed = 314159, max_attempts = 1 ) %>%   ts_recapitate(recombination_rate = 1e-8, Ne = 10000, random_seed = seed) %>%   ts_simplify()  ts #> ╔═════════════════════════╗ #> ║TreeSequence             ║ #> ╠═══════════════╤═════════╣ #> ║Trees          │       87║ #> ╟───────────────┼─────────╢ #> ║Sequence Length│   100000║ #> ╟───────────────┼─────────╢ #> ║Time Units     │    ticks║ #> ╟───────────────┼─────────╢ #> ║Sample Nodes   │      432║ #> ╟───────────────┼─────────╢ #> ║Total Size     │205.8 KiB║ #> ╚═══════════════╧═════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │1173│36.7 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│ 711│71.2 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │   0│ 1.2 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │ 939│35.4 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   7│ 2.7 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   3│37.8 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │   0│16 Bytes│          No║ #> ╚═══════════╧════╧════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"extracting-spatial-tree-sequence-information","dir":"Articles","previous_headings":"","what":"Extracting spatial tree sequence information","title":"Spatially annotated tree sequences","text":"showed basic tutorial, important function data exploration ts_nodes(). function extracts information individuals nodes recorded tree sequence object loaded annotated slendr : completeness, also functions ts_individuals(), ts_nodes() ts_edges() extract tree sequence tables “raw” unprocessed form, ts_nodes() much convenient data exploration analyses. First, combined information low-level tables individuals nodes single table importantly, model generated data spatial model, ts_nodes() automatically annotates node/individual tables position node space (real projected coordinates) time. means can spatial data analysis directly table returned ts_nodes(). Even better, although can see returned object belongs slendr’s class slendr_ts_nodes, internally stored spatial sf object. means can use functionality powerful R package sf well many packages geospatial analyses directly data: Typing object R console presents user-friendly summary spatio-temporal data extracted tree sequence: first part summary, see many individuals (sampled retained) nodes present tree sequence together additional useful information, including section internally stored sf object. crucial point—**can always use internal sf object spatial data directly*. data returned ts_nodes() internally transformed projected CRS used model, can use returned object data class sf. instance, beginning vignette, specified world map model represented projected CRS (EPSG 3035) can verify typing: fact ts_nodes() result just another sf object makes easy visualize overlay contents map, see .","code":"data <- ts_nodes(ts) class(data) #> [1] \"slendr\"       \"slendr_nodes\" \"sf\"           \"tbl_df\"       \"tbl\"          #> [6] \"data.frame\" data #> Simple feature collection with 939 features and 12 fields (with 11 geometries empty) #> Geometry type: POINT #> Dimension:     XY #> Bounding box:  xmin: 2426565 ymin: 257945.3 xmax: 8576618 ymax: 4878988 #> Projected CRS: ETRS89-extended / LAEA Europe #> # A tibble: 939 × 13 #>    name  pop   node_id  time time_tskit           location sampled #>    <chr> <fct>   <int> <dbl>      <dbl>        <POINT [m]> <lgl>   #>  1 AFR_1 AFR       362     0          0  (3716841 1078277) TRUE    #>  2 AFR_1 AFR       363     0          0  (3716841 1078277) TRUE    #>  3 AFR_2 AFR       364     0          0 (3541863 257945.3) TRUE    #>  4 AFR_2 AFR       365     0          0 (3541863 257945.3) TRUE    #>  5 AFR_3 AFR       366     0          0 (3366331 688429.9) TRUE    #>  6 AFR_3 AFR       367     0          0 (3366331 688429.9) TRUE    #>  7 AFR_4 AFR       368     0          0  (3329518 1360326) TRUE    #>  8 AFR_4 AFR       369     0          0  (3329518 1360326) TRUE    #>  9 AFR_5 AFR       370     0          0  (3263687 1123829) TRUE    #> 10 AFR_5 AFR       371     0          0  (3263687 1123829) TRUE    #> # ℹ 929 more rows #> # ℹ 6 more variables: remembered <lgl>, retained <lgl>, alive <lgl>, #> #   pedigree_id <dbl>, ind_id <dbl>, pop_id <int> map #> slendr 'map' object  #> -------------------  #> map: internal coordinate reference system EPSG 3035  #> spatial limits (in degrees longitude and latitude): #>   - vertical -13 ... 70 #>   - horizontal 18 ... 65"},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"using-the-simple-features-interface","dir":"Articles","previous_headings":"","what":"Using the simple features interface","title":"Spatially annotated tree sequences","text":"’s hard overstate powerful R ecosystem around sf package . However, getting familiar package geospatial analysis general can bit hurdle, especially novice users takes time get familiar many new concepts. Although many slendr features encoding programming spatial models handling simulated tree sequence data discussed far designed abstract away complexities underlying low-level details let focus problem hand, spatial data analysis unfortunately whole another matter. Luckily, data generated slendr different source spatial data great free resources disposal. bottom line : spatio-temporal data extracted tree sequences slendr different normal sf object. resource find manipulating, plotting, analysing sf data can applied slendr results well. remainder vignette look couple examples.","code":""},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"plotting-locations-of-simulated-sampled-individuals","dir":"Articles","previous_headings":"","what":"Plotting locations of simulated sampled individuals","title":"Spatially annotated tree sequences","text":"Every spatial object slendr internally class sf. flexibility ggplot2 sf packages means can overlay locations sampled individuals (saved sf format ts_nodes()) top world map (also sf object):  sf simple features objects (, extension, even slendr_spatial objects) internally stored normal data frames couple bells whistles top , powerful tools manipulating tabular data disposal. example, let’s say wanted split sampled individuals tree sequence epochs plot individually using standard ggplot2 features. simply first , adding new column specifying epoch simulated individual belong: chunk code simply adds new column epoch sf spatial data frame object called epochs . can use ggplot2 function geom_sf plot locations sampled individuals map, facet corresponding one epoch (warning can safely ignored):  hope little excursion handling slendr spatial objects (, extension, sf objects) standard data frame manipulation functions ggplot2 visualisation convinced great flexibility analysing spatial slendr data. best introduction -called “tidy” data analysis, encourage read freely-available book R Data Science.","code":"sampled_data <- ts_nodes(ts) %>% filter(sampled)  ggplot() +   geom_sf(data = map, fill = \"lightgray\", color = NA) +   geom_sf(data = sampled_data, aes(shape = pop, color = time)) +   ggtitle(\"Locations of simulated sampled individuals\") +   scale_color_continuous(type = \"viridis\") +   theme_bw() epochs <- sampled_data %>%   mutate(epoch = cut(time, breaks = c(40000, 30000, 10000, 4000, 0)),          epoch = ifelse(is.na(epoch), 0, epoch),          epoch = factor(epoch, labels = c(\"present\", \"(present, 4 ky]\", \"(4 ky, 10 ky]\",                                           \"(10 ky, 30 y]\", \"(30 ky, 40 ky]\"))) ggplot() +   geom_sf(data = map, fill = \"lightgray\", color = NA) +   geom_sf(data = epochs, aes(shape = pop, color = pop)) +   facet_wrap(~ epoch) +   ggtitle(\"Locations of simulated sampled individuals in different epochs\") +   theme_bw()"},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"extracting-spatio-temporal-ancestral-relationships","dir":"Articles","previous_headings":"","what":"Extracting spatio-temporal ancestral relationships","title":"Spatially annotated tree sequences","text":"Perhaps even useful plotting locations simulated individuals accessing locations (times) ancestors particular tree sequence node (“focal node”). Starting focal node individual, can trace geographical location nodes lineage going back way root function ts_ancestors(). record time location every individual happens ancestor least one sampled individual, means know true location every node tree sequence. simplest use case determining locations times every single node genealogical history individual along tree sequence (possible recover ancestral relationships multiple samples ): function starts given node (, name sampled diploid individual provided, two nodes), extracts information parent nodes node entire tree sequence, records locations times, proceeds one level “higher” genealogical history gather information parents parent nodes, etc., reaches root node. result process another sf object row table encodes information single branch genealogy “focal” node individual (example, \"EUR_25\"): row table, two columns location parent_location carry spatial location node (node_id) parent node (parent_id), respectively, columns time parent_time (times nodes) pop parent_pop (populations nodes belong). column connection contains sf geometry object line connecting two nodes coordinate reference system “model world”. column focal_id tells us focal node’s genealogy rows table belong , level column shows deep genealogical past branch (.e. row table) belong . table contains complete information spatio-temporal relationships nodes genealogy given focal sample. spirit demonstrating slendr tree sequence tables interact sf ggplot2 environments, let’s look immediate parent nodes two nodes sampled individual (.e. nodes level 1) using filter function R package dplyr: mentioned , three columns encoding spatial information: location parent_location carry information location child parent node (POINT class), connection object (LINESTRING class) contains line connecting two nodes (branch tree sequence also spatial connection). can plot three spatial features (two points line) individually map:  figure can see red focal node immediate parents tree sequence genealogy (coalescent sense, immediate parents individual!). case ’re surprised see two parents, recall recombination events make history encoded sample complicated can involve ancestors “move ” tree sample, just two ancestors. Looking example detail, can see one node (chromosome) individual “EUR_67” two ancestors, covering portion individuals chromosome, chromosome covered single ancestor (columns left_pos right_pos): convenient way analysis companion function ts_ancestors() called plot_ancestors(). function accepts sf object spatial branching data created ts_ancestors() plots paths nodes map leading focal node root(s) tree sequence (instead just paths immediate parents shown previous figure). case, working single diploid individual, get two sets paths nodes (chromosomes) plot two facets:  can compare result animation recapitulates simulation, presented first vignette. comparing spatial tree sequence figure animation, can immediately notice several things: spatial tree sequence paths trace ancestry single European individual back Africa. fact, also see cluster past ancestral nodes (.e. concentrated coalescent events) place Africa (OOA) migrant population settled around 40,000 thousand years ago (yellow population animation). One chromosome traces ancestry EHG population indicated green square expected programmed Yamnaya migration (descending EHG) east central Europe contribute significant part ancestry present-day Europeans (compare demographic graph top vignette). also see chromosome traces ancestry Anatolia (blue crosses). makes sense, simulated European ancestry part Anatolian. Let’s look spatial ancestry another sample. instance, know simulated history Anatolian population model much simpler. According demographic graph , Anatolians split ancestral population Eurasians Anatolia expanded wave Europe. sampled following individuals: Can see hint spatial dynamics Anatolians spatio-temporal distribution ancestral node locations one sampled individuals? Let’s pick last individual immediately plot spatial ancestry tidyverse-style using pipe operator %>%:  might expect given late age sample, position map (red crossed circle) Anatolia Europe represents one descendants migrants moved Anatolia Europe. can clearly seen position parental nodes tree sequence: nodes represent real individuals lived point past, can see , indeed, lived Anatolia.","code":"ind <- \"EUR_67\"  lineages <- ts_ancestors(ts, ind, verbose = TRUE) #> Collecting ancestors of EUR_67 [1/1]... #>  #> Generating data about spatial relationships of nodes... lineages #> Simple feature collection with 112 features and 12 fields #> Active geometry column: connection #> Geometry type: LINESTRING #> Dimension:     XY #> Bounding box:  xmin: 2736603 ymin: 607014.9 xmax: 8374222 ymax: 4762949 #> Projected CRS: ETRS89-extended / LAEA Europe #> # A tibble: 112 × 15 #>    name  pop   node_id level child_id parent_id child_time parent_time child_pop #>  * <chr> <fct>   <int> <fct>    <int>     <int>      <dbl>       <dbl> <fct>     #>  1 EUR_… EUR       406 1          406       456          0        1900 EUR       #>  2 EUR_… EUR       406 2          456       460       1900        2020 EUR       #>  3 EUR_… EUR       406 3          460       469       2020        3010 EUR       #>  4 EUR_… EUR       406 4          469       488       3010        5080 EUR       #>  5 EUR_… EUR       406 5          488       495       5080        5740 EUR       #>  6 EUR_… EUR       406 6          495       543       5740        9760 EUR       #>  7 EUR_… EUR       406 7          543       608       9760       16090 EUR       #>  8 EUR_… EUR       406 7          543       695       9760       22960 EUR       #>  9 EUR_… EUR       406 8          608       695      16090       22960 EUR       #> 10 EUR_… EUR       406 8          695       704      22960       23560 EUR       #> # ℹ 102 more rows #> # ℹ 6 more variables: parent_pop <fct>, child_location <POINT [m]>, #> #   parent_location <POINT [m]>, connection <LINESTRING [m]>, left_pos <dbl>, #> #   right_pos <dbl> filter(lineages, level == 1) #> Simple feature collection with 2 features and 12 fields #> Active geometry column: connection #> Geometry type: LINESTRING #> Dimension:     XY #> Bounding box:  xmin: 3965813 ymin: 2689630 xmax: 4683571 ymax: 2918506 #> Projected CRS: ETRS89-extended / LAEA Europe #> # A tibble: 2 × 15 #>   name   pop   node_id level child_id parent_id child_time parent_time child_pop #> * <chr>  <fct>   <int> <fct>    <int>     <int>      <dbl>       <dbl> <fct>     #> 1 EUR_67 EUR       406 1          406       456          0        1900 EUR       #> 2 EUR_67 EUR       407 1          407       451          0        1390 EUR       #> # ℹ 6 more variables: parent_pop <fct>, child_location <POINT [m]>, #> #   parent_location <POINT [m]>, connection <LINESTRING [m]>, left_pos <dbl>, #> #   right_pos <dbl> level1_branches <- ts_ancestors(ts, \"EUR_67\") %>% filter(level == 1)  ggplot() +   geom_sf(data = map, fill = \"lightgray\", color = NA) +   geom_sf(data = level1_branches[, ]$child_location, shape = 13, size = 3, color = \"red\") +   geom_sf(data = level1_branches[, ]$connection, linetype = 3) +   geom_sf(data = level1_branches[, ]$parent_location, shape = 20, color = \"blue\") +   theme_bw() +   ggtitle(\"Parent nodes (blue) of a focal individual (red)\") as_tibble(level1_branches)[, c(\"name\", \"node_id\", \"child_id\", \"parent_id\", \"left_pos\", \"right_pos\")] #> # A tibble: 2 × 6 #>   name   node_id child_id parent_id left_pos right_pos #>   <chr>    <int>    <int>     <int>    <dbl>     <dbl> #> 1 EUR_67     406      406       456        0    100000 #> 2 EUR_67     407      407       451        0    100000 ggplot() +   geom_sf(data = map) +   geom_sf(data = lineages, size = 0.5, alpha = 0.2) +   geom_sf(data = sf::st_set_geometry(lineages, \"parent_location\"),           aes(shape = parent_pop, color = parent_pop)) +   geom_sf(data = filter(ts_nodes(ts), name == ind), size = 3) +   guides(alpha = \"none\") +   coord_sf(expand = 0) +   labs(x = \"longitude\", y = \"latitude\") +   facet_grid(. ~ node_id) +   ggtitle(\"Ancestry encoded by two nodes (chromosomes) of EUR_67\") ts_samples(ts) %>% filter(pop == \"ANA\") #> # A tibble: 48 × 3 #>    name    time pop   #>    <chr>  <int> <chr> #>  1 ANA_1  27500 ANA   #>  2 ANA_2  27000 ANA   #>  3 ANA_3  26500 ANA   #>  4 ANA_4  26000 ANA   #>  5 ANA_5  25500 ANA   #>  6 ANA_6  25000 ANA   #>  7 ANA_7  24500 ANA   #>  8 ANA_8  24000 ANA   #>  9 ANA_9  23500 ANA   #> 10 ANA_10 23000 ANA   #> # ℹ 38 more rows lineages <- ts_ancestors(ts, \"ANA_45\")  ggplot() +   geom_sf(data = map) +   geom_sf(data = lineages, size = 0.5, alpha = 0.2) +   geom_sf(data = sf::st_set_geometry(lineages, \"parent_location\"),           aes(shape = parent_pop, color = parent_pop)) +   geom_sf(data = filter(ts_nodes(ts), name == \"ANA_45\"), size = 3) +   guides(alpha = \"none\") +   coord_sf(expand = 0) +   labs(x = \"longitude\", y = \"latitude\") +   facet_grid(. ~ node_id) +   ggtitle(\"Ancestry encoded by two nodes (chromosomes) of ANA_45\")"},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"calculating-distances-and-other-statistics-using-the-sf-package","dir":"Articles","previous_headings":"","what":"Calculating distances and other statistics using the sf package","title":"Spatially annotated tree sequences","text":"can summarise spatial ancestral dynamics figures using statistics? Lets take one look sf object locations times ancestral nodes sampled individuals, focusing following subset columns: can use standard dplyr table manipulation functions compute distances connected notes times separate (.e. branch lengths traditional phylogenetic sense). can use two quantities compute fast () movement ancestral individuals different time periods history sample: Let’s also convert data (absolute distances distance per generation – .e., “speed”) long format easier plotting side side: Let’s try summarise information distances “traveled” nodes different time period fitting spline (rather plotting raw data individual nodes):","code":"lineages <-   ts_samples(ts) %>%   pull(name) %>%   ts_ancestors(ts, x = .)  select(lineages, connection, child_time, parent_time) #> Simple feature collection with 20798 features and 2 fields #> Geometry type: LINESTRING #> Dimension:     XY #> Bounding box:  xmin: 2426565 ymin: 257945.3 xmax: 8576618 ymax: 4878988 #> Projected CRS: ETRS89-extended / LAEA Europe #> # A tibble: 20,798 × 3 #>                              connection child_time parent_time #>                        <LINESTRING [m]>      <dbl>       <dbl> #>  1   (8481422 2958176, 8289917 2751867)      40000       40690 #>  2   (8289917 2751867, 8110419 2813524)      40690       41710 #>  3    (8110419 2813524, 3348636 939123)      41710       67030 #>  4   (8110419 2813524, 2736603 1136489)      41710       83530 #>  5   (8110419 2813524, 3202720 1100178)      41710      203980 #>  6  (8110419 2813524, 3475143 607014.9)      41710      205270 #>  7  (8110419 2813524, 4086614 772146.9)      41710      243100 #>  8    (3348636 939123, 3202720 1100178)      67030      203980 #>  9  (2736603 1136489, 3969941 671890.4)      83530       88810 #> 10 (3969941 671890.4, 3991115 816590.1)      88810       96730 #> # ℹ 20,788 more rows distances <- lineages %>%   mutate(branch_length = abs(parent_time - child_time) / model$generation_time,          distance = sf::st_length(connection) %>% units::set_units(km) %>% as.numeric(),          speed = distance / branch_length,          epoch = cut(parent_time, breaks = c(Inf, seq(60000, 0, by = -3000)), dig.lab = 10, include.lowest = TRUE)) %>%   as_tibble() %>% # strip away the spatial annotation   select(name, pop, node_id, branch_length, distance, speed, parent_pop, parent_time, child_pop, child_time, epoch) distances_long <- distances %>%   filter(child_time < 60000) %>%   filter(!pop %in% c(\"AFR\", \"OOA\")) %>%   tidyr::pivot_longer(cols = c(distance, speed),                       names_to = \"stat\",                       values_to = \"value\") %>%   mutate(facet = case_when(     stat == \"distance\" ~ \"absolute distance of a node from parent\",     stat == \"speed\" ~ \"distance traveled by a node per generation\")) distances_long %>%   ggplot(aes(child_time, value, color = child_pop)) +   geom_smooth(method = \"loess\", aes(group = child_pop)) +   geom_hline(yintercept = 0, linetype = 2, linewidth = 0.5) +   labs(y = \"kilometers\", x = \"time [years ago]\") +   theme(axis.text.x = element_text(hjust = 1, angle = 45),         legend.position = \"bottom\") +   facet_wrap(~ facet, scales = \"free_y\") +   guides(color = guide_legend(\"ancestral node population\")) #> `geom_smooth()` using formula = 'y ~ x'"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"defining-a-model","dir":"Articles","previous_headings":"Detecting gene flow from msprime and SLiM tree sequences","what":"Defining a model","title":"Simulating data with SLiM and msprime backends","text":"now define simple non-spatial model gene flow populations (also known population admixture introgression). involve essentially procedure shown another vignette introducing non-spatial slendr models. Note different normally specify spatial model, except left map argument making population() calls. demonstrate additional features slendr interface tskit Python library, perform gene flow detection test using -called \\(f_4\\) \\(f_4\\)-ratio statistics (briefly introduces vignette). make things little interesting, define two population models: one model without gene flow, another includes gene flow. defining model expressing \\(f\\)-statistics use nomenclature used first study Patterson et al. described \\(f_4-ratio\\) statistic first place. ’re familiar statistical tests gene flow, recommend take look relevant sections linked paper. shiny_graph Let’s start first defining populations splits established figure : Note Ne populations x1 x2 set much higher rest. set \\(N_e\\) populations lower values speed forward SLiM simulations (won’t affect results ’re interested anyway). Higher values \\(N_e\\) x1 x2 populations ensure effect drift acting two populations much smaller. simulate later measure proportion ancestry population b x1, ensure ancestry proportion drift far away expectation make interesting patterns stand clearly. (course, done demonstration purposes speed SLiM simulations making \\(N_e\\) populations smaller. practice, running kinds simulations using msprime back end.)","code":"seq_len <- 100e6 # amount of sequence to simulate rec_rate <- 1e-8 # uniform recombination rate mut_rate <- 1e-8 # mutation rate  o <- population(\"outgroup\", time = 1, N = 100) c <- population(\"c\", time = 2500, N = 100, parent = o) a <- population(\"a\", time = 3000, N = 100, parent = c) b <- population(\"b\", time = 3500, N = 100, parent = a) x1 <- population(\"x1\", time = 3800, N = 5000, parent = c) x2 <- population(\"x2\", time = 4000, N = 5000, parent = x1)"},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"compiling-the-model-and-simulating-data","dir":"Articles","previous_headings":"","what":"Compiling the model and simulating data","title":"Simulating data with SLiM and msprime backends","text":"now use populations compile two models: first model without gene flow (left panel figure ), second model include 10% gene flow b x1 (right panel figure ). also schedule sampling couple individuals end simulation: 50 individuals populations x1 x2 capture bit natural variation b ancestry x1, one individual rest. running simulations, let’s first make sure models set correctly:   Now run models (without gene flow) two slendr backends, SLiM msprime: Note using exactly model configuration object simulation runs! fact, even function interface looks nearly exactly . doesn’t matter specific details demographic models , slendr interpret correctly regardless back end simulation engine choose use.","code":"# no gene flow model model_nogf <- compile_model(populations = list(a, b, x1, x2, c, o), generation_time = 1, simulation_length = 4500)  samples <- schedule_sampling(   model_nogf, times = 4500,   list(a, 1), list(b, 1), list(x1, 50), list(x2, 50), list(c, 1), list(o, 1) )  # model with gene flow gf <- gene_flow(from = b, to = x1, start = 4100, end = 4400, rate = 0.1)  model_gf <- compile_model(populations = list(a, b, x1, x2, c, o), gene_flow = gf, generation_time = 1, simulation_length = 4500)  samples <- schedule_sampling(   model_gf, times = 4500,   list(a, 1), list(b, 1), list(x1, 50), list(x2, 50), list(c, 1), list(o, 1) ) plot_model(model_nogf, sizes = FALSE) plot_model(model_gf, sizes = FALSE, proportions = TRUE) # model without gene flow slim_nogf <- slim(model_nogf, sequence_length = seq_len, recombination_rate = rec_rate, samples = samples, random_seed = seed) msprime_nogf <- msprime(model_nogf, sequence_length = seq_len, recombination_rate = rec_rate, samples = samples, random_seed = seed)  # model with b -> x1 gene flow slim_gf <- slim(model_gf, sequence_length = seq_len, recombination_rate = rec_rate, samples = samples, random_seed = seed) msprime_gf <- msprime(model_gf, sequence_length = seq_len, recombination_rate = rec_rate, samples = samples, random_seed = seed)"},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"comparing-the-outputs-of-msprime-and-slim-runs-of-a-slendr-model","dir":"Articles","previous_headings":"Compiling the model and simulating data","what":"Comparing the outputs of msprime and SLiM runs of a slendr model","title":"Simulating data with SLiM and msprime backends","text":"run introgression models two simulation back ends, let’s load SLiM msprime tree sequence outputs compare contents (hoping !). can see , tree sequence summary data loaded processed slendr’s ts_load() function (data produced SLiM backend) similar got “manually” msprime-produced tree sequence using custom defined functions. somehow obvious nice cheap sanity check indicating tree sequence data structure produced two backends demographic model almost .","code":"# SLiM outputs -- we can use built-in slendr functions for those slim_nogf <-   slim_nogf %>%   ts_recapitate(Ne = 10, recombination_rate = rec_rate, random_seed = seed) %>%   ts_mutate(mut_rate, random_seed = seed)  slim_gf <-   slim_gf %>%   ts_recapitate(Ne = 10, recombination_rate = rec_rate, random_seed = seed) %>%   ts_mutate(mut_rate, random_seed = seed)  # msprime outputs (note that recapitation and simplification doesn't make # sense here because we already have fully coalesced genealogies for our # individuals  of interest msprime_nogf <- ts_mutate(msprime_nogf, mut_rate, random_seed = seed) msprime_gf <- ts_mutate(msprime_gf, mut_rate, random_seed = seed) slim_nogf #> ╔═════════════════════════╗ #> ║TreeSequence             ║ #> ╠═══════════════╤═════════╣ #> ║Trees          │   248158║ #> ╟───────────────┼─────────╢ #> ║Sequence Length│100000000║ #> ╟───────────────┼─────────╢ #> ║Time Units     │    ticks║ #> ╟───────────────┼─────────╢ #> ║Sample Nodes   │    21008║ #> ╟───────────────┼─────────╢ #> ║Total Size     │ 70.6 MiB║ #> ╚═══════════════╧═════════╝ #> ╔═══════════╤══════╤════════╤════════════╗ #> ║Table      │Rows  │Size    │Has Metadata║ #> ╠═══════════╪══════╪════════╪════════════╣ #> ║Edges      │972815│29.7 MiB│          No║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Individuals│131750│12.6 MiB│         Yes║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Migrations │     0│ 8 Bytes│          No║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Mutations  │257348│ 9.1 MiB│          No║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Nodes      │157230│ 5.7 MiB│         Yes║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Populations│     7│ 2.8 KiB│         Yes║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Provenances│     3│38.0 KiB│          No║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Sites      │257040│ 6.1 MiB│          No║ #> ╚═══════════╧══════╧════════╧════════════╝ msprime_nogf #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │      62745║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        208║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   14.4 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │247454│  7.6 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│   104│  2.9 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │ 69172│  2.4 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │ 32685│893.7 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     6│414 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     2│  3.9 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │ 69144│  1.6 MiB│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"comparing-results-of-population-genetics-statistics","dir":"Articles","previous_headings":"Compiling the model and simulating data","what":"Comparing results of population genetics statistics","title":"Simulating data with SLiM and msprime backends","text":"loaded msprime SLiM tree sequences generated model, let’s see can get comparable results calculate statistics interest. Let’s first extract individuals populations x1 x2 estimate values \\(f_4(c, x1 \\textrm{ } x2; b, o)\\) (significantly negative individuals population x1 due ’s ancestry coming b introgression models consistent zero x2) \\(f_4\\textrm{-ratio}(, b, \\textrm{ } x2; c, o)\\), estimates proportion b-like ancestry x1 x2: Now, let’s repeat analysis tree sequences produced msprime backend two slendr demographic models: Finally can proceed plotting SLiM msprime backend results two models compare together. noted , despite fact SLiM forward simulator msprime coalescent simulator, population genetic expectations gene flow gene flow models match closely (barring level uncertainty expected due randomness population genetic processes play).   Things seem looking good behave expected! couple observations: model without gene flow, \\(f_4\\) statistic value zero x1 x2 populations. expected, \\(f_4\\) ~0 consistent relationship lineages conforming standard phylogenetic tree (admixture edges). Essentially, \\(f_4\\) hypothesis test “tree-ness” data. model includes gene flow b x1, \\(f_4\\) value significantly negative. means simple tree hypothesis can rejected gene flow must occurred b x1. Note test reveal direction gene flow, ’s presence. Consistently \\(f_4\\) results, \\(f_4-ratio\\) estimates match expectations. Specifically, statistic estimates around 10% ancestry b x1 gene flow model 0% ancestry cases. importantly, results obtained SLiM msprime back ends nearly exactly . means (non-spatial models) can use SLiM msprime backend interchangeably.","code":"# extract vector of names of the \"test individuals\" in populations `x1` and `x2` X <- ts_samples(slim_gf) %>% filter(pop %in% c(\"x1\", \"x2\")) %>% pull(name) X #>   [1] \"x1_1\"  \"x1_2\"  \"x1_3\"  \"x1_4\"  \"x1_5\"  \"x1_6\"  \"x1_7\"  \"x1_8\"  \"x1_9\"  #>  [10] \"x1_10\" \"x1_11\" \"x1_12\" \"x1_13\" \"x1_14\" \"x1_15\" \"x1_16\" \"x1_17\" \"x1_18\" #>  [19] \"x1_19\" \"x1_20\" \"x1_21\" \"x1_22\" \"x1_23\" \"x1_24\" \"x1_25\" \"x1_26\" \"x1_27\" #>  [28] \"x1_28\" \"x1_29\" \"x1_30\" \"x1_31\" \"x1_32\" \"x1_33\" \"x1_34\" \"x1_35\" \"x1_36\" #>  [37] \"x1_37\" \"x1_38\" \"x1_39\" \"x1_40\" \"x1_41\" \"x1_42\" \"x1_43\" \"x1_44\" \"x1_45\" #>  [46] \"x1_46\" \"x1_47\" \"x1_48\" \"x1_49\" \"x1_50\" \"x2_1\"  \"x2_2\"  \"x2_3\"  \"x2_4\"  #>  [55] \"x2_5\"  \"x2_6\"  \"x2_7\"  \"x2_8\"  \"x2_9\"  \"x2_10\" \"x2_11\" \"x2_12\" \"x2_13\" #>  [64] \"x2_14\" \"x2_15\" \"x2_16\" \"x2_17\" \"x2_18\" \"x2_19\" \"x2_20\" \"x2_21\" \"x2_22\" #>  [73] \"x2_23\" \"x2_24\" \"x2_25\" \"x2_26\" \"x2_27\" \"x2_28\" \"x2_29\" \"x2_30\" \"x2_31\" #>  [82] \"x2_32\" \"x2_33\" \"x2_34\" \"x2_35\" \"x2_36\" \"x2_37\" \"x2_38\" \"x2_39\" \"x2_40\" #>  [91] \"x2_41\" \"x2_42\" \"x2_43\" \"x2_44\" \"x2_45\" \"x2_46\" \"x2_47\" \"x2_48\" \"x2_49\" #> [100] \"x2_50\" # calculate f4-statistics on individuals of `x1` and `x2` populations using data # from the two models (a model with no gene flow and a gene flow model) -- we use # map_dfr to iterate across all individuals from `X_individuals` and binding all # resulting data frames into a single data frame df_slim_f4 <- rbind(   map_dfr(X, ~ ts_f4(slim_nogf, \"c_1\", .x, \"b_1\", \"outgroup_1\")) %>% mutate(model = \"no gene flow\"),   map_dfr(X, ~ ts_f4(slim_gf, \"c_1\", .x, \"b_1\", \"outgroup_1\")) %>% mutate(model = \"gene flow\") ) %>%   select(X, f4, model) %>%   mutate(simulator = \"SLiM backend\")  # compute the proportions of `b` ancestry in `x1` (expected 10%) and `x2` # (expected 0% because this population did not receive any gene flow from `b`) df_slim_f4ratio <- rbind(   ts_f4ratio(slim_nogf, X, \"a_1\", \"b_1\", \"c_1\", \"outgroup_1\") %>% mutate(model = \"no gene flow\"),   ts_f4ratio(slim_gf, X, \"a_1\", \"b_1\", \"c_1\", \"outgroup_1\") %>% mutate(model = \"gene flow\") ) %>%   select(X, alpha, model) %>%   mutate(simulator = \"SLiM backend\") df_msprime_f4 <- rbind(   map_dfr(X, ~ ts_f4(msprime_nogf, \"c_1\", .x, \"b_1\", \"outgroup_1\")) %>% mutate(model = \"no gene flow\"),   map_dfr(X, ~ ts_f4(msprime_gf, \"c_1\", .x, \"b_1\", \"outgroup_1\")) %>% mutate(model = \"gene flow\") ) %>%   select(X, f4, model) %>%   mutate(simulator = \"msprime backend\")  # compute the proportions of `b` ancestry in `x1` (expected 10%) and `x2` # (expected 0% because this population did not receive any gene flow from `b`) df_msprime_f4ratio <- rbind(   ts_f4ratio(msprime_nogf, X, \"a_1\", \"b_1\", \"c_1\", \"outgroup_1\") %>% mutate(model = \"no gene flow\"),   ts_f4ratio(msprime_gf, X, \"a_1\", \"b_1\", \"c_1\", \"outgroup_1\") %>% mutate(model = \"gene flow\") ) %>%   select(X, alpha, model) %>%   mutate(simulator = \"msprime backend\") df_f4 <- rbind(df_slim_f4, df_msprime_f4) %>%   mutate(population = ifelse(grepl(\"x1_\", X),                              \"x1 (received gene flow)\",                              \"x2 (no gene flow)\"))  ggplot(df_f4, aes(f4, fill = population)) +   geom_histogram(bins = 50) +   facet_grid(simulator ~ model) +   geom_vline(xintercept = 0, linetype = 2) +   labs(y = \"number of individuals\", x = \"f4 statistic\",        title = \"f4(c, x1 or x2; b, outgroup)\",        subtitle = \"f4 ~0 is consistent with no gene flow, negative value indicates gene flow with 'b'\") +   theme(legend.position = \"bottom\") df_f4ratio <- rbind(df_slim_f4ratio, df_msprime_f4ratio) %>%   mutate(population = ifelse(grepl(\"x1_\", X),                              \"x1 (received gene flow)\",                              \"x2 (no gene flow)\"))  ggplot(df_f4ratio, aes(alpha, fill = population)) +   geom_histogram(bins = 30) +   facet_grid(simulator ~ model) +   geom_vline(xintercept = 0.1, linetype = 2) +   labs(y = \"number of individuals\", x = \"ancestry proportion (f4-ratio statistic)\",        title = \"f4-ratio estimate of 'b' ancestry calculated from simulated data\",        subtitle = \"f4-ratio = f4(a, outgroup; x1 or x2, c) / f4(a, outgroup; b, c)\") +   theme(legend.position = \"bottom\")"},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"computing-allele-frequency-spectra","dir":"Articles","previous_headings":"","what":"Computing allele frequency spectra","title":"Simulating data with SLiM and msprime backends","text":"Let’s look one example. Imagine following five demographic models describing changes population size abstract population:  models population size change events , depending specifics project, defined using different time units. work radiocarbon-dated samples ancient DNA might used thinking population history units “years present”—case, “present” time (bottom figure) time 0 earlier events specified “5000 years present”, etc. hand, interested modelling theoretical population, might want concern quite happy starting simulation “generation 1” continue specified “generation X”. slendr package makes extremely easy, can specify times whatever direction units want, long kept consistent populations events model. rather unique feature among popgen simulation software programs either make express model “generations forward direction”, explicitly convert time generations going backwards. previous section vignette shown SLiM msprime can used simulation backend. remainder vignette present additional analysis, demonstrating fact can express time models forward backward direction still run models slendr’s SLiM (forward simulator) msprime (backward simulator). Specifically, define five demographic models figure twice, different orientations time – forward backward. compute allele frequency spectrum simulation (five models, forward backward) verify forward backward pairs give result regardless whether use SLiM back end msprime backend.","code":""},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"model-definition","dir":"Articles","previous_headings":"Computing allele frequency spectra","what":"Model definition","title":"Simulating data with SLiM and msprime backends","text":"First, let’s define five population histories forward time direction: Now let’s flip flow time around define population histories units “time ago”: convenience, write helper function , given population history object (ten objects just ), compile slendr model, run SLiM msprime back ends, load tree sequence, compute allele frequency spectrum (AFS) . save us lots code repetition (always good thing), minimize chance copy-pasting bugs (definitely good thing!) compare AFS patterns across different models two slendr simulation back ends (according population genetics theory). Now function doesn’t anything special. simply wraps standard slendr simulation pipeline (1. create population(s); 2. compile model object; 3. simulate model; 4. load results; 5. analyse data) single function. Now whole pipeline, can generate simulated data bind result single data frame: Finally, can plot allele frequency spectra models, two time directions, two simulation back ends:  , couple things note: two models involve population contraction (single step exponential collapse), see decrease low frequency variants. hand, models population expansion show high proportion low frequency variants. model constant population size shows intermediate pattern. nice validation simulation works expected exactly predicted population genetics theory. forward backward time models give exactly result. unexpected anything else bug slendr. However, although unsurprising, hope makes clear free use whatever time unit specification want encode slendr models. non-spatial models, results obtained msprime SLiM back ends consistent one another. , makes sense although completely different software, population genetic theory governing shape allele frequency spectra applies equally . fact, example just implemented important part slendr unit test suite verifies simulations created slendr correct.","code":"N <- 1000 N_factor <- 5 # by what factor should Ne change  seq_len <- 50e6 rec_rate <- 1e-8 mut_rate <- 1e-8 # constant Ne model forward_const <- population(\"const\", time = 1, N = N)  # decreasing step Ne model forward_decr <- population(\"decr\", time = 1, N = N, map = FALSE) %>%   resize(time = 2000, N = N / N_factor, how = \"step\")  # increasing step Ne model forward_incr <- population(\"inc\", time = 1, N = N) %>%   resize(time = 2000, N = N * N_factor, how = \"step\")  # exponential increase in size forward_exp_incr <- population(\"exp_inc\", time = 1, N = N) %>%   resize(time = 2000, end = 3000, N = N * N_factor, how = \"exponential\")  # exponential decrease in size forward_exp_decr <- population(\"exp_decr\", time = 1, N = N) %>%   resize(time = 2000, end = 3000, N = N / N_factor, how = \"exponential\") # constant Ne model backward_const <- population(\"const\", time = 5000, N = N)  # decreasing step Ne model backward_decr <- population(\"decr\", time = 5000, N = N) %>%   resize(time = 3000, N = N / N_factor, how = \"step\")  # increasing step Ne model backward_incr <- population(\"inc\", time = 5000, N = N) %>%   resize(time = 3000, N = N * N_factor, how = \"step\")  # exponential increase in size backward_exp_incr <- population(\"exp_inc\", time = 5000, N = N) %>%   resize(time = 3000, end = 2000, N = N * N_factor, how = \"exponential\")  # exponential decrease in size backward_exp_decr <- population(\"exp_decr\", time = 5000, N = N) %>%   resize(time = 3000, end = 2000, N = N / N_factor, how = \"exponential\") compile_run_afs <- function(model_name, pop, seed = 42) {   # maximum length of the simulation (necessary for forward models which start   # in generation 1)   simulation_length <- 5000    # define sampling times given the direction of time   if (attr(pop, \"history\")[[1]]$time == 1) {     sampling_time <- simulation_length     direction <- \"forward\"   } else {     sampling_time <- 0     direction <- \"backward\"   }    # compile model   model <- compile_model(pop, generation_time = 15, direction = direction, simulation_length = simulation_length)    samples <- schedule_sampling(model, times = sampling_time, list(pop, 50))    # run the model in SLiM   ts_slim <- slim(model, sequence_length = seq_len, recombination_rate = rec_rate,                   samples = samples, random_seed = seed, verbose = FALSE)    # run the same model in msprim   ts_msprime <- msprime(model, sequence_length = seq_len, recombination_rate = rec_rate,                         samples = samples, random_seed = seed, verbose = FALSE)    # load the SLiM tree sequence   ts_slim <- ts_recapitate(ts_slim, Ne = N, recombination_rate = rec_rate, random_seed = seed) %>%     ts_mutate(mut_rate, random_seed = seed)    # load the msprime tree sequence   ts_msprime <- ts_mutate(ts_msprime, mut_rate, random_seed = seed)    # compute the AFS from the SLiM and msprime tree sequences and bind the   # results (derived allele counts per frequency bin) in a data frame   msprime_afs <- ts_afs(ts_msprime, polarised = TRUE)[-1]   slim_afs <- ts_afs(ts_slim, polarised = TRUE)[-1]    rbind(     data.frame(simulator = \"msprime\", model = model_name, f = msprime_afs),     data.frame(simulator = \"SLiM\", model = model_name, f = slim_afs)   ) %>%     group_by(simulator, model) %>%     mutate(n = 1:n(), direction = direction) %>%     ungroup() } afs <- bind_rows(   compile_run_afs(\"constant\", forward_const),   compile_run_afs(\"constant\", backward_const),   compile_run_afs(\"step contraction\", forward_decr),   compile_run_afs(\"step contraction\", backward_decr),   compile_run_afs(\"step increase\", forward_incr),   compile_run_afs(\"step increase\", backward_incr),   compile_run_afs(\"exponential decrease\", forward_exp_decr),   compile_run_afs(\"exponential decrease\", backward_exp_decr),   compile_run_afs(\"exponential increase\", forward_exp_incr),   compile_run_afs(\"exponential increase\", backward_exp_incr) ) %>%   mutate(model = factor(     model, levels = c(\"step contraction\", \"constant\", \"step increase\",                       \"exponential decrease\", \"exponential increase\"))   ) ggplot(afs, aes(n, f, color = direction, linetype = simulator)) +   geom_line(stat = \"identity\") +   facet_wrap(~ model) +   labs(x = \"number of derived alleles\", y = \"frequency\",        title = \"Site frequency spectra obtained from five demographic models\",        subtitle = \"Each model was specified in forward or backward direction of time and executed by two different backend scripts in slendr (SLiM and msprime)\") +   guides(color = guide_legend(\"direction of\\ntime in slendr\"),          linetype = guide_legend(\"slendr backend\\nengine used\")) +   scale_linetype_manual(values = c(3, 2)) +   scale_x_continuous(breaks = c(1, seq(20, 100, 20)), limits = c(1, 100)) +   theme(legend.position = \"bottom\")"},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"conclusion","dir":"Articles","previous_headings":"","what":"Conclusion","title":"Simulating data with SLiM and msprime backends","text":"example shows standard, non-spatial demographic model like implement simulate R, slendr gives way efficiently using ’s msprime back end addition SLiM back end used throughout documentation.","code":""},{"path":"https://www.slendr.net/articles/vignette-08-nonslendr-tskit.html","id":"non-spatial-slim-tree-sequences","dir":"Articles","previous_headings":"","what":"Non-spatial SLiM tree sequences","title":"Analyzing non-slendr tree sequences","text":"Consider following SLiM script, creates couple populations (different \\(N_e\\)) splitting ancestral population p1 (lets save /tmp/nonspatial.slim): run script SLiM, can use slendr load output tree sequence (saved /tmp/nonspatial-slim.trees), simplify , overlay mutations using standard functionality originally developed slendr tree sequences. Note command use loading slendr tree sequences, except direct ts_load() function straight tree-sequence output file rather using ts_load(<model-object>) format used working standard slendr simulations. way, slendr can extract information individual’s names, nodes, population assignments, etc. just slendr tree sequence function ts_nodes(). standard slendr models, function loads “raw” node individual tree-sequences tables, performs couple join operations, presents whole thing nice unified form interactive data analysis (can also include spatial information—see ): Moving tskit statistics, can use data table extract list nodes belonging population (various tskit tree-sequence statistics operate , slendr follows design). computing nucleotide diversity four populations using ts_diversity() function, first creating list lists node IDs (.e. chromosomes) individuals population: Just slendr tree sequences (demonstrated paper) can get individual trees , extracted phylogenetic format provided ape R package. first simplify tree sequence even just 10 nodes make things manageable: R tree object, can use packages like ggtree visualize tree (phylogenetic package work ). Note nodes ‘ape phylo’ trees must conform strict format (must labelled 1...N), extract information node IDs tskit tree-sequence data able plot tree.","code":"initialize() {     setSeed(42);     initializeTreeSeq();     initializeMutationRate(0);     initializeMutationType(\"m1\", 0.5, \"f\", 0.0);     initializeGenomicElementType(\"g1\", m1, 1.0);     initializeGenomicElement(g1, 0, 1e6);     initializeRecombinationRate(1e-8); }  1 early() {     sim.addSubpop(\"p1\", 10); }  1000 early() {     sim.addSubpopSplit(\"p2\", 500, p1); }  3000 early() {     sim.addSubpopSplit(\"p3\", 2500, p1); }  5000 early() {     sim.addSubpopSplit(\"p4\", 10000, p1); }  6000 late() {     sim.treeSeqOutput(\"/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmpx99kMg/filee61c7b568e93\"); } ts <- ts_load(nonspatial_trees_file) %>%   ts_simplify() %>%   ts_mutate(mutation_rate = 1e-7, random_seed = SEED) data <- ts_nodes(ts) %>% dplyr::filter(sampled) data #> # A tibble: 26,020 × 11 #>    pop   node_id  time time_tskit sampled remembered retained alive pedigree_id #>    <chr>   <int> <dbl>      <dbl> <lgl>   <lgl>      <lgl>    <lgl>       <dbl> #>  1 p1          0     0          0 TRUE    FALSE      FALSE    TRUE     20073000 #>  2 p1          1     0          0 TRUE    FALSE      FALSE    TRUE     20073000 #>  3 p1          2     0          0 TRUE    FALSE      FALSE    TRUE     20073001 #>  4 p1          3     0          0 TRUE    FALSE      FALSE    TRUE     20073001 #>  5 p1          4     0          0 TRUE    FALSE      FALSE    TRUE     20073002 #>  6 p1          5     0          0 TRUE    FALSE      FALSE    TRUE     20073002 #>  7 p1          6     0          0 TRUE    FALSE      FALSE    TRUE     20073003 #>  8 p1          7     0          0 TRUE    FALSE      FALSE    TRUE     20073003 #>  9 p1          8     0          0 TRUE    FALSE      FALSE    TRUE     20073004 #> 10 p1          9     0          0 TRUE    FALSE      FALSE    TRUE     20073004 #> # ℹ 26,010 more rows #> # ℹ 2 more variables: ind_id <dbl>, pop_id <int> sample_sets <- split(data$node_id, data$pop)  # compute nucleotide diversity in each population # (any other ts_*() tskit R interface function should work) ts_diversity(ts, sample_sets) #> # A tibble: 4 × 2 #>   set    diversity #>   <chr>      <dbl> #> 1 p1    0.00000755 #> 2 p2    0.000241   #> 3 p3    0.000463   #> 4 p4    0.000201 samples <- sample(data$node_id, 10) ts_small <- ts_simplify(ts, simplify_to = samples)  # extract the 42nd tree in the genealogy to an R 'phylo' format tree <- ts_phylo(ts_small, 42 - 1) #> Starting checking the validity of tree... #> Found number of tips: n = 10  #> Found number of nodes: m = 9  #> Done. tree #>  #> Phylogenetic tree with 10 tips and 9 internal nodes. #>  #> Tip labels: #>   1, 0, 9, 4, 8, 7, ... #> Node labels: #>   45, 28, 22, 31, 35, 36, ... #>  #> Rooted; includes branch lengths. library(ggtree) #> ggtree v3.10.0 For help: https://yulab-smu.top/treedata-book/ #>  #> If you use the ggtree package suite in published research, please cite #> the appropriate paper(s): #>  #> Guangchuang Yu, David Smith, Huachen Zhu, Yi Guan, Tommy Tsan-Yuk Lam. #> ggtree: an R package for visualization and annotation of phylogenetic #> trees with their covariates and other associated data. Methods in #> Ecology and Evolution. 2017, 8(1):28-36. doi:10.1111/2041-210X.12628 #>  #> Guangchuang Yu.  Data Integration, Manipulation and Visualization of #> Phylogenetic Trees (1st edition). Chapman and Hall/CRC. 2022, #> doi:10.1201/9781003279242 #>  #> LG Wang, TTY Lam, S Xu, Z Dai, L Zhou, T Feng, P Guo, CW Dunn, BR #> Jones, T Bradley, H Zhu, Y Guan, Y Jiang, G Yu. treeio: an R package #> for phylogenetic tree input and output with richly annotated and #> associated data. Molecular Biology and Evolution. 2020, 37(2):599-603. #> doi: 10.1093/molbev/msz240  #>  #> Attaching package: 'ggtree' #> The following object is masked from 'package:magrittr': #>  #>     inset  labels <- ts_nodes(tree) %>% select(node = phylo_id, tskit_id = node_id)  ggtree(tree, branch.length = \"none\") %<+% labels +   geom_label(aes(label = tskit_id)) library(ape) plot(tree, show.tip.label = FALSE) nodelabels() tiplabels()"},{"path":"https://www.slendr.net/articles/vignette-08-nonslendr-tskit.html","id":"msprime-non-slendr-tree-sequences","dir":"Articles","previous_headings":"","what":"msprime (non-slendr) tree sequences","title":"Analyzing non-slendr tree sequences","text":"applies also msprime tree sequences (really surprising, given ’s tskit hood). can start Python: can proceed loading msprime tree sequence R analyze slendr functionality:","code":"import msprime ts = msprime.sim_ancestry(100) ts.dump(<trees file>) ts <- ts_load(msprime_trees_file)  ts_nodes(ts) #> # A tibble: 199 × 7 #>    pop   ind_id node_id  time time_tskit sampled pop_id #>    <chr>  <dbl>   <int> <dbl>      <dbl> <lgl>    <int> #>  1 0         NA       0     0          0 TRUE         0 #>  2 0         NA       1     0          0 TRUE         0 #>  3 0         NA       2     0          0 TRUE         0 #>  4 0         NA       3     0          0 TRUE         0 #>  5 0         NA       4     0          0 TRUE         0 #>  6 0         NA       5     0          0 TRUE         0 #>  7 0         NA       6     0          0 TRUE         0 #>  8 0         NA       7     0          0 TRUE         0 #>  9 0         NA       8     0          0 TRUE         0 #> 10 0         NA       9     0          0 TRUE         0 #> # ℹ 189 more rows"},{"path":"https://www.slendr.net/articles/vignette-08-nonslendr-tskit.html","id":"spatial-slim-non-slendr-tree-sequences","dir":"Articles","previous_headings":"","what":"Spatial SLiM (non-slendr) tree sequences","title":"Analyzing non-slendr tree sequences","text":"Furthermore, generalized interface also supports slendr’s spatial tree-sequence features, bells whistles. instance, lets take following spatial SLiM script (modified SLiM manual) execute SLiM usual way: can load simplify output tree sequence just vignette (anywhere slendr documentation): Finally, can access spatio-temporal data embedded output tree sequence standard slendr way (note spatial sf column location POINT data type): get tree sequence converted spatial sf data format, can use standard geospatial packages use spatial data analysis methods packages provide. briefly demonstrate means, can trivially plot location recorded node:  can also collect spatio-temporal ancestry information particular node (.e. times locations ancestors way root, “link” plot signifying parent-child edge somewhere along tree sequence) plot 2D surface (x y dimensions [0, 1]). plot little chaotic, hopefully conveys idea (“focal node” 0 highlighted red). essentially plot last figure paper.","code":"initialize() {     setSeed(42);       initializeSLiMOptions(keepPedigrees=T, dimensionality=\"xy\");       initializeTreeSeq();       initializeMutationRate(1e-7);       initializeMutationType(\"m1\", 0.5, \"f\", 0.0);       initializeGenomicElementType(\"g1\", m1, 1.0);       initializeGenomicElement(g1, 0, 1e6);       initializeRecombinationRate(1e-8); } 1 early() {     sim.addSubpop(\"p1\", 500);      // initial positions are random in ([0,1], [0,1])     p1.individuals.x = runif(p1.individualCount);       p1.individuals.y = runif(p1.individualCount); } modifyChild() {       // draw a child position near the first parent, within bounds       do child.x = parent1.x + rnorm(1, 0, 0.02);       while ((child.x < 0.0) | (child.x > 1.0));        do child.y = parent1.y + rnorm(1, 0, 0.02);       while ((child.y < 0.0) | (child.y > 1.0));        return T; } 1: late() {     sim.treeSeqRememberIndividuals(sim.subpopulations.individuals, permanent = F); }  10000 late() {     sim.treeSeqOutput(\"/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmpx99kMg/filee61c27e4c5b7\"); } ts <- ts_load(spatial_trees_file) %>% ts_simplify() data <- ts_nodes(ts) data #> Simple feature collection with 1941 features and 11 fields #> Geometry type: POINT #> Dimension:     XY #> Bounding box:  xmin: 0.02032986 ymin: 0.001180405 xmax: 0.9990375 ymax: 0.9979969 #> CRS:           NA #> # A tibble: 1,941 × 12 #>    pop   node_id  time time_tskit               location sampled remembered #>    <chr>   <int> <dbl>      <dbl>                <POINT> <lgl>   <lgl>      #>  1 p1          0     0          0  (0.3158008 0.8489413) TRUE    FALSE      #>  2 p1          1     0          0  (0.3158008 0.8489413) TRUE    FALSE      #>  3 p1          2     0          0  (0.5946915 0.5792152) TRUE    FALSE      #>  4 p1          3     0          0  (0.5946915 0.5792152) TRUE    FALSE      #>  5 p1          4     0          0  (0.8320243 0.5595862) TRUE    FALSE      #>  6 p1          5     0          0  (0.8320243 0.5595862) TRUE    FALSE      #>  7 p1          6     0          0  (0.5532792 0.1833623) TRUE    FALSE      #>  8 p1          7     0          0  (0.5532792 0.1833623) TRUE    FALSE      #>  9 p1          8     0          0 (0.7333698 0.01834573) TRUE    FALSE      #> 10 p1          9     0          0 (0.7333698 0.01834573) TRUE    FALSE      #> # ℹ 1,931 more rows #> # ℹ 5 more variables: retained <lgl>, alive <lgl>, pedigree_id <dbl>, #> #   ind_id <dbl>, pop_id <int> ggplot() + geom_sf(data = data, aes(color = time), alpha = 0.5) ancestral_links <- ts_ancestors(ts, 0)  ggplot() +     geom_sf(data = ancestral_links, size = 0.5, aes(alpha = parent_time)) +     geom_sf(data = sf::st_set_geometry(ancestral_links, \"parent_location\"), aes(color = parent_time)) +     geom_sf(data = data[data$node_id == 0, ], size = 3, color = \"red\")"},{"path":"https://www.slendr.net/articles/vignette-08-nonslendr-tskit.html","id":"conclusion","dir":"Articles","previous_headings":"","what":"Conclusion","title":"Analyzing non-slendr tree sequences","text":"vignette gave brief overview using slendr’s R-tskit interface loading, processing, analyzing “pure” non-slendr tree sequences produced msprime SLiM scripts. Although touched upon basic features R-tskit interface standard tree sequences, important note far slendr concerned, matter tree sequence produced, long conforms tskit specification. means regardless source tree sequence data, able use slendr’s tskit functionality run analyses.","code":""},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"note","dir":"Articles","previous_headings":"","what":"Note","title":"Examples from the slendr paper","text":"Since publication slendr paper, new versions SLiM slendr released. Specifically, results paper based SLiM version 4.0.1 slendr version 0.5. Whenever new version SLiM released, ’s always chance aspect randomness simulation changes slightly, affects results even though random seed kept value. , figures paper (specifically figure Example 4 paper) look slightly different. However, results still relevant applicable, just presentation changed.","code":"# workaround for GitHub actions getting \"Killed\" due to out of memory issues # -- when running unit tests on GitHub, smaller amount of sequence will be # simulated if (Sys.getenv(\"RUNNER_OS\") != \"\") {   scaling <- 10 # use 10X less sequence on GitHub actions } else {   scaling <- 1 # but simulate normal amount of data otherwise }  library(slendr)  library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union library(ggplot2) library(purrr) library(tidyr) library(cowplot) library(forcats)  init_env() #> The interface to all required Python modules has been activated.  SEED <- 42 set.seed(SEED) # placeholder \"figure\" where a code chunk will be pasted p_code <- ggplot() +   geom_text(aes(x = 1, y = 1), label = \"code will be here\") +   theme_void()"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"script-from-panel-a","dir":"Articles","previous_headings":"Example 1 (Figure 2)","what":"Script from panel A","title":"Examples from the slendr paper","text":"","code":"o <- population(\"o\", time = 1, N = 100) c <- population(\"c\", time = 2500, N = 100, parent = o) a <- population(\"a\", time = 2800, N = 100, parent = c) b <- population(\"b\", time = 3700, N = 100, parent = a) x1 <- population(\"x1\", time = 4000, N = 15000, parent = c) x2 <- population(\"x2\", time = 4300, N = 15000, parent = x1)  gf <- gene_flow(from = b, to = x1, start = 5400, end = 5800, 0.1)  model <- compile_model(   populations = list(o, a, b, c, x1, x2), gene_flow = gf,   generation_time = 1, simulation_length = 6000 )  plot_model(model, sizes = FALSE, proportions = TRUE) # panel B  ts <- msprime(model, sequence_length = 100e6 / scaling, recombination_rate = 1e-8, random_seed = SEED) %>%   ts_mutate(mutation_rate = 1e-8, random_seed = SEED)  samples <- ts_samples(ts) %>% group_by(pop) %>% sample_n(100)  # panel C divergence <- ts_divergence(ts, split(samples$name, samples$pop))  # panel D f4ratio <- ts_f4ratio(   ts, X = filter(samples, pop %in% c(\"x1\", \"x2\"))$name,   A = \"a_1\", B = \"b_1\", C = \"c_1\", O = \"o_1\" )"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"plotting-code","dir":"Articles","previous_headings":"Example 1 (Figure 2)","what":"Plotting code","title":"Examples from the slendr paper","text":"","code":"divergence <- divergence %>% mutate(pair = paste(x, \"-\", y))  f4ratio <- f4ratio %>% mutate(population = gsub(\"^(.*)_.*$\", \"\\\\1\", X), alpha = alpha * 100)  p_ex1_divergence <- divergence %>%   ggplot(aes(fct_reorder(pair, divergence), divergence)) +   geom_point(size = 2.5) +   xlab(\"population pair\") + ylab(\"pairwise divergence\") +   theme_minimal() +   theme(legend.position = \"bottom\",         legend.text = element_text(size = 10),         axis.text.x = element_text(hjust = 1, angle = 45, size = 8),         axis.title.x = element_blank())  p_ex1_f4ratio <- f4ratio %>%   ggplot(aes(population, alpha)) +   geom_hline(yintercept = 0, linetype = 2) +   geom_jitter(alpha = 0.5) +   geom_boxplot(outlier.shape = NA, alpha = 0.7) +   ylab(base::expression(italic(\"f\")[4]~\"-ratio ancestry proportion [%]\")) +   theme_minimal() +   coord_cartesian(ylim = c(0, 20)) +   theme(legend.position = \"none\",         axis.text.x = element_text(size = 11),         axis.title.x = element_blank(),         panel.grid.major.x = element_blank())#; p_ex3_f4ratio  # let's avoid ggpubr as another dependency: # https://github.com/kassambara/ggpubr/blob/master/R/as_ggplot.R#L27 p_ex1_legend <- ggdraw() + draw_grob(grid::grobTree(get_legend(p_ex1_divergence)))  p_ex1_model <- plot_model(model, sizes = FALSE, proportions = TRUE)  p_ex1 <- plot_grid(   p_code,   plot_grid(     p_ex1_model,     plot_grid(       p_ex1_divergence + theme(legend.position = \"none\"),       p_ex1_f4ratio,       ncol = 2, rel_widths = c(1, 0.8), labels = c(\"C\", \"D\")     ),     p_ex1_legend, nrow = 3, rel_heights = c(1, 1, 0.1),     labels = \"B\"   ),   nrow = 1, labels = c(\"A\", \"\") )  p_ex1"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"script-from-panel-a-1","dir":"Articles","previous_headings":"Example 2 (Figure 3)","what":"Script from panel A","title":"Examples from the slendr paper","text":"","code":"map <- world(xrange = c(0, 10), yrange = c(0, 10),              landscape = region(center = c(5, 5), radius = 5))  p1 <- population(\"pop1\", time = 1, N = 2000, map = map, competition = 0) p2 <- population(\"pop2\", time = 1, N = 2000, map = map, competition = 9) p3 <- population(\"pop3\", time = 1, N = 2000, map = map, competition = 6) p4 <- population(\"pop4\", time = 1, N = 2000, map = map, competition = 5) p5 <- population(\"pop5\", time = 1, N = 2000, map = map, competition = 4) p6 <- population(\"pop6\", time = 1, N = 2000, map = map, competition = 3) p7 <- population(\"pop7\", time = 1, N = 2000, map = map, competition = 2) p8 <- population(\"pop8\", time = 1, N = 2000, map = map, competition = 1)  model <- compile_model(   populations = list(p1, p2, p3, p4, p5, p6, p7, p8),   generation_time = 1, simulation_length = 5000, resolution = 0.1,   mating = 0.1, dispersal = 0.05 )  ts <-   slim(model, sequence_length = 10e6 / scaling, recombination_rate = 1e-8, random_seed = SEED) %>%   ts_simplify() %>%   ts_mutate(mutation_rate = 1e-7, random_seed = SEED) #> Warning: Simplifying a non-recapitated tree sequence. Make sure this is what #> you really want  locations <- ts_nodes(ts) %>% filter(time == max(time))  heterozygosity <- ts_samples(ts) %>%   group_by(pop) %>%   sample_n(100) %>%   mutate(pi = ts_diversity(ts, name)$diversity)"},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"plotting-code-1","dir":"Articles","previous_headings":"Example 2 (Figure 3)","what":"Plotting code","title":"Examples from the slendr paper","text":"","code":"p_ex2_clustering <- ggplot() +   geom_sf(data = map) +   geom_sf(data = locations, aes(color = pop), size = 0.05, alpha = 0.25) +   facet_grid(. ~ pop, switch = \"x\") +   xlab(\"spatial distributions emerged in the simulation\") +   theme(     strip.background = element_blank(),     strip.text = element_text(size = 11),     panel.grid = element_blank(),     axis.ticks = element_blank(),     axis.text = element_blank(),     panel.background = element_blank()   ) +   guides(color = \"none\")  p_ex2_diversity <- ggplot(heterozygosity, aes(pop, pi, color = pop)) +   geom_violin(color = \"black\") +   geom_jitter(alpha = 0.5) +   labs(y = \"individual heterozygosity\") +   guides(color = \"none\") +   theme_minimal() +   theme(axis.title.x = element_blank(),         axis.text.x = element_blank(), panel.grid.major.x = element_blank(),         plot.margin = margin(t = 0.2, r = 0.2, b = -0.1, l = 0.2, \"cm\"))  p_ex2 <- plot_grid(   p_code,   plot_grid(     p_ex2_diversity,     p_ex2_clustering +       theme(plot.margin = margin(t = 0, r = 0.4, b = 0, l = 1.8, \"cm\")),     nrow = 2,     rel_heights = c(1, 0.5),     labels = c(\"B\", \"C\")   ),   nrow = 2, labels = c(\"A\", \"\"), rel_heights = c(1.5, 1) )  p_ex2"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"script-from-panel-a-2","dir":"Articles","previous_headings":"Example 3 (Figure 4)","what":"Script from panel A","title":"Examples from the slendr paper","text":"","code":"map <- world(xrange = c(-13, 70), yrange = c(18, 65), crs = 3035) #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpdblWjc/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84  R1 <- region(   \"EHG range\", map,   polygon = list(c(26, 55), c(38, 53), c(48, 53), c(60, 53),                  c(60, 60), c(48, 63), c(38, 63), c(26, 60)) ) R2 <- region(   \"Europe\", map,   polygon = list(     c(-8, 35), c(-5, 36), c(10, 38), c(20, 35), c(25, 35),     c(33, 45), c(20, 58), c(-5, 60), c(-15, 50)   ) ) R3 <- region(   \"Anatolia\", map,   polygon = list(c(28, 35), c(40, 35), c(42, 40),                  c(30, 43), c(27, 40), c(25, 38)) ) R4 <- join(R2, R3) R5 <- region(   \"YAM range\", map,   polygon = list(c(26, 50), c(38, 49), c(48, 50),                  c(48, 56), c(38, 59), c(26, 56)) )  ooa_trajectory <- list(c(40, 30), c(50, 30), c(60, 40), c(45, 55)) map <- world(xrange = c(-13, 70), yrange = c(18, 65), crs = 3035) #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpdblWjc/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84  ooa <- population(   \"OOA\", time = 50000, N = 500, remove = 23000,    map = map, center = c(33, 30), radius = 400e3 ) %>%   move(trajectory = ooa_trajectory, start = 50000, end = 40000, snapshots = 30)  ehg <- population(   \"EHG\", time = 28000, N = 1000, parent = ooa, remove = 6000,   map = map, polygon = R1 )  eur <- population(   \"EUR\", time = 30000, N = 2000, parent = ooa,   map = map, polygon = R2 ) %>%   resize(N = 10000, time = 5000, end = 0, how = \"exponential\")  ana <- population(   \"ANA\", time = 25000, N = 4000, parent = ooa, remove = 3000,   map = map, polygon = R3 ) %>%   expand_range(by = 3e6, start = 10000, end = 7000, polygon = R4, snapshots = 15)  yam <- population(   \"YAM\", time = 7000, N = 600, parent = ehg, remove = 2500,   map = map, polygon = R5 ) %>%   move(trajectory = list(c(15, 50)), start = 5000, end = 3000, snapshots = 10)  gf <- list(   gene_flow(ana, to = yam, rate = 0.5, start = 6500, end = 5000),   gene_flow(ana, to = eur, rate = 0.6, start = 8000, end = 6000),   gene_flow(yam, to = eur, rate = 0.7, start = 3500, end = 3000) )  model <- compile_model(   populations = list(ooa, ehg, eur, ana, yam), gene_flow = gf,   generation_time = 30, resolution = 10e3,   competition = 150e3, mating = 120e3, dispersal = 90e3 )  samples <- schedule_sampling(   model, times = seq(0, 50000, by = 1000),   list(ehg, 20), list(ana, 20), list(yam, 20), list(eur, 20) )  plot_model(model, sizes = FALSE) plot_map(model)  ts <- slim(   model, burnin = 200000, samples = samples, random_seed = SEED,   sequence_length = 200000, recombination_rate = 1e-8 )"},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"plotting-code-2","dir":"Articles","previous_headings":"Example 3 (Figure 4)","what":"Plotting code","title":"Examples from the slendr paper","text":"","code":"p_map <- plot_map(model) +   theme(legend.position = \"bottom\") +   guides(alpha = \"none\")  p_ex3 <- plot_grid(   p_code,   plot_grid(     plot_model(model, sizes = FALSE),     p_map,     labels = c(\"B\", \"C\"), nrow = 2, rel_heights = c(1, 1)   ),   ncol = 2, labels = c(\"A\", \"\"), rel_widths = c(1, 1.2) )  p_ex3"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"script-from-panel-a-3","dir":"Articles","previous_headings":"Example 4 (Figure 5)","what":"Script from panel A","title":"Examples from the slendr paper","text":"","code":"ts_small <- ts_simplify(ts, simplify_to = c(\"EUR_578\", \"YAM_75\", \"ANA_163\", \"EHG_208\"))  # tskit uses zero-based indexing tree <- ts_phylo(ts_small, i = 15 / scaling) #> Starting checking the validity of tree... #> Found number of tips: n = 8  #> Found number of nodes: m = 7  #> Done. nodes <- ts_nodes(tree) edges <- ts_edges(tree)  ancestors <- ts_ancestors(ts, \"EUR_578\")"},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"plotting-code-3","dir":"Articles","previous_headings":"Example 4 (Figure 5)","what":"Plotting code","title":"Examples from the slendr paper","text":"explained top page, random number generation current SLiM slightly different version 4.0.1 available time writing paper. figure looks little different well (different random generation implies different movement individuals, different individuals sampled). record, ’s original figure used paper also originally part vignette:","code":"library(ggtree) #> ggtree v3.10.0 For help: https://yulab-smu.top/treedata-book/ #>  #> If you use the ggtree package suite in published research, please cite #> the appropriate paper(s): #>  #> Guangchuang Yu, David Smith, Huachen Zhu, Yi Guan, Tommy Tsan-Yuk Lam. #> ggtree: an R package for visualization and annotation of phylogenetic #> trees with their covariates and other associated data. Methods in #> Ecology and Evolution. 2017, 8(1):28-36. doi:10.1111/2041-210X.12628 #>  #> Guangchuang Yu.  Data Integration, Manipulation and Visualization of #> Phylogenetic Trees (1st edition). Chapman and Hall/CRC. 2022, #> doi:10.1201/9781003279242 #>  #> Guangchuang Yu. Using ggtree to visualize data on tree-like structures. #> Current Protocols in Bioinformatics. 2020, 69:e96. doi:10.1002/cpbi.96 #>  #>  #> Attaching package: 'ggtree' #> The following object is masked from 'package:tidyr': #>  #>     expand  # prepare annotation table for ggtree linking R phylo node ID (not tskit integer # ID!) of each node with its population name df <- as_tibble(nodes) %>% select(node = phylo_id, pop)  abs_comma <- function (x, ...) {   format(abs(x) / 1000, ..., scientific = FALSE, trim = TRUE) }  highlight_nodes <- as_tibble(nodes) %>% dplyr::filter(name == \"EUR_578\") %>% .$phylo_id  p_tree <- ggtree(tree, aes(color = pop, fill = pop)) %<+% df +   geom_tiplab(align = TRUE, geom = \"label\", offset = 2000,               color = \"white\", fontface = \"bold\", size = 2.7) +   geom_tiplab(align = TRUE, geom = NULL, linetype = \"dotted\", size = 0) +   geom_point2(aes(subset = (node %in% highlight_nodes)), color = \"black\", size = 2.7) +   geom_label2(aes(label = label, subset = !isTip),               color = \"black\", size = 2.7) +   theme_tree2() +   theme(legend.position = \"none\") +   xlab(\"time before present [thousand years ago]\") +   scale_x_continuous(limits = c(-80000, 31000), labels = abs_comma,                      breaks = -c(100, 80, 60, 40, 20, 0) * 1000) p_tree <- revts(p_tree)  # nodes$label <- ifelse(is.na(nodes$name), nodes$node_id, nodes$name) nodes$label <- sapply(1:nrow(nodes), function(i) {   if (is.na(nodes[i, ]$name))     nodes[i, ]$node_id   else {     ind <- nodes[i, ]$name     paste(nodes[!is.na(nodes$name) & nodes$name == ind, ]$node_id, collapse = \"&\")   } })  p_map <- ggplot() +   geom_sf(data = map) +   geom_sf(data = edges, aes(color = parent_pop), size = 0.5) +   geom_sf_label(data = nodes[!nodes$sampled, ],                 aes(label = node_id, fill = pop), size = 3) +   geom_sf_label(data = nodes[nodes$sampled, ],                 aes(label = label, fill = pop), size = 3,                 fontface = \"bold\", color = \"white\") +   coord_sf(xlim = c(3177066.1, 7188656.9),            ylim = c(757021.7, 5202983.3), expand = 0) +   guides(fill = guide_legend(\"\", override.aes = aes(label = \"\"))) +   guides(color = \"none\") +   scale_colour_discrete(drop = FALSE) +   scale_fill_discrete(drop = FALSE) +   theme_bw() +   theme(legend.position = \"bottom\",         axis.title.x = element_blank(),         axis.title.y = element_blank())  chrom_names <- stats::setNames(   c(\"EUR_578 (node 6)\", \"EUR_578 (node 7)\"),   unique(ancestors$node_id) )  p_ancestors <- ggplot() +   geom_sf(data = map) +   geom_sf(data = ancestors, size = 0.5, alpha = 0.25) +   geom_sf(data = sf::st_set_geometry(ancestors, \"parent_location\"),           aes(shape = parent_pop, color = parent_pop)) +   geom_sf(data = filter(ts_nodes(ts), name == \"EUR_578\"), size = 3) +   coord_sf(expand = 0) +   labs(x = \"longitude\", y = \"latitude\") +   theme_bw() +   facet_grid(. ~ node_id, labeller = labeller(node_id = chrom_names)) +   theme(legend.position = \"none\")  p_legend <- ggdraw() + draw_grob(grid::grobTree(get_legend(p_map)))  p_ex4 <- plot_grid(   p_code,   plot_grid(p_tree + theme(legend.position = \"none\"),             p_map + theme(legend.position = \"none\"),             labels = c(\"B\", \"C\"), rel_widths = c(1, 0.9)),   p_ancestors,   p_legend,   labels = c(\"A\", \"\", \"D\", \"\"),   nrow = 4, rel_heights = c(0.5, 1, 1, 0.1) )  p_ex4"},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"run-time-of-each-code-example-from-the-paper","dir":"Articles","previous_headings":"","what":"Run time of each code example from the paper","title":"Examples from the slendr paper","text":"following times measured 16’’ MacBook Pro Apple M1 Pro chip (2021 model), 32 GB RAM, running macOS Ventura Version 13.1. table , ex1, ex2, ex3, ex4 correspond runtimes code shown one four example figures slendr paper.","code":""},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"demographic-model","dir":"Articles","previous_headings":"","what":"Demographic model","title":"Extracting true ancestry tracts","text":"Let’s imagine following demographic model Neanderthal introgression ancestors non-Africans (represented “EUR” “PAP” populations, approximating European Papuan people living today), followed Denisovan introgression ancestors Papuans:","code":"anc_all <- population(\"ancestor_all\", time = 700e3, N = 10000, remove = 640e3) afr <- population(\"AFR\", parent = anc_all, time = 650e3, N = 10000) anc_arch <- population(\"ancestor_archaics\", parent = anc_all, time = 650e3, N = 10000, remove = 390e3) nea <- population(\"NEA\", parent = anc_arch, time = 400e3, N = 2000, remove = 30e3) den <- population(\"DEN\", parent = anc_arch, time = 400e3, N = 2000, remove = 30e3) nonafr <- population(\"nonAFR\", parent = afr, time = 100e3, N = 3000, remove = 39e3) eur <- population(\"EUR\", parent = nonafr, time = 45e3, N = 5000) pap <- population(\"PAP\", parent = nonafr, time = 45e3, N = 5000)  gf <- list(   gene_flow(from = nea, to = nonafr, rate = 0.03, start = 55000, end = 50000),   gene_flow(from = den, to = pap, rate = 0.07, start = 35000, end = 30000) )  model <- compile_model(   populations = list(anc_all, afr, anc_arch, nea, den, nonafr, eur, pap),   gene_flow = gf,   generation_time = 30,   serialize = FALSE )  plot_model(   model, sizes = FALSE,   order = c(\"AFR\", \"EUR\", \"nonAFR\", \"PAP\", \"ancestor_all\", \"DEN\", \"ancestor_archaics\", \"NEA\") )"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"tree-sequence-simulation","dir":"Articles","previous_headings":"","what":"Tree sequence simulation","title":"Extracting true ancestry tracts","text":"Let’s now simulate 50Mb tree sequence model, recording 50 diploid individuals EUR PAP populations:","code":"samples <- schedule_sampling(model, times = 0, list(eur, 50), list(pap, 50))  ts <- msprime(model, sequence_length = 100e6, recombination_rate = 1e-8, samples = samples, random_seed = 42)"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"extracting-ancestry-tracts","dir":"Articles","previous_headings":"","what":"Extracting ancestry tracts","title":"Extracting true ancestry tracts","text":"order extract tracts Neanderthal Denisovan ancestry, can use slendr’s new function ts_tracts() serves simplified R-friendly interface Python method tspop.get_pop_ancestry(). important piece information used function -called “census time”, records time recording “ancestral population” identity node ancestral subsegment sample set. Please see excellent vignette tspop information inner workings algorithm. case, let’s extract ancestry tracts corresponding ancestral nodes present 55 thousand years ago – time corresponds moment start archaic introgression: table ancestry tracts looks like. expect, see column indicating name individual, left right coordinates tract individual, well population name source ancestry tract:","code":"nea_tracts <- ts_tracts(ts, census = 55000) #>  #> PopAncestry summary #> Number of ancestral populations:     3 #> Number of sample chromosomes:        200 #> Number of ancestors:             44052 #> Total length of genomes:         20000000000.000000 #> Ancestral coverage:          20000000000.000000 den_tracts <- ts_tracts(ts, census = 35000) #>  #> PopAncestry summary #> Number of ancestral populations:     3 #> Number of sample chromosomes:        200 #> Number of ancestors:             55373 #> Total length of genomes:         20000000000.000000 #> Ancestral coverage:          20000000000.000000  tracts <- bind_rows(nea_tracts, den_tracts) tracts #> # A tibble: 17,002 × 8 #>    name  node_id pop   source_pop     left    right length source_pop_id #>    <chr>   <dbl> <fct> <fct>         <dbl>    <dbl>  <dbl>         <dbl> #>  1 EUR_1       0 EUR   NEA          781887   852668  70781             3 #>  2 EUR_1       0 EUR   NEA         1451389  1463837  12448             3 #>  3 EUR_1       0 EUR   NEA         1596995  1601441   4446             3 #>  4 EUR_1       0 EUR   NEA         1629857  1689656  59799             3 #>  5 EUR_1       0 EUR   NEA         3203711  3339565 135854             3 #>  6 EUR_1       0 EUR   NEA         3850629  3947923  97294             3 #>  7 EUR_1       0 EUR   NEA         6190700  6205627  14927             3 #>  8 EUR_1       0 EUR   NEA        10394211 10426667  32456             3 #>  9 EUR_1       0 EUR   NEA        13044685 13180258 135573             3 #> 10 EUR_1       0 EUR   NEA        14858956 14912068  53112             3 #> # ℹ 16,992 more rows"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"summaries-of-ancestral-proportions","dir":"Articles","previous_headings":"","what":"Summaries of ancestral proportions","title":"Extracting true ancestry tracts","text":"summarise ancestry proportions target EUR PAP populations, see EUR population carries ~3% Neanderthal ancestry also true PAP population. However, also see Papuans carry 7% Denisovan ancestry. consistent model, also expectation empirical data. Let’s visualize proportions individual level:","code":"summary <- tracts %>%   group_by(name, node_id, pop, source_pop) %>%   summarise(prop = sum(length) / 100e6) #> `summarise()` has grouped output by 'name', 'node_id', 'pop'. You can override #> using the `.groups` argument.  summary %>% group_by(pop, source_pop) %>% summarise(mean(prop)) %>% arrange(source_pop, pop) #> `summarise()` has grouped output by 'pop'. You can override using the `.groups` #> argument. #> # A tibble: 3 × 3 #> # Groups:   pop [2] #>   pop   source_pop `mean(prop)` #>   <fct> <fct>             <dbl> #> 1 EUR   NEA              0.0303 #> 2 PAP   NEA              0.0348 #> 3 PAP   DEN              0.0739 summary %>% ggplot(aes(source_pop, prop, color = source_pop, fill = source_pop)) +   geom_jitter() +   coord_cartesian(ylim = c(0, 0.2)) +   geom_hline(yintercept = c(0.03, 0.08), linetype = 2) +   ylab(\"ancestry proportion\") +   facet_wrap(~ pop) +   ggtitle(\"Ancestry proportions in each individual\",           \"(vertical lines represent 3% and 7% baseline expectations\")"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"chromosome-painting-of-ancestry-tracts","dir":"Articles","previous_headings":"","what":"“Chromosome painting” of ancestry tracts","title":"Extracting true ancestry tracts","text":"tracts object contains coordinates every single ancestry segment simulated individuals, can “paint” chromosome two archaic human ancestries:  lining NEA & DEN ancestry tracts EUR PAP populations, can see common origin Neanderthal ancestry non-African populations manifests significant overlap NEA tracts populations.","code":"tracts %>% mutate(chrom = paste(name, \" (node\", node_id, \")\")) %>% ggplot(aes(x = left, xend = right, y = chrom, yend = chrom, color = source_pop)) +   geom_segment(linewidth = 3) +   theme_minimal() +   labs(x = \"position [bp]\", y = \"haplotype\") +   ggtitle(\"True ancestry tracts along each chromosome\") +   theme(axis.text.y = element_blank(), panel.grid = element_blank()) +   facet_grid(pop ~ ., scales = \"free_y\")"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"average-tract-lengths","dir":"Articles","previous_headings":"","what":"Average tract lengths:","title":"Extracting true ancestry tracts","text":"Let’s compute simple summaries tract lengths simulated data, compare theoretical expectations. Theoretical expectations (Racimo Slatkin 2015, Box 1) Neanderthal tracts: Denisovan tracts: can see, simulations far theoretical expectations, giving us confidence simulations (ancestry tract extraction algorithm) working expected.","code":"tracts %>%   group_by(pop, source_pop) %>%   summarise(mean(length)) #> `summarise()` has grouped output by 'pop'. You can override using the `.groups` #> argument. #> # A tibble: 3 × 3 #> # Groups:   pop [2] #>   pop   source_pop `mean(length)` #>   <fct> <fct>               <dbl> #> 1 EUR   NEA                65765. #> 2 PAP   NEA                69204. #> 3 PAP   DEN               100317. m <- 0.03 t <- 52500 / 30 r <- 1e-8  mean_nea <- 1 / ((1 - m) * r * (t - 1)) mean_nea #> [1] 58943.84 m <- 0.07 t <- 37500 / 30 r <- 1e-8  mean_den <- 1 / ((1 - m) * r * (t - 1)) mean_den #> [1] 86090.38"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"distribution-of-ancestry-tract-lengths","dir":"Articles","previous_headings":"","what":"Distribution of ancestry tract lengths","title":"Extracting true ancestry tracts","text":"Finally, let’s plot distributions lengths ancestry tracts. case archaic human introgression well studied ’s perhaps exciting look figures. said, less well studied species, might interesting use kinds simulations inference introgression times proportions via Approximate Bayesian Computation another method:","code":"expectation_df <- data.frame(   pop = c(\"EUR\", \"PAP\", \"PAP\"),   source_pop = c(\"NEA\", \"NEA\", \"DEN\"),   length = c(mean_nea, mean_nea, mean_den) ) p_densities <- tracts %>% ggplot(aes(length, color = source_pop)) +   geom_density() +   geom_vline(data = expectation_df, aes(xintercept = length, color = source_pop),              linetype = 2) +   facet_wrap(~ pop) +   ggtitle(\"Distribution of tract lengths per different ancestries\")  cowplot::plot_grid(p_densities, p_densities + scale_x_log10(), nrow = 2)"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"pure-msprime-tree-sequence","dir":"Articles","previous_headings":"","what":"Pure msprime tree sequence","title":"Extracting true ancestry tracts","text":"Finally, sanity check, let’s use pure msprime simulation example official tspop documentation test ts_tracts() behaves expected even standard msprime tree-sequence object. First, let’s run simulation code exactly : Let’s save msprime tree sequence disk can load R (.e., approximating might want want use ts_tracts() without running slendr simulation first): Now let’s move R , load tree sequence slendr extract ancestry tracts using ts_tracts(): setting squashed = FALSE, get full, un-squashed ancestry segments, appropriate ancestral node ID: comparing two tables pandas data frames tspop documentation, can see obtained results.","code":"import msprime  pop_size = 500 sequence_length = 1e7 seed = 98765 rho = 3e-8  # Make the Demography object. demography = msprime.Demography() demography.add_population(name=\"RED\", initial_size=pop_size) #> Population(initial_size=500, growth_rate=0, name='RED', description='', extra_metadata={}, default_sampling_time=None, initially_active=None, id=0) demography.add_population(name=\"BLUE\", initial_size=pop_size) #> Population(initial_size=500, growth_rate=0, name='BLUE', description='', extra_metadata={}, default_sampling_time=None, initially_active=None, id=1) demography.add_population(name=\"ADMIX\", initial_size=pop_size) #> Population(initial_size=500, growth_rate=0, name='ADMIX', description='', extra_metadata={}, default_sampling_time=None, initially_active=None, id=2) demography.add_population(name=\"ANC\", initial_size=pop_size) #> Population(initial_size=500, growth_rate=0, name='ANC', description='', extra_metadata={}, default_sampling_time=None, initially_active=None, id=3) demography.add_admixture(     time=100, derived=\"ADMIX\", ancestral=[\"RED\", \"BLUE\"], proportions=[0.5, 0.5] ) #> Admixture(time=100, derived='ADMIX', ancestral=['RED', 'BLUE'], proportions=[0.5, 0.5]) demography.add_census(time=100.01) # Census is here! #> CensusEvent(time=100.01) demography.add_population_split(     time=1000, derived=[\"RED\", \"BLUE\"], ancestral=\"ANC\" ) #> PopulationSplit(time=1000, derived=['RED', 'BLUE'], ancestral='ANC')  # Simulate. ts = msprime.sim_ancestry(     samples={\"RED\": 0, \"BLUE\": 0, \"ADMIX\" : 2},     demography=demography,     random_seed=seed,     sequence_length=sequence_length,     recombination_rate=rho ) import tempfile path = tempfile.NamedTemporaryFile(suffix=\".trees\").name  ts.dump(path) sim_ts <- ts_load(reticulate::py$path)  squashed_tracts <- ts_tracts(sim_ts, census = 100.01, squashed = TRUE) #>  #> PopAncestry summary #> Number of ancestral populations:     2 #> Number of sample chromosomes:        4 #> Number of ancestors:             118 #> Total length of genomes:         40000000.000000 #> Ancestral coverage:          40000000.000000  head(squashed_tracts) #> # A tibble: 6 × 4 #>   sample    left   right population #>    <dbl>   <dbl>   <dbl>      <dbl> #> 1      0       0  419848          0 #> 2      0  419848  483009          1 #> 3      0  483009 1475765          0 #> 4      0 1475765 2427904          1 #> 5      0 2427904 3635390          0 #> 6      0 3635390 4606954          1 tail(squashed_tracts) #> # A tibble: 6 × 4 #>   sample    left    right population #>    <dbl>   <dbl>    <dbl>      <dbl> #> 1      3 7134130  7362300          1 #> 2      3 7362300  7369409          0 #> 3      3 7369409  7596783          1 #> 4      3 7596783  8289015          0 #> 5      3 8289015  8918727          1 #> 6      3 8918727 10000000          0 full_tracts <- ts_tracts(sim_ts, census = 100.01, squashed = FALSE) #>  #> PopAncestry summary #> Number of ancestral populations:     2 #> Number of sample chromosomes:        4 #> Number of ancestors:             118 #> Total length of genomes:         40000000.000000 #> Ancestral coverage:          40000000.000000  head(full_tracts) #> # A tibble: 6 × 5 #>   sample   left   right ancestor population #>    <int>  <dbl>   <dbl>    <int>      <int> #> 1      0      0   33027       74          0 #> 2      0  33027  155453       33          0 #> 3      0 155453  290542       46          0 #> 4      0 290542  419848       18          0 #> 5      0 419848  483009       83          1 #> 6      0 483009 1475765       28          0 tail(full_tracts) #> # A tibble: 6 × 5 #>   sample    left    right ancestor population #>    <int>   <dbl>    <dbl>    <int>      <int> #> 1      3 8477625  8672850       94          1 #> 2      3 8672850  8849756       95          1 #> 3      3 8849756  8918727      131          1 #> 4      3 8918727  9165035       44          0 #> 5      3 9165035  9176562       47          0 #> 6      3 9176562 10000000       58          0"},{"path":"https://www.slendr.net/authors.html","id":null,"dir":"","previous_headings":"","what":"Authors","title":"Authors and Citation","text":"Martin Petr. Author, maintainer.","code":""},{"path":"https://www.slendr.net/authors.html","id":"citation","dir":"","previous_headings":"","what":"Citation","title":"Authors and Citation","text":"Petr M (2024). slendr: Simulation Framework Spatiotemporal Population Genetics. R package version 0.9.0, https://github.com/bodkan/slendr.","code":"@Manual{,   title = {slendr: A Simulation Framework for Spatiotemporal Population Genetics},   author = {Martin Petr},   year = {2024},   note = {R package version 0.9.0},   url = {https://github.com/bodkan/slendr}, }"},{"path":[]},{"path":"https://www.slendr.net/index.html","id":"overview-","dir":"","previous_headings":"","what":"Overview","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"slendr R package toolbox defining population genetic models simulating genomic data entirely R. originally conceived framework simulating spatially-explicit genomic data real geographic landscapes grown much . page briefly summarizes slendr’s important features. much detailed description slendr architecture extensive set practical code examples can found paper PCI journal website.","code":""},{"path":"https://www.slendr.net/index.html","id":"citing-slendr","dir":"","previous_headings":"","what":"Citing slendr","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"slendr paper now published Peer Community Journal! use slendr work, please cite : Petr, Martin; Haller, Benjamin C.; Ralph, Peter L.; Racimo, Fernando. slendr: framework spatio-temporal population genomic simulations geographic landscapes. Peer Community Journal, Volume 3 (2023), article . e121. doi : 10.24072/pcjournal.354. Citations help justify development fixing bugs! Thank ! ❤️","code":""},{"path":"https://www.slendr.net/index.html","id":"main-features","dir":"","previous_headings":"","what":"Main features","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"brief summary slendr’s important features. R package allows : Program demographic models, including population splits, population size changes, gene-flow events using extremely simple declarative language entirely R (see vignette example model-definition interface). Even complex models can written little code require bare minimum R programming knowledge (thing user needs know call R function R data frame look like). Execute slendr models using efficient, tailor-made SLiM msprime simulation scripts bundled R package. simulation engines save outputs form efficient tree-sequence data structure. SLiM msprime programming needed! Load, process, analyse tree-sequence outputs via slendr’s built-R interface tree-sequence library tskit. can compute many population genetic statistics R immediately simulation finishes directly output tree sequences, without convert files formats (VCF, EIGENSTRAT) analysis different software. Encode complex models population movements landscape (see brief example model , extended explanation tutorial). knowledge cartographic geospatial analysis concepts needed. Simulate dynamic spatial demographic models using SLiM’s continuous-space simulation capabilities directly R (, SLiM programming required). outputs simulations saved tree sequences can analysed using standard R geospatial data analysis libraries. slendr performs conversion tree sequence tables appropriate spatial R data type automatically. Specify within-population individual dispersal dynamics R interface leveraging SLiM’s individual interaction parameters implemented SLiM back-end script. Schedule sampling events specify many individuals’ genomes, populations, times (optionally, locations) recorded simulation engine (SLiM msprime) output tree-sequence files. Utilizing flexibility R wealth libraries statistics, geospatial analysis graphics, combining power population genetic simulation frameworks SLiM msprime, slendr R package makes possible write entire simulation analytic pipelines without need leave R environment.","code":""},{"path":"https://www.slendr.net/index.html","id":"testing-the-r-package-in-an-online-rstudio-session","dir":"","previous_headings":"","what":"Testing the R package in an online RStudio session","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"can open RStudio session test examples vignettes directly web browser clicking button (installation needed!):  case RStudio instance appears starting slowly, please patient (Binder freely available service limited computational resources provided community). Binder crashes, try reloading web page, restart cloud session. get browser-based RStudio session, can navigate vignettes/ directory test examples !","code":""},{"path":"https://www.slendr.net/index.html","id":"installation","dir":"","previous_headings":"","what":"Installation","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"slendr now available CRAN means can install simply entering install.packages(\"slendr\") R console. like test latest features software (perhaps need bug fixes), can install devtools::install_github(\"bodkan/slendr\") (note requires R package devtools). ⚠️⚠️⚠️ Note: default slendr installation longer hard dependency geospatial R packages sf, stars, rnaturalearth. indend use slendr spatial genomic simulations data analysis, run install.packages(c(\"sf\", \"stars\", \"rnaturalearth\")) first. made change ease installation burden researchers might need slendr’s support traditional, non-spatial simulations (represent majority use-cases population genetics general). geospatial R packages imply non-negligible amount software dependencies , proven unnecessary hurdle many users might need (least first). ⚠️⚠️⚠️ software active development! like stay updated: Click “Watch” button project’s GitHub website. Follow social media posting progress updates (can find links homepage). time time, take look changelog post updates new features, breaking changes, etc.","code":""},{"path":"https://www.slendr.net/index.html","id":"example","dir":"","previous_headings":"","what":"Example","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"brief demonstration kind spatial model slendr originally designed simulate data . Please note although spatially-explicit population genetic model, slendr extensive support traditional, non-spatial simulations well. Furthermore, example shows specify simulate model R. doesn’t show analyse tree-sequence outputs compute population genetic statistics (important feature demonstrated tutorial).","code":""},{"path":"https://www.slendr.net/index.html","id":"id_1-setup-the-spatial-context","dir":"","previous_headings":"Example","what":"1. Setup the spatial context","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"Imagine wanted simulate spatio-temporal genomic data toy model history modern humans West Eurasia Africa migration. First, define spatial context simulation. represent “world” occupied populations model. can visualize defined world map using function plot_map provided package.  Although example use real Earth landscape, map can completely abstract (either blank user-defined landscape features continents, islands, corridors barriers).","code":"library(slendr)  # this sets up internal Python environment and needs to be ran only once! # (do not put this in your R scripts, run this command in the R console # after you (re-)install slendr) setup_env()  # activate the internal Python environment needed for simulation and # tree-sequence processing init_env()  map <- world(   xrange = c(-13, 70), # min-max longitude   yrange = c(18, 65),  # min-max latitude   crs = \"EPSG:3035\"    # coordinate reference system (CRS) for West Eurasia ) plot_map(map)"},{"path":"https://www.slendr.net/index.html","id":"id_2-define-broader-geographic-regions","dir":"","previous_headings":"Example","what":"2. Define broader geographic regions","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"order make definitions population ranges () easier, can define smaller regions map using function region. Note coordinates slendr specified geographic coordinate system (.e., degrees longitude latitude), internally represented projected CRS (case, EPSG 3035 specified ). makes easier us define spatial features simply reading coordinates regular map internal projected CRS makes simulations accurate (distances shapes distorted can use CRS tailored region world working ). projected CRS takes care projection part world ’re interested three-dimensional Earth surface two-dimensional map. , can use generic plot_map function visualize objects, making sure specified correctly:","code":"africa <- region(   \"Africa\", map,   polygon = list(c(-18, 20), c(38, 20), c(30, 33),                  c(20, 33), c(10, 38), c(-6, 35)) ) europe <- region(   \"Europe\", map,   polygon = list(     c(-8, 35), c(-5, 36), c(10, 38), c(20, 35), c(25, 35),     c(33, 45), c(20, 58), c(-5, 60), c(-15, 50)   ) ) anatolia <- region(   \"Anatolia\", map,   polygon = list(c(28, 35), c(40, 35), c(42, 40),                  c(30, 43), c(27, 40), c(25, 38)) ) plot_map(africa, europe, anatolia)"},{"path":"https://www.slendr.net/index.html","id":"id_3-define-demographic-history-and-population-boundaries","dir":"","previous_headings":"Example","what":"3. Define demographic history and population boundaries","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"important function slendr package population(), used define names, split times, sizes spatial ranges populations. , specify times years present, distances kilometers. makes sense models, times can also given forward direction. also note functions move() expand_range() designed take slendr population object schedule spatial dynamics appropriate times model simulation (happen later step). Note order make example executable reasonable time extremely old laptop, decreased sizes populations unrealistic levels. speed SLiM simulation later step. can use function plot_map get “compressed” overview spatio-temporal range dynamics encoded model far (prior simulation ).","code":"afr <- population( # African ancestral population   \"AFR\", time = 52000, N = 3000, map = map, polygon = africa )  ooa <- population( # population of the first migrants out of Africa   \"OOA\", parent = afr, time = 51000, N = 500, remove = 25000,   center = c(33, 30), radius = 400e3 ) %>%   move(     trajectory = list(c(40, 30), c(50, 30), c(60, 40)),     start = 50000, end = 40000, snapshots = 20   )  ehg <- population( # Eastern hunter-gatherers   \"EHG\", parent = ooa, time = 28000, N = 1000, remove = 6000,   polygon = list(     c(26, 55), c(38, 53), c(48, 53), c(60, 53),     c(60, 60), c(48, 63), c(38, 63), c(26, 60)) )  eur <- population( # European population   name = \"EUR\", parent = ehg, time = 25000, N = 2000,   polygon = europe )  ana <- population( # Anatolian farmers   name = \"ANA\", time = 28000, N = 3000, parent = ooa, remove = 4000,   center = c(34, 38), radius = 500e3, polygon = anatolia ) %>%   expand_range( # expand the range by 2.500 km     by = 2500e3, start = 10000, end = 7000,     polygon = join(europe, anatolia), snapshots = 20   )  yam <- population( # Yamnaya steppe population   name = \"YAM\", time = 7000, N = 500, parent = ehg, remove = 2500,   polygon = list(c(26, 50), c(38, 49), c(48, 50),                  c(48, 56), c(38, 59), c(26, 56)) ) %>%   move(trajectory = list(c(15, 50)), start = 5000, end = 3000, snapshots = 10) plot_map(afr, ooa, ehg, eur, ana, yam)"},{"path":"https://www.slendr.net/index.html","id":"id_4-define-gene-flow-events","dir":"","previous_headings":"Example","what":"4. Define gene-flow events","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"default, populations slendr mix even overlapping. order schedule gene-flow event two populations, can use function gene_flow. want specify multiple events , can collect events simple R list:","code":"gf <- list(   gene_flow(from = ana, to = yam, rate = 0.5, start = 6500, end = 6400, overlap = FALSE),   gene_flow(from = ana, to = eur, rate = 0.5, start = 8000, end = 6000),   gene_flow(from = yam, to = eur, rate = 0.75, start = 4000, end = 3000) )"},{"path":"https://www.slendr.net/index.html","id":"id_5-compile-the-model-to-a-set-of-configuration-files","dir":"","previous_headings":"Example","what":"5. Compile the model to a set of configuration files","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"run simulation, compile individual model components (population objects gene-flow events) single R object, specifying additional model parameters. Additionally, performs internal consistency checks, making sure model parameters (split times, gene flow times, etc.) make sense (potentially quite computationally costly) simulation even run. Compiled model kept R object can passed different functions. example, simulate data slim() engine (also run simulation coalescent engine via msprime() function).","code":"model <- compile_model(   populations = list(afr, ooa, ehg, eur, ana, yam), # populations defined above   gene_flow = gf,   generation_time = 30,   resolution = 10e3,                   # resolution in meters per pixel   competition = 130e3, mating = 100e3, # spatial interaction parameters   dispersal = 70e3,                    # how far can offspring end up from their parents )"},{"path":"https://www.slendr.net/index.html","id":"id_6-visualize-the-model","dir":"","previous_headings":"Example","what":"6. Visualize the model","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"package provides R shiny-based browser app explore_model() checking model dynamics interactively visually. complex models, much better static spatial plots one showed step 2 : function two modes: Plotting (“playing”) spatial map dynamics:  Displaying demographic history graph (splits gene-flow events) embedded specified model:","code":"explore_model(model)"},{"path":"https://www.slendr.net/index.html","id":"id_7-run-the-model-in-slim","dir":"","previous_headings":"Example","what":"7. Run the model in SLiM","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"Finally, can execute compiled model SLiM. run simulation batch mode, also run SLiMgui setting method = \"gui\". allow us inspect spatial simulation happens real time. slim function generates complete SLiM script tailored run spatial model defined . saves , user, tremendous amount time, don’t write new SLiM code every time design new demographic model. output simulation run slendr model always tree sequence, loaded object ts_slim. specified , slendr’s SLiM backend simulate 10 Mb sequence individual, produce tree sequence output simulation run can analysed many built-population genetic functions. default, individuals living end simulation recorded samples tree sequence. specific set samples (ancient modern) needed, can defined accordingly using dedicated function. Note although defined spatial model, just easily simulated standard, non-spatial data running model slendr’s msprime() back end without need make changes: quick overview SLiM simulation run summarised GIF animation. , please note simulation extremely simplified. simulated small number individuals population, also didn’t specify dispersal dynamics populations look clumped.  point, either compute population genetic statistics interest perhaps analyse spatial features genealogies simulated model.","code":"ts_slim <- slim(model, sequence_length = 10e6, recombination_rate = 1e-8,                 method = \"batch\", random_seed = 314159) ts_msprime <- msprime(model, sequence_length = 10e6, recombination_rate = 1e-8) animate_model(model = model, file = locations_file, steps = 50, width = 500, height = 300)"},{"path":"https://www.slendr.net/index.html","id":"further-information","dir":"","previous_headings":"","what":"Further information","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"example provides brief incomplete overview full functionality slendr package. much slendr demonstrated . instance: can tweak parameters influencing dispersal dynamics (“clumpy” populations , far can offspring migrate parents, etc.) define change time. instance, can see animation , African population forms single “blob” really isn’t spread across entire population range. Tweaking dispersal parameters show vignette helps avoid . can use slendr program non-spatial models, means conceivable traditional, random-mating demographic model can simulated lines R code. can learn vignette (detail vignette). SLiM simulations can often quite slow compared coalescent counterparts, also provide functionality allowing simulate slendr models (without change!) using built-msprime back end script. See vignette tutorial works. can build complex spatial models still abstract (assuming real geographic location), including traditional simulations demes lattice structure. complete example shown vignette. slendr & SLiM save data tree-sequence file format, thanks R package reticulate interfacing Python code, full power tskit pyslim manipulating tree-sequence data right fingertips, within convenient environment R. extended example can found vignette. spatially explicit population models, slendr package automatically converts simulated output data format makes possible analyse many available R packages geospatial data analysis. brief description functionality can found vignette. can find complete reproducible code behind examples paper dedicated R vignette .","code":""},{"path":"https://www.slendr.net/reference/animate_model.html","id":null,"dir":"Reference","previous_headings":"","what":"Animate the simulated population dynamics — animate_model","title":"Animate the simulated population dynamics — animate_model","text":"Animate simulated population dynamics","code":""},{"path":"https://www.slendr.net/reference/animate_model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Animate the simulated population dynamics — animate_model","text":"","code":"animate_model(model, file, steps, gif = NULL, width = 800, height = 560)"},{"path":"https://www.slendr.net/reference/animate_model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Animate the simulated population dynamics — animate_model","text":"model Compiled slendr_model model object file Path table saved individual locations steps many frames animation ? gif Path output GIF file (animation object returned default) width, height Dimensions animation pixels","code":""},{"path":"https://www.slendr.net/reference/animate_model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Animate the simulated population dynamics — animate_model","text":"gif = NULL, return gganimate animation object. Otherwise GIF file saved value returned.","code":""},{"path":"https://www.slendr.net/reference/area.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the area covered by the given slendr object — area","title":"Calculate the area covered by the given slendr object — area","text":"Calculate area covered given slendr object","code":""},{"path":"https://www.slendr.net/reference/area.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the area covered by the given slendr object — area","text":"","code":"area(x)"},{"path":"https://www.slendr.net/reference/area.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the area covered by the given slendr object — area","text":"x Object class slendr","code":""},{"path":"https://www.slendr.net/reference/area.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the area covered by the given slendr object — area","text":"Area covered input object. slendr_pop given, table population range area time point returned. slendr_region slendr_world object specified, total area covered object's spatial boundary returned.","code":""},{"path":"https://www.slendr.net/reference/area.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the area covered by the given slendr object — area","text":"","code":"region_a <- region(\"A\", center = c(20, 50), radius = 20) region_b <- region(\"B\", polygon = list(c(50, 40), c(70, 40), c(70, 60), c(50, 60))) plot_map(region_a, region_b)   # note that area won't be *exactly* equal to pi*r^2: #   https://stackoverflow.com/a/65280376 area(region_a) #> [1] 1256.063  area(region_b) #> [1] 400"},{"path":"https://www.slendr.net/reference/as.phylo.slendr_phylo.html","id":null,"dir":"Reference","previous_headings":"","what":"Convert an annotated slendr_phylo object to a phylo object — as.phylo.slendr_phylo","title":"Convert an annotated slendr_phylo object to a phylo object — as.phylo.slendr_phylo","text":"function servers workaround around ggtree error: Error UseMethod(\".phylo\") :     applicable method '.phylo' applied object class     \"c('phylo', 'slendr_phylo')\"","code":""},{"path":"https://www.slendr.net/reference/as.phylo.slendr_phylo.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Convert an annotated slendr_phylo object to a phylo object — as.phylo.slendr_phylo","text":"","code":"# S3 method for slendr_phylo as.phylo(x, ...)"},{"path":"https://www.slendr.net/reference/as.phylo.slendr_phylo.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Convert an annotated slendr_phylo object to a phylo object — as.phylo.slendr_phylo","text":"x Tree object class slendr_phylo ... Additional (unused) arguments .phylo S3 method","code":""},{"path":"https://www.slendr.net/reference/as.phylo.slendr_phylo.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Convert an annotated slendr_phylo object to a phylo object — as.phylo.slendr_phylo","text":"Standard phylogenetic tree object implemented R package ape","code":""},{"path":"https://www.slendr.net/reference/check_dependencies.html","id":null,"dir":"Reference","previous_headings":"","what":"Check that the required dependencies are available for slendr to work — check_dependencies","title":"Check that the required dependencies are available for slendr to work — check_dependencies","text":"Check required dependencies available slendr work","code":""},{"path":"https://www.slendr.net/reference/check_dependencies.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check that the required dependencies are available for slendr to work — check_dependencies","text":"","code":"check_dependencies(python = FALSE, slim = FALSE, quit = FALSE)"},{"path":"https://www.slendr.net/reference/check_dependencies.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check that the required dependencies are available for slendr to work — check_dependencies","text":"python slendr Python environment required? slim SLiM required? quit R interpreter quit required slendr dependencies missing? option (turned default, set FALSE) used mainly avoiding running slendr man page examples machines lack dependencies. set TRUE, logical value returned.","code":""},{"path":"https://www.slendr.net/reference/check_dependencies.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check that the required dependencies are available for slendr to work — check_dependencies","text":"quit = TRUE, values returned, quit = FALSE, scalar logical value returned indicating whether dependencies present.","code":""},{"path":"https://www.slendr.net/reference/check_env.html","id":null,"dir":"Reference","previous_headings":"","what":"Check that the active Python environment is setup for slendr — check_env","title":"Check that the active Python environment is setup for slendr — check_env","text":"function inspects Python environment activated reticulate package prints versions slendr Python dependencies console.","code":""},{"path":"https://www.slendr.net/reference/check_env.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check that the active Python environment is setup for slendr — check_env","text":"","code":"check_env(verbose = TRUE)"},{"path":"https://www.slendr.net/reference/check_env.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check that the active Python environment is setup for slendr — check_env","text":"verbose log message printed? FALSE, logical value returned (invisibly).","code":""},{"path":"https://www.slendr.net/reference/check_env.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check that the active Python environment is setup for slendr — check_env","text":"Either TRUE (slendr Python environment present) FALSE (slendr Python environment present).","code":""},{"path":"https://www.slendr.net/reference/check_env.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Check that the active Python environment is setup for slendr — check_env","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated. check_env() #> Summary of the currently active Python environment: #>  #> Python binary: /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.0_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python  #> Python version: 3.12.1 | packaged by conda-forge | (main, Dec 23 2023, 08:01:35) [Clang 16.0.6 ]  #>  #> slendr requirements: #>  - tskit: version 0.5.6 ✓  #>  - msprime: version 1.3.0 ✓  #>  - pyslim: version 1.0.4 ✓  #>  - tspop: present ✓"},{"path":"https://www.slendr.net/reference/clear_env.html","id":null,"dir":"Reference","previous_headings":"","what":"Remove the automatically created slendr Python environment — clear_env","title":"Remove the automatically created slendr Python environment — clear_env","text":"Remove automatically created slendr Python environment","code":""},{"path":"https://www.slendr.net/reference/clear_env.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Remove the automatically created slendr Python environment — clear_env","text":"","code":"clear_env(force = FALSE)"},{"path":"https://www.slendr.net/reference/clear_env.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Remove the automatically created slendr Python environment — clear_env","text":"force Ask deleting environment?","code":""},{"path":"https://www.slendr.net/reference/clear_env.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Remove the automatically created slendr Python environment — clear_env","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/compile_model.html","id":null,"dir":"Reference","previous_headings":"","what":"Compile a slendr demographic model — compile_model","title":"Compile a slendr demographic model — compile_model","text":"First, compiles vectorized population spatial maps series binary raster PNG files, format SLiM understands uses define population boundaries. extracts demographic model defined user (.e. population divergences gene flow events) series tables later used built-SLiM script program timing simulation events.","code":""},{"path":"https://www.slendr.net/reference/compile_model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compile a slendr demographic model — compile_model","text":"","code":"compile_model(   populations,   generation_time,   path = NULL,   resolution = NULL,   competition = NULL,   mating = NULL,   dispersal = NULL,   gene_flow = list(),   overwrite = FALSE,   force = FALSE,   simulation_length = NULL,   direction = NULL,   serialize = TRUE,   slim_script = NULL,   description = \"\",   sim_length = NULL )"},{"path":"https://www.slendr.net/reference/compile_model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compile a slendr demographic model — compile_model","text":"populations Object(s) slendr_pop class (multiple objects need specified list) generation_time Generation time (model time units) path Output directory model configuration files loaded backend SLiM script. NULL, model configuration files saved temporary directory. resolution many distance units per pixel? competition, mating Maximum spatial competition mating choice distance dispersal Standard deviation normal distribution parent-offspring distance gene_flow Gene flow events generated gene_flow function (either list data.frame objects format defined gene_flow function, single data.frame) overwrite Completely delete specified directory, case already exists, create new one? force Force deletion model directory already present? Useful non-interactive uses. interactive mode, user asked confirm deletion manually. simulation_length Total length simulation (required forward time models, optional models specified backward time units default run \"present time\") direction Intended direction time. normal circumstances parameter inferred model need set manually. serialize model files serialized disk? , R model object returned files created. speeds simulation msprime prevents using SLiM back end. slim_script Path SLiM script used executing model (default, bundled backend script used). NULL, SLiM script bundled slendr used. description Optional short description model sim_length Deprecated. Use simulation_length instead.","code":""},{"path":"https://www.slendr.net/reference/compile_model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Compile a slendr demographic model — compile_model","text":"Compiled slendr_model model object encapsulates information specified model (populations involved, much gene flow occur, spatial resolution map, spatial dispersal mating parameters used SLiM simulation, applicable)","code":""},{"path":"https://www.slendr.net/reference/compile_model.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Compile a slendr demographic model — compile_model","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/distance.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the distance between a pair of spatial boundaries — distance","title":"Calculate the distance between a pair of spatial boundaries — distance","text":"Calculate distance pair spatial boundaries","code":""},{"path":"https://www.slendr.net/reference/distance.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the distance between a pair of spatial boundaries — distance","text":"","code":"distance(x, y, measure, time = NULL)"},{"path":"https://www.slendr.net/reference/distance.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the distance between a pair of spatial boundaries — distance","text":"x, y Objects class slendr measure measure distance? can either 'border' (distance borders x y) 'center' (distance centroids). time Time closest spatial maps x y represent slendr_pop population boundaries (ignored general slendr_region objects)","code":""},{"path":"https://www.slendr.net/reference/distance.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the distance between a pair of spatial boundaries — distance","text":"coordinate reference system specified, distance projected units (.e. meters) returned. Otherwise function returns normal Euclidean distance.","code":""},{"path":"https://www.slendr.net/reference/distance.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the distance between a pair of spatial boundaries — distance","text":"","code":"# create two regions on a blank abstract landscape region_a <- region(\"A\", center = c(20, 50), radius = 20) region_b <- region(\"B\", center = c(80, 50), radius = 20) plot_map(region_a, region_b)   # compute the distance between the centers of both population ranges distance(region_a, region_b, measure = \"center\") #> [1] 60  # compute the distance between the borders of both population ranges distance(region_a, region_b, measure = \"border\") #> [1] 20"},{"path":"https://www.slendr.net/reference/expand_range.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand the population range — expand_range","title":"Expand the population range — expand_range","text":"Expands spatial population range specified distance given time-window","code":""},{"path":"https://www.slendr.net/reference/expand_range.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand the population range — expand_range","text":"","code":"expand_range(   pop,   by,   end,   start,   overlap = 0.8,   snapshots = NULL,   polygon = NULL,   lock = FALSE,   verbose = TRUE )"},{"path":"https://www.slendr.net/reference/expand_range.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand the population range — expand_range","text":"pop Object class slendr_pop many units distance expand ? start, end expansion start/end? overlap Minimum overlap subsequent spatial boundaries snapshots number intermediate snapshots (overrides overlap parameter) polygon Geographic region restrict expansion lock Maintain density individuals. FALSE (default), number individuals population change. TRUE, number individuals simulated changed (increased decreased) appropriately, match new population range area. verbose Report progress generating intermediate spatial boundaries?","code":""},{"path":"https://www.slendr.net/reference/expand_range.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand the population range — expand_range","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/expand_range.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Expand the population range — expand_range","text":"Note slendr models accomodate SLiM msprime back ends, population sizes times events rounded nearest integer value.","code":""},{"path":"https://www.slendr.net/reference/expand_range.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Expand the population range — expand_range","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/explore_model.html","id":null,"dir":"Reference","previous_headings":"","what":"Open an interactive browser of the spatial model — explore_model","title":"Open an interactive browser of the spatial model — explore_model","text":"Open interactive browser spatial model","code":""},{"path":"https://www.slendr.net/reference/explore_model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Open an interactive browser of the spatial model — explore_model","text":"","code":"explore_model(model)"},{"path":"https://www.slendr.net/reference/explore_model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Open an interactive browser of the spatial model — explore_model","text":"model Compiled slendr_model model object","code":""},{"path":"https://www.slendr.net/reference/explore_model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Open an interactive browser of the spatial model — explore_model","text":"return value, called order start interactive browser-based interface explore dynamics slendr model","code":""},{"path":"https://www.slendr.net/reference/gene_flow.html","id":null,"dir":"Reference","previous_headings":"","what":"Define a gene-flow event between two populations — gene_flow","title":"Define a gene-flow event between two populations — gene_flow","text":"Define gene-flow event two populations","code":""},{"path":"https://www.slendr.net/reference/gene_flow.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Define a gene-flow event between two populations — gene_flow","text":"","code":"gene_flow(from, to, rate, start, end, overlap = TRUE)"},{"path":"https://www.slendr.net/reference/gene_flow.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Define a gene-flow event between two populations — gene_flow","text":", Objects class slendr_pop rate Scalar value range (0, 1] specifying proportion migration given time period start, end Start end gene-flow event overlap Require spatial overlap admixing populations?  (default TRUE)","code":""},{"path":"https://www.slendr.net/reference/gene_flow.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Define a gene-flow event between two populations — gene_flow","text":"Object class data.frame containing parameters specified gene-flow event.","code":""},{"path":"https://www.slendr.net/reference/gene_flow.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Define a gene-flow event between two populations — gene_flow","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/init_env.html","id":null,"dir":"Reference","previous_headings":"","what":"Activate slendr's own dedicated Python environment — init_env","title":"Activate slendr's own dedicated Python environment — init_env","text":"function attempts activate dedicated slendr Miniconda Python environment previously set via setup_env.","code":""},{"path":"https://www.slendr.net/reference/init_env.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Activate slendr's own dedicated Python environment — init_env","text":"","code":"init_env(quiet = FALSE)"},{"path":"https://www.slendr.net/reference/init_env.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Activate slendr's own dedicated Python environment — init_env","text":"quiet informative messages printed console? Default FALSE.","code":""},{"path":"https://www.slendr.net/reference/init_env.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Activate slendr's own dedicated Python environment — init_env","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/join.html","id":null,"dir":"Reference","previous_headings":"","what":"Merge two spatial slendr objects into one — join","title":"Merge two spatial slendr objects into one — join","text":"Merge two spatial slendr objects one","code":""},{"path":"https://www.slendr.net/reference/join.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Merge two spatial slendr objects into one — join","text":"","code":"join(x, y, name = NULL)"},{"path":"https://www.slendr.net/reference/join.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Merge two spatial slendr objects into one — join","text":"x Object class slendr y Object class slendr name Optional name resulting geographic region. missing, name constructed function arguments.","code":""},{"path":"https://www.slendr.net/reference/join.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Merge two spatial slendr objects into one — join","text":"Object class slendr_region encodes standard spatial object class sf several additional attributes (importantly corresponding slendr_map object, applicable).","code":""},{"path":"https://www.slendr.net/reference/join.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Merge two spatial slendr objects into one — join","text":"","code":"# create a blank abstract world 1000x1000 distance units in size blank_map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # it is possible to construct custom landscapes (islands, corridors, etc.) island1 <- region(\"island1\", polygon = list(c(10, 30), c(50, 30), c(40, 50), c(0, 40))) island2 <- region(\"island2\", polygon = list(c(60, 60), c(80, 40), c(100, 60), c(80, 80))) island3 <- region(\"island3\", center = c(20, 80), radius = 10) archipelago <- island1 %>% join(island2) %>% join(island3)  custom_map <- world(xrange = c(1, 100), c(1, 100), landscape = archipelago)  # real Earth landscapes can be defined using freely-available Natural Earth # project data and with the possibility to specify an appropriate Coordinate # Reference System, such as this example of a map of Europe  real_map <- world(xrange = c(-15, 40), yrange = c(30, 60), crs = \"EPSG:3035\") #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpFhhY5X/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84"},{"path":"https://www.slendr.net/reference/move.html","id":null,"dir":"Reference","previous_headings":"","what":"Move the population to a new location in a given amount of time — move","title":"Move the population to a new location in a given amount of time — move","text":"function defines displacement population along given trajectory given time frame","code":""},{"path":"https://www.slendr.net/reference/move.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Move the population to a new location in a given amount of time — move","text":"","code":"move(   pop,   trajectory,   end,   start,   overlap = 0.8,   snapshots = NULL,   verbose = TRUE )"},{"path":"https://www.slendr.net/reference/move.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Move the population to a new location in a given amount of time — move","text":"pop Object class slendr_pop trajectory List two-dimensional vectors (longitude, latitude) specifying migration trajectory start, end Start/end points population migration overlap Minimum overlap subsequent spatial boundaries snapshots number intermediate snapshots (overrides overlap parameter) verbose Show progress searching number sufficient snapshots?","code":""},{"path":"https://www.slendr.net/reference/move.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Move the population to a new location in a given amount of time — move","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/move.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Move the population to a new location in a given amount of time — move","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/msprime.html","id":null,"dir":"Reference","previous_headings":"","what":"Run a slendr model in msprime — msprime","title":"Run a slendr model in msprime — msprime","text":"function execute built-msprime script run compiled slendr demographic model.","code":""},{"path":"https://www.slendr.net/reference/msprime.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Run a slendr model in msprime — msprime","text":"","code":"msprime(   model,   sequence_length,   recombination_rate,   samples = NULL,   output = NULL,   random_seed = NULL,   load = TRUE,   verbose = FALSE,   debug = FALSE,   run = TRUE )"},{"path":"https://www.slendr.net/reference/msprime.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Run a slendr model in msprime — msprime","text":"model Model object created compile function sequence_length Total length simulated sequence (base-pairs) recombination_rate Recombination rate simulated sequence (recombinations per basepair per generation) samples data frame times given number individuals remembered tree-sequence (see schedule_sampling function can generate sampling schedule correct format). missing, individuals present end simulation recorded tree-sequence output file. output Path output tree-sequence file. NULL (default), tree sequence saved temporary file. random_seed Random seed (NULL, seed generated 0 maximum integer number available) load final tree sequence immediately loaded returned? Default TRUE. alternative (FALSE) useful tree-sequence file written custom location loaded later point. verbose Write output log console (default FALSE)? debug Write msprime's debug log console (default FALSE)? run msprime engine run? FALSE, command line msprime command printed (returned invisibly character vector) executed.","code":""},{"path":"https://www.slendr.net/reference/msprime.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Run a slendr model in msprime — msprime","text":"tree-sequence object loaded via Python-R reticulate interface function ts_load (internally represented Python object tskit.trees.TreeSequence). Optionally, depending value arguments load = run =, nothing character vector, respectively.","code":""},{"path":"https://www.slendr.net/reference/msprime.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Run a slendr model in msprime — msprime","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # afr and eur objects would normally be created before slendr model compilation, # but here we take them out of the model object already compiled for this # example (in a standard slendr simulation pipeline, this wouldn't be necessary) afr <- model$populations[[\"AFR\"]] eur <- model$populations[[\"EUR\"]] chimp <- model$populations[[\"CH\"]]  # schedule the sampling of a couple of ancient and present-day individuals # given model at 20 ky, 10 ky, 5ky ago and at present-day (time 0) modern_samples <- schedule_sampling(model, times = 0, list(afr, 10), list(eur, 100), list(chimp, 1)) ancient_samples <- schedule_sampling(model, times = c(40000, 30000, 20000, 10000), list(eur, 1))  # sampling schedules are just data frames and can be merged easily samples <- rbind(modern_samples, ancient_samples)  # run a simulation using the msprime back end from a compiled slendr model object ts <- msprime(model, sequence_length = 1e5, recombination_rate = 0, samples = samples)  # automatic loading of a simulated output can be prevented by `load = FALSE`, which can be # useful when a custom path to a tree-sequence output is given for later downstream analyses output_file <- tempfile(fileext = \".trees\") msprime(model, sequence_length = 1e5, recombination_rate = 0, samples = samples,         output = output_file, load = FALSE, random_seed = 42) # ... at a later stage: ts <- ts_load(output_file, model)  summary(ts) #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │          1║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│     100000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        230║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   39.4 KiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤════╤═════════╤════════════╗ #> ║Table      │Rows│Size     │Has Metadata║ #> ╠═══════════╪════╪═════════╪════════════╣ #> ║Edges      │ 471│ 14.7 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Individuals│ 115│  3.2 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Migrations │   0│  8 Bytes│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Mutations  │   0│ 16 Bytes│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Nodes      │ 472│ 12.9 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Populations│   4│338 Bytes│         Yes║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Provenances│   1│  2.9 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Sites      │   0│ 16 Bytes│          No║ #> ╚═══════════╧════╧═════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/overlap.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate the overlap of two slendr objects — overlap","title":"Generate the overlap of two slendr objects — overlap","text":"Generate overlap two slendr objects","code":""},{"path":"https://www.slendr.net/reference/overlap.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate the overlap of two slendr objects — overlap","text":"","code":"overlap(x, y, name = NULL)"},{"path":"https://www.slendr.net/reference/overlap.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate the overlap of two slendr objects — overlap","text":"x Object class slendr y Object class slendr name Optional name resulting geographic region. missing, name constructed function arguments.","code":""},{"path":"https://www.slendr.net/reference/overlap.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generate the overlap of two slendr objects — overlap","text":"Object class slendr_region encodes standard spatial object class sf several additional attributes (importantly corresponding slendr_map object, applicable).","code":""},{"path":"https://www.slendr.net/reference/pipe.html","id":null,"dir":"Reference","previous_headings":"","what":"Pipe operator — %>%","title":"Pipe operator — %>%","text":"Pipe operator","code":""},{"path":"https://www.slendr.net/reference/pipe.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Pipe operator — %>%","text":"","code":"lhs %>% rhs"},{"path":"https://www.slendr.net/reference/pipe.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Pipe operator — %>%","text":"See magrittr::%>% details.","code":""},{"path":"https://www.slendr.net/reference/plot_map.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot slendr geographic features on a map — plot_map","title":"Plot slendr geographic features on a map — plot_map","text":"Plots objects three slendr spatial classes (slendr_map, slendr_region, slendr_pop).","code":""},{"path":"https://www.slendr.net/reference/plot_map.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot slendr geographic features on a map — plot_map","text":"","code":"plot_map(   ...,   time = NULL,   gene_flow = FALSE,   graticules = \"original\",   intersect = TRUE,   show_map = TRUE,   title = NULL,   interpolated_maps = NULL )"},{"path":"https://www.slendr.net/reference/plot_map.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot slendr geographic features on a map — plot_map","text":"... Objects classes slendr_map, slendr_region, slendr_pop time Plot concrete time point gene_flow Indicate geneflow events arrow graticules Plot graticules original Coordinate Reference System (longitude-latitude), internal CRS (meters)? intersect Intersect population boundaries landscape geographic boundaries (default TRUE)? show_map Show underlying world map title Title plot interpolated_maps Interpolated spatial boundaries populations time points (used plotting using explore shiny app)","code":""},{"path":"https://www.slendr.net/reference/plot_map.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot slendr geographic features on a map — plot_map","text":"ggplot2 object visualized slendr map","code":""},{"path":"https://www.slendr.net/reference/plot_model.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot demographic history encoded in a slendr model — plot_model","title":"Plot demographic history encoded in a slendr model — plot_model","text":"Plot demographic history encoded slendr model","code":""},{"path":"https://www.slendr.net/reference/plot_model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot demographic history encoded in a slendr model — plot_model","text":"","code":"plot_model(   model,   sizes = TRUE,   proportions = FALSE,   gene_flow = TRUE,   log = FALSE,   order = NULL,   file = NULL,   ... )"},{"path":"https://www.slendr.net/reference/plot_model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot demographic history encoded in a slendr model — plot_model","text":"model Compiled slendr_model model object sizes population size changes visualized? proportions gene flow proportions visualized (FALSE default prevent cluttering overplotting) gene_flow gene-flow arrows visualized (default TRUE). log y-axis plotted log scale? Useful models long time-scales. order Order populations along x-axis, given character vector population names. NULL (default), default plotting algorithm used, ordering populations ancestral recent using -order tree traversal. file Output file figure saved via ggsave ... Optional argument passed ggsave","code":""},{"path":"https://www.slendr.net/reference/plot_model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot demographic history encoded in a slendr model — plot_model","text":"ggplot2 object visualized slendr model","code":""},{"path":"https://www.slendr.net/reference/plot_model.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot demographic history encoded in a slendr model — plot_model","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence path <- system.file(\"extdata/models/introgression\", package = \"slendr\") model <- read_model(path)  plot_model(model, sizes = FALSE, log = TRUE)"},{"path":"https://www.slendr.net/reference/population.html","id":null,"dir":"Reference","previous_headings":"","what":"Define a population — population","title":"Define a population — population","text":"Defines parameters population (non-spatial spatial).","code":""},{"path":"https://www.slendr.net/reference/population.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Define a population — population","text":"","code":"population(   name,   time,   N,   parent = NULL,   map = FALSE,   center = NULL,   radius = NULL,   polygon = NULL,   remove = NULL,   intersect = TRUE,   competition = NA,   mating = NA,   dispersal = NA,   dispersal_fun = NULL,   aquatic = FALSE )"},{"path":"https://www.slendr.net/reference/population.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Define a population — population","text":"name Name population time Time population's first appearance N Number individuals time first appearance parent Parent population object NULL (indicates population ancestor, first population \"lineage\") map Object type slendr_map defines world context (created using world function). value FALSE provided, non-spatial model run. center Two-dimensional vector specifying center circular range radius Radius circular range polygon List vector pairs, defining corners polygon range geographic region class slendr_region polygon coordinates extracted (see region() function) remove Time population removed intersect Intersect population's boundaries landscape features? competition, mating Maximum spatial competition mating choice distance dispersal Standard deviation normal distribution distance offspring disperses parent dispersal_fun Distribution function governing dispersal offspring. One \"normal\", \"uniform\", \"cauchy\", \"exponential\", \"brownian\" (vertical horizontal displacements drawn normal distribution independently). aquatic species aquatic (FALSE default, .e. terrestrial species)?","code":""},{"path":"https://www.slendr.net/reference/population.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Define a population — population","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/population.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Define a population — population","text":"four ways specify spatial boundary: ) circular range specified using center coordinate radius, ii) polygon specified list two-dimensional vector coordinates, iii) polygon ii), defined (named) using region function, iv) just world map specified (circular polygon range parameters set default NULL value), population allowed occupy entire landscape. Note slendr models accomodate SLiM msprime back ends, population sizes split times rounded nearest integer value.","code":""},{"path":"https://www.slendr.net/reference/population.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Define a population — population","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/print.slendr_pop.html","id":null,"dir":"Reference","previous_headings":"","what":"Print a short summary of a slendr object — print.slendr_pop","title":"Print a short summary of a slendr object — print.slendr_pop","text":"spatial objects slendr package internally represented Simple Features (sf) objects. fact hidden circumstances , goal slendr package provide functionality much higher level (population boundaries, geographic regions, instead individual polygons \"low-level\" geometric objects), without users worry low-level details involved handling spatial geometries. However, full sf object representation can always printed calling x[].","code":""},{"path":"https://www.slendr.net/reference/print.slendr_pop.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print a short summary of a slendr object — print.slendr_pop","text":"","code":"# S3 method for slendr_pop print(x, ...)  # S3 method for slendr_region print(x, ...)  # S3 method for slendr_map print(x, ...)  # S3 method for slendr_model print(x, ...)"},{"path":"https://www.slendr.net/reference/print.slendr_pop.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print a short summary of a slendr object — print.slendr_pop","text":"x Object class slendr (either slendr_pop, slendr_map, slendr_region, slendr_table) ... Additional arguments passed print","code":""},{"path":"https://www.slendr.net/reference/print.slendr_pop.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print a short summary of a slendr object — print.slendr_pop","text":"return value, used printing","code":""},{"path":"https://www.slendr.net/reference/print.slendr_ts.html","id":null,"dir":"Reference","previous_headings":"","what":"Print tskit's summary table of the Python tree-sequence object — print.slendr_ts","title":"Print tskit's summary table of the Python tree-sequence object — print.slendr_ts","text":"Print tskit's summary table Python tree-sequence object","code":""},{"path":"https://www.slendr.net/reference/print.slendr_ts.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print tskit's summary table of the Python tree-sequence object — print.slendr_ts","text":"","code":"# S3 method for slendr_ts print(x, ...)"},{"path":"https://www.slendr.net/reference/print.slendr_ts.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print tskit's summary table of the Python tree-sequence object — print.slendr_ts","text":"x Tree object class slendr_phylo ... Additional arguments normally passed print (used case)","code":""},{"path":"https://www.slendr.net/reference/print.slendr_ts.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print tskit's summary table of the Python tree-sequence object — print.slendr_ts","text":"return value, simply prints tskit summary table terminal","code":""},{"path":"https://www.slendr.net/reference/read_model.html","id":null,"dir":"Reference","previous_headings":"","what":"Read a previously serialized model configuration — read_model","title":"Read a previously serialized model configuration — read_model","text":"Reads configuration tables model data location previously compiled compile function.","code":""},{"path":"https://www.slendr.net/reference/read_model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Read a previously serialized model configuration — read_model","text":"","code":"read_model(path)"},{"path":"https://www.slendr.net/reference/read_model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Read a previously serialized model configuration — read_model","text":"path Directory required configuration files","code":""},{"path":"https://www.slendr.net/reference/read_model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Read a previously serialized model configuration — read_model","text":"Compiled slendr_model model object encapsulates information specified model (populations involved, much gene flow occur, spatial resolution map, spatial dispersal mating parameters used SLiM simulation, applicable)","code":""},{"path":"https://www.slendr.net/reference/read_model.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Read a previously serialized model configuration — read_model","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence path <- system.file(\"extdata/models/introgression\", package = \"slendr\") model <- read_model(path)  plot_model(model, sizes = FALSE, log = TRUE)"},{"path":"https://www.slendr.net/reference/region.html","id":null,"dir":"Reference","previous_headings":"","what":"Define a geographic region — region","title":"Define a geographic region — region","text":"Creates geographic region (polygon) given map gives name. can used define objects can reused multiple places slendr script (region arguments population) without repeatedly define polygon coordinates.","code":""},{"path":"https://www.slendr.net/reference/region.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Define a geographic region — region","text":"","code":"region(name = NULL, map = NULL, center = NULL, radius = NULL, polygon = NULL)"},{"path":"https://www.slendr.net/reference/region.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Define a geographic region — region","text":"name Name geographic region map Object type sf defines map center Two-dimensional vector specifying center circular range radius Radius circular range polygon List vector pairs, defining corners polygon range geographic region class slendr_region polygon coordinates extracted (see region() function)","code":""},{"path":"https://www.slendr.net/reference/region.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Define a geographic region — region","text":"Object class slendr_region encodes standard spatial object class sf several additional attributes (importantly corresponding slendr_map object, applicable).","code":""},{"path":"https://www.slendr.net/reference/region.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Define a geographic region — region","text":"","code":"# create a blank abstract world 1000x1000 distance units in size blank_map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # it is possible to construct custom landscapes (islands, corridors, etc.) island1 <- region(\"island1\", polygon = list(c(10, 30), c(50, 30), c(40, 50), c(0, 40))) island2 <- region(\"island2\", polygon = list(c(60, 60), c(80, 40), c(100, 60), c(80, 80))) island3 <- region(\"island3\", center = c(20, 80), radius = 10) archipelago <- island1 %>% join(island2) %>% join(island3)  custom_map <- world(xrange = c(1, 100), c(1, 100), landscape = archipelago)  # real Earth landscapes can be defined using freely-available Natural Earth # project data and with the possibility to specify an appropriate Coordinate # Reference System, such as this example of a map of Europe  real_map <- world(xrange = c(-15, 40), yrange = c(30, 60), crs = \"EPSG:3035\") #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpFhhY5X/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84"},{"path":"https://www.slendr.net/reference/reproject.html","id":null,"dir":"Reference","previous_headings":"","what":"Reproject coordinates between coordinate systems — reproject","title":"Reproject coordinates between coordinate systems — reproject","text":"Converts coordinates compiled raster map (.e. pixel units) different Geographic Coordinate Systems (CRS).","code":""},{"path":"https://www.slendr.net/reference/reproject.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Reproject coordinates between coordinate systems — reproject","text":"","code":"reproject(   from,   to,   x = NULL,   y = NULL,   coords = NULL,   model = NULL,   add = FALSE,   input_prefix = \"\",   output_prefix = \"new\" )"},{"path":"https://www.slendr.net/reference/reproject.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Reproject coordinates between coordinate systems — reproject","text":", Either CRS code accepted GDAL, valid integer EPSG value, object class crs, value \"raster\" (converting /pixel coordinates), \"world\" (converting /whatever CRS set underlying map) x, y Coordinates two dimensions (missing, coordinates expected data.frame specified coords parameter columns \"x\" \"y\") coords data.frame-like object coordinates columns \"x\" \"y\" model Object class slendr_model add Add column coordinates input data.frame coords (coordinates otherwise returned separate object)? input_prefix, output_prefix Input output prefixes data frame columns spatial coordinates","code":""},{"path":"https://www.slendr.net/reference/reproject.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Reproject coordinates between coordinate systems — reproject","text":"Data.frame converted two-dimensional coordinates given input","code":""},{"path":"https://www.slendr.net/reference/reproject.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Reproject coordinates between coordinate systems — reproject","text":"","code":"lon_lat_df <- data.frame(x = c(30, 0, 15), y = c(60, 40, 10))  reproject(   from = \"epsg:4326\",   to = \"epsg:3035\",   coords = lon_lat_df,   add = TRUE # add converted [lon,lat] coordinates as a new column ) #> # A tibble: 3 × 4 #>       x     y     newx      newy #>   <dbl> <dbl>    <dbl>     <dbl> #> 1    30    60 5422493.  4256803. #> 2     0    40 3465349.  1934879. #> 3    15    10 4907297. -1328914."},{"path":"https://www.slendr.net/reference/resize.html","id":null,"dir":"Reference","previous_headings":"","what":"Change the population size — resize","title":"Change the population size — resize","text":"Resizes population starting current value N individuals specified value","code":""},{"path":"https://www.slendr.net/reference/resize.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Change the population size — resize","text":"","code":"resize(pop, N, how, time, end = NULL)"},{"path":"https://www.slendr.net/reference/resize.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Change the population size — resize","text":"pop Object class slendr_pop N Population size change change population size (options \"step\" \"exponential\") time Time population size change end End population size change period (used exponential change events)","code":""},{"path":"https://www.slendr.net/reference/resize.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Change the population size — resize","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/resize.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Change the population size — resize","text":"case exponential size change, final N larger current size, population exponentially growing specified time period reaches N individuals. N smaller, population shrink exponentially. Note slendr models accomodate SLiM msprime back ends, population sizes split times rounded nearest integer value.","code":""},{"path":"https://www.slendr.net/reference/resize.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Change the population size — resize","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":null,"dir":"Reference","previous_headings":"","what":"Define sampling events for a given set of populations — schedule_sampling","title":"Define sampling events for a given set of populations — schedule_sampling","text":"Schedule sampling events specified times , optionally, given set locations landscape","code":""},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Define sampling events for a given set of populations — schedule_sampling","text":"","code":"schedule_sampling(model, times, ..., locations = NULL, strict = FALSE)"},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Define sampling events for a given set of populations — schedule_sampling","text":"model Object class slendr_model times Integer vector times (model time units) schedule remembering individuals tree-sequence ... Lists two elements (slendr_pop population object-<number individuals sample), representing populations many individuals remembered times given times locations List vector pairs, defining two-dimensional coordinates locations closest number individuals given populations sampled. NULL (default), individuals sampled randomly throughout spatial boundary. strict occurence population present given time result error? Default FALSE, meaning invalid sampling times populations quietly ignored.","code":""},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Define sampling events for a given set of populations — schedule_sampling","text":"Data frame three columns: time sampling, population sample , many individuals sample","code":""},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Define sampling events for a given set of populations — schedule_sampling","text":"times locations given, sampling scheduled specified location given time-point. Note time-, interest simplicity, sanity checks performed locations given except restriction sampling points must fall within bounding box around simulated world map. , slendr simply instruct SLiM backend script sample individuals close sampling points given possible, regardless whether points lie within population spatial boundary particular moment time.","code":""},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Define sampling events for a given set of populations — schedule_sampling","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence path <- system.file(\"extdata/models/introgression\", package = \"slendr\") model <- read_model(path)  # afr and eur objects would normally be created before slendr model compilation, # but here we take them out of the model object already compiled for this # example (in a standard slendr simulation pipeline, this wouldn't be necessary) afr <- model$populations[[\"AFR\"]] eur <- model$populations[[\"EUR\"]]  # schedule the recording of 10 African and 100 European individuals from a # given model at 20 ky, 10 ky, 5ky ago and at present-day (time 0) schedule <- schedule_sampling(   model, times = c(20000, 10000, 5000, 0),   list(afr, 10), list(eur, 100) )  # the result of `schedule_sampling` is a simple data frame (note that the locations # of sampling locations have `NA` values because the model is non-spatial) schedule #> # A tibble: 8 × 7 #>    time pop       n y_orig x_orig y     x     #>   <int> <chr> <int> <lgl>  <lgl>  <lgl> <lgl> #> 1     0 AFR      10 NA     NA     NA    NA    #> 2     0 EUR     100 NA     NA     NA    NA    #> 3  5000 AFR      10 NA     NA     NA    NA    #> 4  5000 EUR     100 NA     NA     NA    NA    #> 5 10000 AFR      10 NA     NA     NA    NA    #> 6 10000 EUR     100 NA     NA     NA    NA    #> 7 20000 AFR      10 NA     NA     NA    NA    #> 8 20000 EUR     100 NA     NA     NA    NA"},{"path":"https://www.slendr.net/reference/set_dispersal.html","id":null,"dir":"Reference","previous_headings":"","what":"Change dispersal parameters — set_dispersal","title":"Change dispersal parameters — set_dispersal","text":"Changes either competition interactive distance, mating choice distance, dispersal offspring parent","code":""},{"path":"https://www.slendr.net/reference/set_dispersal.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Change dispersal parameters — set_dispersal","text":"","code":"set_dispersal(   pop,   time,   competition = NA,   mating = NA,   dispersal = NA,   dispersal_fun = NULL )"},{"path":"https://www.slendr.net/reference/set_dispersal.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Change dispersal parameters — set_dispersal","text":"pop Object class slendr_pop time Time population size change competition, mating Maximum spatial competition mating choice distance dispersal Standard deviation normal distribution distance offspring disperses parent dispersal_fun Distribution function governing dispersal offspring. One \"normal\", \"uniform\", \"cauchy\", \"exponential\", \"brownian\" (vertical horizontal displacements drawn normal distribution independently).","code":""},{"path":"https://www.slendr.net/reference/set_dispersal.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Change dispersal parameters — set_dispersal","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/set_dispersal.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Change dispersal parameters — set_dispersal","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/set_range.html","id":null,"dir":"Reference","previous_headings":"","what":"Update the population range — set_range","title":"Update the population range — set_range","text":"function allows manual control spatial map changes addition expand move functions","code":""},{"path":"https://www.slendr.net/reference/set_range.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Update the population range — set_range","text":"","code":"set_range(   pop,   time,   center = NULL,   radius = NULL,   polygon = NULL,   lock = FALSE )"},{"path":"https://www.slendr.net/reference/set_range.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Update the population range — set_range","text":"pop Object class slendr_pop time Time change center Two-dimensional vector specifying center circular range radius Radius circular range polygon List vector pairs, defining corners polygon range (see also region argument) geographic region class slendr_region polygon coordinates extracted lock Maintain density individuals. FALSE (default), number individuals population change. TRUE, number individuals simulated changed (increased decreased) appropriately, match new population range area.","code":""},{"path":"https://www.slendr.net/reference/set_range.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Update the population range — set_range","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/set_range.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Update the population range — set_range","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/setup_env.html","id":null,"dir":"Reference","previous_headings":"","what":"Setup a dedicated Python virtual environment for slendr — setup_env","title":"Setup a dedicated Python virtual environment for slendr — setup_env","text":"function automatically download Python miniconda distribution dedicated R-Python interface. also create slendr-specific Python environment required Python dependencies.","code":""},{"path":"https://www.slendr.net/reference/setup_env.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Setup a dedicated Python virtual environment for slendr — setup_env","text":"","code":"setup_env(quiet = FALSE, agree = FALSE, pip = NULL)"},{"path":"https://www.slendr.net/reference/setup_env.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Setup a dedicated Python virtual environment for slendr — setup_env","text":"quiet informative messages printed console? Default FALSE. agree Automatically agree questions? pip pip used instead conda installing slendr's Python dependencies? Note still use conda distribution install Python , change repository slendr install Python dependencies. Unless explicitly set TRUE, Python dependencies installed conda repositories default, expect case osx-arm64 Mac architecture, conda dependencies broken.","code":""},{"path":"https://www.slendr.net/reference/setup_env.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Setup a dedicated Python virtual environment for slendr — setup_env","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/shrink_range.html","id":null,"dir":"Reference","previous_headings":"","what":"Shrink the population range — shrink_range","title":"Shrink the population range — shrink_range","text":"Shrinks spatial population range specified distance given time-window","code":""},{"path":"https://www.slendr.net/reference/shrink_range.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Shrink the population range — shrink_range","text":"","code":"shrink_range(   pop,   by,   end,   start,   overlap = 0.8,   snapshots = NULL,   lock = FALSE,   verbose = TRUE )"},{"path":"https://www.slendr.net/reference/shrink_range.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Shrink the population range — shrink_range","text":"pop Object class slendr_pop many units distance shrink ? start, end boundary shrinking start/end? overlap Minimum overlap subsequent spatial boundaries snapshots number intermediate snapshots (overrides overlap parameter) lock Maintain density individuals. FALSE (default), number individuals population change. TRUE, number individuals simulated changed (increased decreased) appropriately, match new population range area. verbose Report progress generating intermediate spatial boundaries?","code":""},{"path":"https://www.slendr.net/reference/shrink_range.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Shrink the population range — shrink_range","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/shrink_range.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Shrink the population range — shrink_range","text":"Note slendr models accomodate SLiM msprime back ends, population sizes split times rounded nearest integer value.","code":""},{"path":"https://www.slendr.net/reference/shrink_range.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Shrink the population range — shrink_range","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/slendr.html","id":null,"dir":"Reference","previous_headings":"","what":"A Simulation Framework for Spatiotemporal Population Genetics — slendr","title":"A Simulation Framework for Spatiotemporal Population Genetics — slendr","text":"framework simulating spatially explicit genomic data leverages real cartographic information programmatic visual encoding spatiotemporal population dynamics real geographic landscapes. Population genetic models automatically executed 'SLiM' software behind scenes, using custom built-simulation 'SLiM' script. Additionally, fully abstract spatial models tied specific geographic location supported, users can also simulate data standard, non-spatial, random-mating models. can simulated either 'SLiM' built-back-end script, using efficient coalescent population genetics simulator 'msprime' custom-built 'Python' script bundled R package. Simulated genomic data saved tree-sequence format can loaded, manipulated, summarised using tree-sequence functionality via R interface 'Python' module 'tskit'. Complete model configuration, simulation analysis pipelines can therefore constructed without need leave R environment, eliminating friction disparate tools population genetic simulations data analysis.","code":""},{"path":"https://www.slendr.net/reference/slendr.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"A Simulation Framework for Spatiotemporal Population Genetics — slendr","text":"can find installation instructions, reference manual, tutorials https://www.slendr.net.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/slendr.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"A Simulation Framework for Spatiotemporal Population Genetics — slendr","text":"Maintainer: Martin Petr contact@bodkan.net (ORCID)","code":""},{"path":"https://www.slendr.net/reference/slim.html","id":null,"dir":"Reference","previous_headings":"","what":"Run a slendr model in SLiM — slim","title":"Run a slendr model in SLiM — slim","text":"function execute SLiM script generated compile function compilation slendr demographic model.","code":""},{"path":"https://www.slendr.net/reference/slim.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Run a slendr model in SLiM — slim","text":"","code":"slim(   model,   sequence_length,   recombination_rate,   samples = NULL,   output = NULL,   burnin = 0,   max_attempts = 1,   spatial = !is.null(model$world),   coalescent_only = TRUE,   method = c(\"batch\", \"gui\"),   random_seed = NULL,   run = TRUE,   verbose = FALSE,   load = TRUE,   locations = NULL,   slim_path = NULL )"},{"path":"https://www.slendr.net/reference/slim.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Run a slendr model in SLiM — slim","text":"model Model object created compile function sequence_length Total length simulated sequence (base-pairs) recombination_rate Recombination rate simulated sequence (recombinations per basepair per generation) samples data frame times given number individuals remembered tree-sequence (see schedule_sampling function can generate sampling schedule correct format). missing, individuals present end simulation recorded tree-sequence output file. output Path output tree-sequence file. NULL (default), tree sequence saved temporary file. burnin Length burnin (model's time units, .e. years) max_attempts many attempts made place offspring near one parents? Serves prevent infinite loops SLiM backend. Default value 1. spatial model executed spatial mode? default, world map specified model definition, simulation proceed spatial mode. coalescent_only initializeTreeSeq(retainCoalescentOnly =   <...>) set TRUE (default) FALSE? See \"retainCoalescentOnly\" SLiM manual detail. method run script? (\"gui\" - open SLiMgui, \"batch\" - run command line) random_seed Random seed (NULL, seed generated 0 maximum integer number available) run SLiM engine run? FALSE, command line SLiM command printed (returned invisibly character vector) executed. verbose Write SLiM output log console (default FALSE)? load final tree sequence immediately loaded returned? Default TRUE. alternative (FALSE) useful tree-sequence file written custom location loaded later point. locations NULL, locations saved. Otherwise, path file locations individual throughout simulation saved (likely use animate_model). slim_path Path appropriate SLiM binary (useful slim binary $PATH). Note argument must specified function run Windows.","code":""},{"path":"https://www.slendr.net/reference/slim.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Run a slendr model in SLiM — slim","text":"tree-sequence object loaded via Python-R reticulate interface function ts_load (internally represented Python object tskit.trees.TreeSequence). Optionally, depending value arguments load = run =, nothing character vector, respectively.","code":""},{"path":"https://www.slendr.net/reference/slim.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Run a slendr model in SLiM — slim","text":"","code":"check_dependencies(python = TRUE, slim = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # afr and eur objects would normally be created before slendr model compilation, # but here we take them out of the model object already compiled for this # example (in a standard slendr simulation pipeline, this wouldn't be necessary) afr <- model$populations[[\"AFR\"]] eur <- model$populations[[\"EUR\"]] chimp <- model$populations[[\"CH\"]]  # schedule the sampling of a couple of ancient and present-day individuals # given model at 20 ky, 10 ky, 5ky ago and at present-day (time 0) modern_samples <- schedule_sampling(model, times = 0, list(afr, 5), list(eur, 5), list(chimp, 1)) ancient_samples <- schedule_sampling(model, times = c(30000, 20000, 10000), list(eur, 1))  # sampling schedules are just data frames and can be merged easily samples <- rbind(modern_samples, ancient_samples)  # run a simulation using the SLiM back end from a compiled slendr model object and return # a tree-sequence output ts <- slim(model, sequence_length = 1e5, recombination_rate = 0, samples = samples)  # automatic loading of a simulated output can be prevented by `load = FALSE`, which can be # useful when a custom path to a tree-sequence output is given for later downstream analyses output_file <- tempfile(fileext = \".trees\") slim(model, sequence_length = 1e5, recombination_rate = 0, samples = samples,      output = output_file, load = FALSE) # ... at a later stage: ts <- ts_load(output_file, model)  ts #> ╔═══════════════════════╗ #> ║TreeSequence           ║ #> ╠═══════════════╤═══════╣ #> ║Trees          │      1║ #> ╟───────────────┼───────╢ #> ║Sequence Length│ 100000║ #> ╟───────────────┼───────╢ #> ║Time Units     │  ticks║ #> ╟───────────────┼───────╢ #> ║Sample Nodes   │  10046║ #> ╟───────────────┼───────╢ #> ║Total Size     │2.6 MiB║ #> ╚═══════════════╧═══════╝ #> ╔═══════════╤═════╤═════════╤════════════╗ #> ║Table      │Rows │Size     │Has Metadata║ #> ╠═══════════╪═════╪═════════╪════════════╣ #> ║Edges      │18361│573.8 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Individuals│12858│  1.2 MiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Migrations │    0│  8 Bytes│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Mutations  │    0│  1.2 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Nodes      │18362│682.1 KiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Populations│    4│  2.5 KiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Provenances│    1│ 34.7 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Sites      │    0│ 16 Bytes│          No║ #> ╚═══════════╧═════╧═════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/subtract.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate the difference between two slendr objects — subtract","title":"Generate the difference between two slendr objects — subtract","text":"Generate difference two slendr objects","code":""},{"path":"https://www.slendr.net/reference/subtract.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate the difference between two slendr objects — subtract","text":"","code":"subtract(x, y, name = NULL)"},{"path":"https://www.slendr.net/reference/subtract.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate the difference between two slendr objects — subtract","text":"x Object class slendr y Object class slendr name Optional name resulting geographic region. missing, name constructed function arguments.","code":""},{"path":"https://www.slendr.net/reference/subtract.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generate the difference between two slendr objects — subtract","text":"Object class slendr_region encodes standard spatial object class sf several additional attributes (importantly corresponding slendr_map object, applicable).","code":""},{"path":"https://www.slendr.net/reference/summary.slendr_nodes.html","id":null,"dir":"Reference","previous_headings":"","what":"Summarise the contents of a ts_nodes result — summary.slendr_nodes","title":"Summarise the contents of a ts_nodes result — summary.slendr_nodes","text":"Summarise contents ts_nodes result","code":""},{"path":"https://www.slendr.net/reference/summary.slendr_nodes.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Summarise the contents of a ts_nodes result — summary.slendr_nodes","text":"","code":"# S3 method for slendr_nodes summary(object, ...)"},{"path":"https://www.slendr.net/reference/summary.slendr_nodes.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Summarise the contents of a ts_nodes result — summary.slendr_nodes","text":"object Data frame produced function ts_nodes ... Additional formal arguments summary method (unused )","code":""},{"path":"https://www.slendr.net/reference/summary.slendr_nodes.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Summarise the contents of a ts_nodes result — summary.slendr_nodes","text":"Used output terminal","code":""},{"path":"https://www.slendr.net/reference/ts_afs.html","id":null,"dir":"Reference","previous_headings":"","what":"Compute the allele frequency spectrum (AFS) — ts_afs","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"function computes AFS respect given set individuals nodes.","code":""},{"path":"https://www.slendr.net/reference/ts_afs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"","code":"ts_afs(   ts,   sample_sets = NULL,   mode = c(\"site\", \"branch\", \"node\"),   windows = NULL,   span_normalise = FALSE,   polarised = TRUE )"},{"path":"https://www.slendr.net/reference/ts_afs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set). NULL, allele frequency spectrum individuals tree sequence computed. mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) added automatically) span_normalise Argument passed tskit's allele_frequency_spectrum method polarised TRUE (default) allele frequency spectrum folded (.e. counts assume knowledge allele ancestral, derived, known simulation)","code":""},{"path":"https://www.slendr.net/reference/ts_afs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"Allele frequency spectrum values given sample set. Note contents first last elements AFS might surprise . Read links description detail tskit handles things.","code":""},{"path":"https://www.slendr.net/reference/ts_afs.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"information format result dimensions, particular interpretation first last element AFS (complete = TRUE), please see tskit manual https://tskit.dev/tskit/docs/stable/python-api.html example section dedicated AFS https://tskit.dev/tutorials/analysing_tree_sequences.html#allele-frequency-spectra.","code":""},{"path":"https://www.slendr.net/reference/ts_afs.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  samples <- ts_samples(ts) %>% .[.$pop %in% c(\"AFR\", \"EUR\"), ]  # compute AFS for the given set of individuals ts_afs(ts, sample_sets = list(samples$name)) #>  [1] 1080   73   21    8    1    4    0    0    0    0   11    0    0    0    0 #> [16]    1    0    3    5   10  935"},{"path":"https://www.slendr.net/reference/ts_ancestors.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","title":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","text":"Extract (spatio-)temporal ancestral history given nodes/individuals","code":""},{"path":"https://www.slendr.net/reference/ts_ancestors.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","text":"","code":"ts_ancestors(ts, x, verbose = FALSE, complete = TRUE)"},{"path":"https://www.slendr.net/reference/ts_ancestors.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","text":"ts Tree sequence object class slendr_ts x Either individual name integer node ID verbose Report progress ancestry path generation? complete every individual tree sequence need complete metadata recorded? TRUE, individuals/nodes complete metadata included reconstruction ancestral relationships. instance, nodes added coalescent recapitation phase included spatial information associated .","code":""},{"path":"https://www.slendr.net/reference/ts_ancestors.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","text":"table ancestral nodes given tree-sequence node way root tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_ancestors.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # find the complete ancestry information for a given individual ts_ancestors(ts, \"EUR_1\", verbose = TRUE) #> Collecting ancestors of EUR_1 [1/1]... #>  #> Generating data about spatial relationships of nodes... #> # A tibble: 222 × 12 #>    name  pop   node_id level child_id parent_id child_time parent_time child_pop #>    <chr> <fct>   <int> <fct>    <int>     <int>      <dbl>       <dbl> <fct>     #>  1 EUR_1 EUR        16 1           16        33          0         500 EUR       #>  2 EUR_1 EUR        16 2           33        43        500       21830 EUR       #>  3 EUR_1 EUR        16 2           33        47        500       28730 EUR       #>  4 EUR_1 EUR        16 2           33        49        500       30380 EUR       #>  5 EUR_1 EUR        16 2           33        51        500       35360 EUR       #>  6 EUR_1 EUR        16 2           33        54        500       39380 EUR       #>  7 EUR_1 EUR        16 2           33        57        500       41510 EUR       #>  8 EUR_1 EUR        16 2           33        58        500       44810 EUR       #>  9 EUR_1 EUR        16 2           33        59        500       46040 EUR       #> 10 EUR_1 EUR        16 2           33        62        500       51680 EUR       #> # ℹ 212 more rows #> # ℹ 3 more variables: parent_pop <fct>, left_pos <dbl>, right_pos <dbl>"},{"path":"https://www.slendr.net/reference/ts_coalesced.html","id":null,"dir":"Reference","previous_headings":"","what":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","title":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","text":"Check trees tree sequence fully coalesced","code":""},{"path":"https://www.slendr.net/reference/ts_coalesced.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","text":"","code":"ts_coalesced(ts, return_failed = FALSE)"},{"path":"https://www.slendr.net/reference/ts_coalesced.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","text":"ts Tree sequence object class slendr_ts return_failed Report back trees failed coalescence check?","code":""},{"path":"https://www.slendr.net/reference/ts_coalesced.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","text":"TRUE FALSE value return_failed = FALSE, otherwise vector (tskit Python 0-based) indices trees failed coalescence test","code":""},{"path":"https://www.slendr.net/reference/ts_coalesced.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  ts_coalesced(ts) # is the tree sequence fully coalesced? (TRUE or FALSE) #> [1] TRUE  # returns a vector of tree sequence segments which are not coalesced not_coalesced <- ts_coalesced(ts, return_failed = TRUE)"},{"path":"https://www.slendr.net/reference/ts_descendants.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract all descendants of a given tree-sequence node — ts_descendants","title":"Extract all descendants of a given tree-sequence node — ts_descendants","text":"Extract descendants given tree-sequence node","code":""},{"path":"https://www.slendr.net/reference/ts_descendants.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract all descendants of a given tree-sequence node — ts_descendants","text":"","code":"ts_descendants(ts, x, verbose = FALSE, complete = TRUE)"},{"path":"https://www.slendr.net/reference/ts_descendants.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract all descendants of a given tree-sequence node — ts_descendants","text":"ts Tree sequence object class slendr_ts x integer node ID ancestral node verbose Report progress ancestry path generation? complete every individual tree sequence need complete metadata recorded? TRUE, individuals/nodes complete metadata included reconstruction ancestral relationships. instance, nodes added coalescent recapitation phase included spatial information associated .","code":""},{"path":"https://www.slendr.net/reference/ts_descendants.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract all descendants of a given tree-sequence node — ts_descendants","text":"table descendant nodes given tree-sequence node way leaves tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_descendants.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract all descendants of a given tree-sequence node — ts_descendants","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # find the complete descendancy information for a given individual ts_descendants(ts, x = 62, verbose = TRUE) #>  #> Generating data about spatial relationships of nodes... #> # A tibble: 4 × 12 #>   name  pop   node_id level child_id parent_id child_time parent_time child_pop #>   <chr> <fct>   <dbl> <fct>    <int>     <int>      <dbl>       <dbl> <fct>     #> 1 EUR_4 EUR        62 1           22        62          0       51680 EUR       #> 2 NA    EUR        62 1           33        62        500       51680 EUR       #> 3 EUR_1 EUR        62 2           16        33          0         500 EUR       #> 4 EUR_2 EUR        62 2           18        33          0         500 EUR       #> # ℹ 3 more variables: parent_pop <fct>, left_pos <dbl>, right_pos <dbl>"},{"path":"https://www.slendr.net/reference/ts_divergence.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate pairwise divergence between sets of individuals — ts_divergence","title":"Calculate pairwise divergence between sets of individuals — ts_divergence","text":"Calculate pairwise divergence sets individuals","code":""},{"path":"https://www.slendr.net/reference/ts_divergence.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate pairwise divergence between sets of individuals — ts_divergence","text":"","code":"ts_divergence(   ts,   sample_sets,   mode = c(\"site\", \"branch\", \"node\"),   windows = NULL,   span_normalise = TRUE )"},{"path":"https://www.slendr.net/reference/ts_divergence.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate pairwise divergence between sets of individuals — ts_divergence","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set) mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) specified added automatically. span_normalise Divide result span window? Default TRUE, see tskit documentation detail.","code":""},{"path":"https://www.slendr.net/reference/ts_divergence.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate pairwise divergence between sets of individuals — ts_divergence","text":"pairwise calculation, either single divergence value vector divergence values (one window)","code":""},{"path":"https://www.slendr.net/reference/ts_divergence.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate pairwise divergence between sets of individuals — ts_divergence","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # collect sampled individuals from all populations in a list sample_sets <- ts_samples(ts) %>%   split(., .$pop) %>%   lapply(function(pop) pop$name)  # compute the divergence between individuals from each sample set (list of # individual names generated in the previous step) ts_divergence(ts, sample_sets) %>% .[order(.$divergence), ] #> # A tibble: 6 × 3 #>   x     y     divergence #>   <chr> <chr>      <dbl> #> 1 AFR   EUR    0.0000602 #> 2 EUR   NEA    0.000289  #> 3 AFR   NEA    0.000305  #> 4 CH    NEA    0.00397   #> 5 AFR   CH     0.00398   #> 6 CH    EUR    0.00398"},{"path":"https://www.slendr.net/reference/ts_diversity.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate diversity in given sets of individuals — ts_diversity","title":"Calculate diversity in given sets of individuals — ts_diversity","text":"Calculate diversity given sets individuals","code":""},{"path":"https://www.slendr.net/reference/ts_diversity.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate diversity in given sets of individuals — ts_diversity","text":"","code":"ts_diversity(   ts,   sample_sets,   mode = c(\"site\", \"branch\", \"node\"),   windows = NULL,   span_normalise = TRUE )"},{"path":"https://www.slendr.net/reference/ts_diversity.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate diversity in given sets of individuals — ts_diversity","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set). simple vector provided, interpreted .list(sample_sets), meaning given statistic calculated individual separately. mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) added automatically) span_normalise Divide result span window? Default TRUE, see tskit documentation detail.","code":""},{"path":"https://www.slendr.net/reference/ts_diversity.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate diversity in given sets of individuals — ts_diversity","text":"set individuals either single diversity value vector diversity values (one window)","code":""},{"path":"https://www.slendr.net/reference/ts_diversity.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate diversity in given sets of individuals — ts_diversity","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # collect sampled individuals from all populations in a list sample_sets <- ts_samples(ts) %>%   split(., .$pop) %>%   lapply(function(pop) pop$name)  # compute diversity in each population based on sample sets extracted # in the previous step ts_diversity(ts, sample_sets) %>% .[order(.$diversity), ] #> # A tibble: 4 × 2 #>   set    diversity #>   <chr>      <dbl> #> 1 CH    0          #> 2 AFR   0.0000004  #> 3 NEA   0.00000933 #> 4 EUR   0.0000661"},{"path":"https://www.slendr.net/reference/ts_draw.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot a graphical representation of a single tree — ts_draw","title":"Plot a graphical representation of a single tree — ts_draw","text":"function first obtains SVG representation tree calling draw_svg method tskit renders bitmap image R. many optional keyword arguments draw_svg method can provided automatically passed method behind scenes.","code":""},{"path":"https://www.slendr.net/reference/ts_draw.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot a graphical representation of a single tree — ts_draw","text":"","code":"ts_draw(   x,   width = 1000,   height = 1000,   labels = FALSE,   sampled_only = TRUE,   title = NULL,   ... )"},{"path":"https://www.slendr.net/reference/ts_draw.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot a graphical representation of a single tree — ts_draw","text":"x single tree extracted ts_tree width, height Pixel dimensions rendered bitmap labels Label node individual name? sampled_only individuals explicitly sampled simplification labeled? relevant situations sampled individuals can among ancestral nodes. title Optional title figure ... Keyword arguments tskit draw_svg function.","code":""},{"path":"https://www.slendr.net/reference/ts_draw.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot a graphical representation of a single tree — ts_draw","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/ts_draw.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot a graphical representation of a single tree — ts_draw","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract the first tree in the tree sequence and draw it tree <- ts_tree(ts, i = 1)  # ts_draw accepts various optional arguments of tskit.Tree.draw_svg ts_draw(tree, time_scale = \"rank\")"},{"path":"https://www.slendr.net/reference/ts_edges.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","title":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","text":"Extract spatio-temporal edge annotation table given tree tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_edges.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","text":"","code":"ts_edges(x)"},{"path":"https://www.slendr.net/reference/ts_edges.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","text":"x Tree object generated ts_phylo slendr tree sequence object produced ts_load, ts_recapitate, ts_simplify, ts_mutate","code":""},{"path":"https://www.slendr.net/reference/ts_edges.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","text":"Data frame sf type containing times nodes start-end coordinates edges across space","code":""},{"path":"https://www.slendr.net/reference/ts_edges.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract an annotated table with (spatio-)temporal edge information ts_edges(ts) #> # A tibble: 298 × 6 #>    child_node_id parent_node_id child_time parent_time child_pop parent_pop #>            <int>          <int>      <dbl>       <dbl> <fct>     <fct>      #>  1             0             74      70000       70040 NEA       NEA        #>  2             1             74      70000       70040 NEA       NEA        #>  3             2             55      40000       40220 NEA       NEA        #>  4             3             55      40000       40220 NEA       NEA        #>  5             4             34          0         740 AFR       AFR        #>  6             5             28          0          80 AFR       AFR        #>  7             6             26          0          20 AFR       AFR        #>  8             7             27          0          80 AFR       AFR        #>  9             8             27          0          80 AFR       AFR        #> 10             9             36          0        1160 AFR       AFR        #> # ℹ 288 more rows"},{"path":"https://www.slendr.net/reference/ts_eigenstrat.html","id":null,"dir":"Reference","previous_headings":"","what":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","title":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","text":"EIGENSTRAT data produced function can used admixr R package (https://bodkan.net/admixr/).","code":""},{"path":"https://www.slendr.net/reference/ts_eigenstrat.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","text":"","code":"ts_eigenstrat(ts, prefix, chrom = \"chr1\", outgroup = NULL)"},{"path":"https://www.slendr.net/reference/ts_eigenstrat.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","text":"ts Tree sequence object class slendr_ts prefix EIGENSTRAT trio prefix chrom name chromosome EIGENSTRAT snp file (default \"chr1\") outgroup formal, artificial outgroup added? NULL (default), outgroup added. non-NULL character name serve name outgroup ind file.","code":""},{"path":"https://www.slendr.net/reference/ts_eigenstrat.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","text":"Object class EIGENSTRAT created admixr package","code":""},{"path":"https://www.slendr.net/reference/ts_eigenstrat.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","text":"case outgroup formally specified slendr model generated tree sequence data, possible artificially create outgroup sample name specified outgroup argument, carry ancestral alleles (.e. value \"2\" geno file position snp file).","code":""},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"functions present R interface corresponding f-statistics methods tskit.","code":""},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"","code":"ts_f2(   ts,   A,   B,   mode = c(\"site\", \"branch\", \"node\"),   span_normalise = TRUE,   windows = NULL )  ts_f3(   ts,   A,   B,   C,   mode = c(\"site\", \"branch\", \"node\"),   span_normalise = TRUE,   windows = NULL )  ts_f4(   ts,   W,   X,   Y,   Z,   mode = c(\"site\", \"branch\", \"node\"),   span_normalise = TRUE,   windows = NULL )  ts_f4ratio(   ts,   X,   A,   B,   C,   O,   mode = c(\"site\", \"branch\"),   span_normalise = TRUE )"},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"ts Tree sequence object class slendr_ts mode mode calculation (\"sites\" \"branch\") span_normalise Divide result span window? Default TRUE, see tskit documentation detail. windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) specified added automatically. W, X, Y, Z, , B, C, O Character vectors individual names (largely following nomenclature Patterson 2021, see crucial differences tskit ADMIXTOOLS Details)","code":""},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"Data frame statistics calculated given sets individuals","code":""},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"Note order populations f3 statistic implemented tskit (https://tskit.dev/tskit/docs/stable/python-api.html#tskit.TreeSequence.f3) different might expect ADMIXTOOLS, defined Patterson 2012 (see https://academic.oup.com/genetics/article/192/3/1065/5935193 heading \"three-population test introduction f-statistics\", well ADMIXTOOLS documentation https://github.com/DReichLab/AdmixTools/blob/master/README.3PopTest#L5). Specifically, widely used notation introduced Patterson assumes population triplet f3(C; , B), C \"focal\" sample (.e., either outgroup sample tested admixture). contrast, tskit implements f3(; B, C), \"focal sample\" . Although likely confuse many ADMIXTOOLS users, slendr much choice , ts_*() functions designed broadly compatible raw tskit methods.","code":""},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk and add mutations to it ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # calculate f2 for two individuals in a previously loaded tree sequence ts_f2(ts, A = \"AFR_1\", B = \"EUR_1\") #> # A tibble: 1 × 3 #>   A     B           f2 #>   <chr> <chr>    <dbl> #> 1 AFR_1 EUR_1 0.000024  # calculate f2 for two sets of individuals ts_f2(ts, A = c(\"AFR_1\", \"AFR_2\"), B = c(\"EUR_1\", \"EUR_3\")) #> # A tibble: 1 × 3 #>   A           B                  f2 #>   <chr>       <chr>           <dbl> #> 1 AFR_1+AFR_2 EUR_1+EUR_3 0.0000247  # calculate f3 for two individuals in a previously loaded tree sequence ts_f3(ts, A = \"EUR_1\", B = \"AFR_1\", C = \"NEA_1\") #> # A tibble: 1 × 4 #>   A     B     C            f3 #>   <chr> <chr> <chr>     <dbl> #> 1 EUR_1 AFR_1 NEA_1 -0.000012  # calculate f3 for two sets of individuals ts_f3(ts, A = c(\"AFR_1\", \"AFR_2\", \"EUR_1\", \"EUR_2\"),           B = c(\"NEA_1\", \"NEA_2\"),           C = \"CH_1\") #> # A tibble: 1 × 4 #>   A                       B           C           f3 #>   <chr>                   <chr>       <chr>    <dbl> #> 1 AFR_1+AFR_2+EUR_1+EUR_2 NEA_1+NEA_2 CH_1  0.000126  # calculate f4 for single individuals ts_f4(ts, W = \"EUR_1\", X = \"AFR_1\", Y = \"NEA_1\", Z = \"CH_1\") #> # A tibble: 1 × 5 #>   W     X     Y     Z           f4 #>   <chr> <chr> <chr> <chr>    <dbl> #> 1 EUR_1 AFR_1 NEA_1 CH_1  0.000009  # calculate f4 for sets of individuals ts_f4(ts, W = c(\"EUR_1\", \"EUR_2\"),           X = c(\"AFR_1\", \"AFR_2\"),           Y = \"NEA_1\",           Z = \"CH_1\") #> # A tibble: 1 × 5 #>   W           X           Y     Z           f4 #>   <chr>       <chr>       <chr> <chr>    <dbl> #> 1 EUR_1+EUR_2 AFR_1+AFR_2 NEA_1 CH_1  0.000018  # calculate f4-ratio for a given set of target individuals X ts_f4ratio(ts, X = c(\"EUR_1\", \"EUR_2\", \"EUR_4\", \"EUR_5\"),                A = \"NEA_1\", B = \"NEA_2\", C = \"AFR_1\", O = \"CH_1\") #> # A tibble: 4 × 6 #>   X     A     B     C     O      alpha #>   <chr> <chr> <chr> <chr> <chr>  <dbl> #> 1 EUR_1 NEA_1 NEA_2 AFR_1 CH_1  0.0486 #> 2 EUR_2 NEA_1 NEA_2 AFR_1 CH_1  0.181  #> 3 EUR_4 NEA_1 NEA_2 AFR_1 CH_1  0.0417 #> 4 EUR_5 NEA_1 NEA_2 AFR_1 CH_1  0"},{"path":"https://www.slendr.net/reference/ts_fst.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate pairwise statistics between sets of individuals — ts_fst","title":"Calculate pairwise statistics between sets of individuals — ts_fst","text":"discussion difference \"site\", \"branch\", \"node\" options mode argument, please see tskit documentation https://tskit.dev/tskit/docs/stable/stats.html#sec-stats-mode.","code":""},{"path":"https://www.slendr.net/reference/ts_fst.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate pairwise statistics between sets of individuals — ts_fst","text":"","code":"ts_fst(   ts,   sample_sets,   mode = c(\"site\", \"branch\", \"node\"),   windows = NULL,   span_normalise = TRUE )"},{"path":"https://www.slendr.net/reference/ts_fst.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate pairwise statistics between sets of individuals — ts_fst","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set) mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) specified added automatically. span_normalise Divide result span window? Default TRUE, see tskit documentation detail.","code":""},{"path":"https://www.slendr.net/reference/ts_fst.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate pairwise statistics between sets of individuals — ts_fst","text":"pairwise calculation, either single Fst value vector Fst values (one window)","code":""},{"path":"https://www.slendr.net/reference/ts_fst.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate pairwise statistics between sets of individuals — ts_fst","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # compute F_st between two sets of individuals in a given tree sequence ts ts_fst(ts, sample_sets = list(afr = c(\"AFR_1\", \"AFR_2\", \"AFR_3\"),                               eur = c(\"EUR_1\", \"EUR_2\"))) #> # A tibble: 1 × 3 #>   x     y       Fst #>   <chr> <chr> <dbl> #> 1 afr   eur   0.303"},{"path":"https://www.slendr.net/reference/ts_genotypes.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract genotype table from the tree sequence — ts_genotypes","title":"Extract genotype table from the tree sequence — ts_genotypes","text":"Extract genotype table tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_genotypes.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract genotype table from the tree sequence — ts_genotypes","text":"","code":"ts_genotypes(ts)"},{"path":"https://www.slendr.net/reference/ts_genotypes.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract genotype table from the tree sequence — ts_genotypes","text":"ts Tree sequence object class slendr_ts","code":""},{"path":"https://www.slendr.net/reference/ts_genotypes.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract genotype table from the tree sequence — ts_genotypes","text":"Data frame object class tibble containing genotypes simulated individuals columns","code":""},{"path":"https://www.slendr.net/reference/ts_genotypes.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract genotype table from the tree sequence — ts_genotypes","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk, recapitate it, simplify it, and mutate it ts <- ts_load(slendr_ts, model) %>%   ts_recapitate(Ne = 10000, recombination_rate = 1e-8) %>%   ts_simplify() %>%   ts_mutate(mutation_rate = 1e-8)  # extract the genotype matrix (this could take  a long time consume lots # of memory!) gts <- ts_genotypes(ts) #> 3 multiallelic sites (0.138% out of 2173 total) detected and removed"},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":null,"dir":"Reference","previous_headings":"","what":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"function iterates tree sequence returns IBD tracts pairs individuals nodes","code":""},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"","code":"ts_ibd(   ts,   coordinates = FALSE,   within = NULL,   between = NULL,   squash = FALSE,   minimum_length = NULL,   maximum_time = NULL,   sf = TRUE )"},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"ts Tree sequence object class slendr_ts coordinates coordinates detected IBD tracts reported? FALSE (default), total length shared IBD segments numbers reported. TRUE, coordinates segment returned (note can massive impact memory usage). See details information. within character vector individual names integer vector node IDs indicating set nodes within look IBD segments. list lists character vectors individual names integer vectors node IDs, indicating set nodes look shared IBD segments. squash adjacent IBD segments pairs nodes squashed differ 'genealogical paths' MRCA? Default FALSE. context, see https://github.com/tskit-dev/tskit/issues/2459. option EXPERIMENTAL! minimum_length Minimum length IBD segment return results. useful reducing total amount IBD returned (see Details). maximum_time Oldest MRCA node considered IBD ancestor return IBD segment results. useful reducing total amount IBD returned. sf IBD segments spatial tree sequence analyzed, returned table spatial sf object? Default TRUE.","code":""},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"data frame IBD results (either coordinates IBD segment shared pair nodes, summary statistics total IBD sharing pair)","code":""},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"function considered experimental. full control IBD segment detection tree-sequence data, users can (perhaps, time , ) rely tskit method ibd_segments (see https://tskit.dev/tskit/docs/stable/python-api.html#tskit.TreeSequence.ibd_segments). Iternally, function leverages tskit TreeSequence method ibd_segments. However, note ts_ibd function always returns data frame IBD tracts, provide option iterate individual IBD segments shown official tskit documentation https://tskit.dev/tskit/docs/stable/ibd.html. general, R handles heavy iteration poorly, function attempt serve full wrapper ibd_segments. Unfortunately, distinction \"squashed IBD\" (many consider expected definition IBD) tskit’s IBD defined via distinct genealogical paths (see https://github.com/tskit-dev/tskit/issues/2459 discussion topic), makes meaning filtering parameter ibd_segments() method tskit minimum_length somewhat unintuitive. moment, function argument filters IBD segments tskit level, level squashed IBD segments!","code":""},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # find IBD segments between specified Neanderthals and Europeans ts_ibd(   ts,   coordinates = TRUE,   between = list(c(\"NEA_1\", \"NEA_2\"), c(\"EUR_1\", \"EUR_2\")),   minimum_length = 40000 ) #> # A tibble: 34 × 13 #>    node1 node2 length  mrca node1_time node2_time tmrca   left  right name1 #>    <int> <int>  <dbl> <dbl>      <dbl>      <dbl> <dbl>  <dbl>  <dbl> <chr> #>  1     0    16  52762    83      70000          0 20001 228046 280808 NEA_1 #>  2     0    16  45006    83      70000          0 20001 449167 494173 NEA_1 #>  3     0    17  52762    83      70000          0 20001 228046 280808 NEA_1 #>  4     0    18  52762    83      70000          0 20001 228046 280808 NEA_1 #>  5     0    18  45006    83      70000          0 20001 449167 494173 NEA_1 #>  6     0    19  57765    83      70000          0 20001  47447 105212 NEA_1 #>  7     0    19  47935    83      70000          0 20001 105212 153147 NEA_1 #>  8     0    19  46028    83      70000          0 20001 178539 224567 NEA_1 #>  9     0    19  58054    79      70000          0  2340 227185 285239 NEA_1 #> 10     1    16  52762    83      70000          0 20001 228046 280808 NEA_1 #> # ℹ 24 more rows #> # ℹ 3 more variables: name2 <chr>, pop1 <fct>, pop2 <fct>"},{"path":"https://www.slendr.net/reference/ts_load.html","id":null,"dir":"Reference","previous_headings":"","what":"Load a tree sequence file produced by a given model — ts_load","title":"Load a tree sequence file produced by a given model — ts_load","text":"function loads tree sequence file simulated given slendr model. Optionally, tree sequence can recapitated simplified.","code":""},{"path":"https://www.slendr.net/reference/ts_load.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Load a tree sequence file produced by a given model — ts_load","text":"","code":"ts_load(file, model = NULL)"},{"path":"https://www.slendr.net/reference/ts_load.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Load a tree sequence file produced by a given model — ts_load","text":"file path tree-sequence file (either originating slendr model standard non-slendr tree sequence). model Optional slendr_model object produced tree-sequence file. Used adding various annotation data metadata standard tskit tree-sequence object.","code":""},{"path":"https://www.slendr.net/reference/ts_load.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Load a tree sequence file produced by a given model — ts_load","text":"Tree-sequence object class slendr_ts, serves interface point Python module tskit using slendr functions ts_ prefix.","code":""},{"path":"https://www.slendr.net/reference/ts_load.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Load a tree sequence file produced by a given model — ts_load","text":"loading, recapitation simplification performed using Python module pyslim serves link tree sequences generated SLiM tskit module manipulation tree sequence data. steps modelled official pyslim tutorial documentation available : https://tskit.dev/pyslim/docs/latest/tutorial.html. recapitation simplification steps can also performed individually using functions ts_recapitate ts_simplify.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_load.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Load a tree sequence file produced by a given model — ts_load","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load tree sequence generated by a given model ts <- ts_load(slendr_ts, model)  # even tree sequences generated by non-slendr models can be msprime_ts <- system.file(\"extdata/models/msprime.trees\", package = \"slendr\") ts <- ts_load(msprime_ts)  # load tree sequence and immediately simplify it only to sampled individuals # (note that the example tree sequence is already simplified so this operation # does not do anything in this case) ts <- ts_load(slendr_ts, model = model) %>% ts_simplify(keep_input_roots = TRUE)  # load tree sequence and simplify it to a subset of sampled individuals ts_small <- ts_simplify(ts, simplify_to = c(\"CH_1\", \"NEA_1\", \"NEA_2\",                                             \"AFR_1\", \"AFR_2\", \"EUR_1\", \"EUR_2\"))  # load tree sequence, recapitate it and simplify it ts <- ts_load(slendr_ts, model) %>%   ts_recapitate(recombination_rate = 1e-8, Ne = 10000, random_seed = 42) %>%   ts_simplify()  # load tree sequence, recapitate it, simplify it and overlay neutral mutations ts <- ts_load(slendr_ts, model) %>%   ts_recapitate(recombination_rate = 1e-8, Ne = 10000, random_seed = 42) %>%   ts_simplify() %>%   ts_mutate(mutation_rate = 1e-8)  ts #> ╔═════════════════════════╗ #> ║TreeSequence             ║ #> ╠═══════════════╤═════════╣ #> ║Trees          │       70║ #> ╟───────────────┼─────────╢ #> ║Sequence Length│   500000║ #> ╟───────────────┼─────────╢ #> ║Time Units     │    ticks║ #> ╟───────────────┼─────────╢ #> ║Sample Nodes   │       26║ #> ╟───────────────┼─────────╢ #> ║Total Size     │205.3 KiB║ #> ╚═══════════════╧═════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │ 297│ 9.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│  72│ 8.8 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │2206│80.9 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │  85│ 3.8 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   6│ 2.7 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   6│41.0 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │2201│53.8 KiB│          No║ #> ╚═══════════╧════╧════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/ts_metadata.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","title":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","text":"Extract list tree sequence metadata saved SLiM","code":""},{"path":"https://www.slendr.net/reference/ts_metadata.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","text":"","code":"ts_metadata(ts)"},{"path":"https://www.slendr.net/reference/ts_metadata.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","text":"ts Tree sequence object class slendr_ts","code":""},{"path":"https://www.slendr.net/reference/ts_metadata.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","text":"List metadata fields extracted tree-sequence object","code":""},{"path":"https://www.slendr.net/reference/ts_metadata.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract the list of metadata information from the tree sequence ts_metadata(ts) #> $version #> [1] \"slendr_0.8.1.9000\" #>  #> $description #> [1] \"\" #>  #> $sampling #> # A tibble: 13 × 3 #>    name   time pop   #>    <chr> <int> <chr> #>  1 NEA_1 70000 NEA   #>  2 NEA_2 40000 NEA   #>  3 AFR_1     0 AFR   #>  4 AFR_2     0 AFR   #>  5 AFR_3     0 AFR   #>  6 AFR_4     0 AFR   #>  7 AFR_5     0 AFR   #>  8 CH_1      0 CH    #>  9 EUR_1     0 EUR   #> 10 EUR_2     0 EUR   #> 11 EUR_3     0 EUR   #> 12 EUR_4     0 EUR   #> 13 EUR_5     0 EUR   #>  #> $sample_names #>  [1] \"NEA_1\" \"NEA_2\" \"AFR_1\" \"AFR_2\" \"AFR_3\" \"AFR_4\" \"AFR_5\" \"CH_1\"  \"EUR_1\" #> [10] \"EUR_2\" \"EUR_3\" \"EUR_4\" \"EUR_5\" #>  #> $sample_ids #> NULL #>  #> $map #> NULL #>  #> $arguments #> $arguments$BURNIN_LENGTH #> [1] 0 #>  #> $arguments$MAX_ATTEMPTS #> [1] 1 #>  #> $arguments$RECOMB_RATE #> [1] 1e-08 #>  #> $arguments$SEED #> [1] 314159 #>  #> $arguments$SEQUENCE_LENGTH #> [1] 500000 #>  #> $arguments$SIMULATION_LENGTH #> [1] 216667 #>  #>"},{"path":"https://www.slendr.net/reference/ts_mutate.html","id":null,"dir":"Reference","previous_headings":"","what":"Add mutations to the given tree sequence — ts_mutate","title":"Add mutations to the given tree sequence — ts_mutate","text":"Add mutations given tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_mutate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add mutations to the given tree sequence — ts_mutate","text":"","code":"ts_mutate(   ts,   mutation_rate,   random_seed = NULL,   keep_existing = TRUE,   mut_type = NULL )"},{"path":"https://www.slendr.net/reference/ts_mutate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add mutations to the given tree sequence — ts_mutate","text":"ts Tree sequence object class slendr_ts mutation_rate Mutation rate used msprime simulate mutations random_seed Random seed passed msprime's mutate method (NULL, seed generated 0 maximum integer number available) keep_existing Keep existing mutations? mut_type Assign SLiM mutation type neutral mutations? NULL (default), special mutation type used. integer number given, mutations SLiM mutation type integer identifier created.","code":""},{"path":"https://www.slendr.net/reference/ts_mutate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add mutations to the given tree sequence — ts_mutate","text":"Tree-sequence object class slendr_ts, serves interface point Python module tskit using slendr functions ts_ prefix.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_mutate.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add mutations to the given tree sequence — ts_mutate","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  ts <- ts_load(slendr_ts, model) ts_mutate <- ts_mutate(ts, mutation_rate = 1e-8, random_seed = 42)  ts_mutate #> ╔═════════════════════════╗ #> ║TreeSequence             ║ #> ╠═══════════════╤═════════╣ #> ║Trees          │       70║ #> ╟───────────────┼─────────╢ #> ║Sequence Length│   500000║ #> ╟───────────────┼─────────╢ #> ║Time Units     │    ticks║ #> ╟───────────────┼─────────╢ #> ║Sample Nodes   │       26║ #> ╟───────────────┼─────────╢ #> ║Total Size     │204.8 KiB║ #> ╚═══════════════╧═════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │ 298│ 9.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│  73│ 8.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │2246│82.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │  86│ 3.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   5│ 2.6 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   4│38.0 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │2240│54.7 KiB│          No║ #> ╚═══════════╧════╧════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/ts_names.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract names of individuals in a tree sequence — ts_names","title":"Extract names of individuals in a tree sequence — ts_names","text":"Extract names individuals tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_names.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract names of individuals in a tree sequence — ts_names","text":"","code":"ts_names(ts, split = NULL)"},{"path":"https://www.slendr.net/reference/ts_names.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract names of individuals in a tree sequence — ts_names","text":"ts Tree sequence object class slendr_ts split sample names tree sequence split column (population time column)? Default NULL names samples returned single character vector. set \"pop\" \"time\", list character vectors returned, one vector unique \"pop\" \"time\" grouping.","code":""},{"path":"https://www.slendr.net/reference/ts_names.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract names of individuals in a tree sequence — ts_names","text":"vector character sample names. split specified, list vectors returned, one element list per population sampling time.","code":""},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract combined annotated table of individuals and nodes — ts_nodes","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"function combines information table individuals table nodes single data frame can used downstream analyses.","code":""},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"","code":"ts_nodes(x, sf = TRUE)"},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"x Tree sequence object class slendr_ts phylo object extracted ts_phylo sf spatial data returned sf format? FALSE, spatial geometries returned simply x y columns, instead standard POINT data type.","code":""},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"Data frame processed information tree sequence object. model generated data spatial, result returned spatial object class sf.","code":""},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"source data (tables individuals nodes recorded tree sequence generated SLiM) combined single data frame. model generated data spatial, coordinates nodes (pixel-based default SLiM spatial simulations occur raster), coordinates automatically converted explicit spatial object sf class unless spatial = FALSE. See https://r-spatial.github.io/sf/ extensive introduction sf package ways spatial data can processed, analysed, visualised.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract an annotated table with (spatio-)temporal node information ts_nodes(ts) #> # A tibble: 86 × 12 #>    name  pop   node_id  time time_tskit sampled remembered retained alive #>    <chr> <fct>   <int> <dbl>      <dbl> <lgl>   <lgl>      <lgl>    <lgl> #>  1 NEA_1 NEA         0 70000       2334 TRUE    TRUE       TRUE     FALSE #>  2 NEA_1 NEA         1 70000       2334 TRUE    TRUE       TRUE     FALSE #>  3 NEA_2 NEA         2 40000       1334 TRUE    TRUE       TRUE     FALSE #>  4 NEA_2 NEA         3 40000       1334 TRUE    TRUE       TRUE     FALSE #>  5 AFR_1 AFR         4     0          0 TRUE    TRUE       TRUE     TRUE  #>  6 AFR_1 AFR         5     0          0 TRUE    TRUE       TRUE     TRUE  #>  7 AFR_2 AFR         6     0          0 TRUE    TRUE       TRUE     TRUE  #>  8 AFR_2 AFR         7     0          0 TRUE    TRUE       TRUE     TRUE  #>  9 AFR_3 AFR         8     0          0 TRUE    TRUE       TRUE     TRUE  #> 10 AFR_3 AFR         9     0          0 TRUE    TRUE       TRUE     TRUE  #> # ℹ 76 more rows #> # ℹ 3 more variables: pedigree_id <dbl>, pop_id <int>, ind_id <dbl>"},{"path":"https://www.slendr.net/reference/ts_phylo.html","id":null,"dir":"Reference","previous_headings":"","what":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","title":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","text":"Convert tree tree sequence object class phylo","code":""},{"path":"https://www.slendr.net/reference/ts_phylo.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","text":"","code":"ts_phylo(   ts,   i,   mode = c(\"index\", \"position\"),   labels = c(\"tskit\", \"pop\"),   quiet = FALSE )"},{"path":"https://www.slendr.net/reference/ts_phylo.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","text":"ts Tree sequence object class slendr_ts Position tree tree sequence. mode = \"index\", -th tree returned (zero-based indexing tskit), mode = \"position\", tree covering -th base simulated genome returned (, tskit's indexing). mode argument interpreted? Either \"index\" -th tree sequence genealogies, \"position\" along simulated genome. labels stored node labels final phylo object? Options either population name tskit integer node ID (different thing phylo class node integer index). quiet ape's internal phylo validity test printed ?","code":""},{"path":"https://www.slendr.net/reference/ts_phylo.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","text":"Standard phylogenetic tree object implemented R package ape","code":""},{"path":"https://www.slendr.net/reference/ts_phylo.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>%   ts_recapitate(Ne = 10000, recombination_rate = 1e-8) %>%   ts_simplify()  # extract the 1st tree from a given tree sequence, return ape object tree <- ts_phylo(ts, i = 1, mode = \"index\", quiet = TRUE) tree #>  #> Phylogenetic tree with 26 tips and 24 internal nodes. #>  #> Tip labels: #>   25 (EUR_5), 24 (EUR_5), 23 (EUR_4), 22 (EUR_4), 21 (EUR_3), 20 (EUR_3), ... #> Node labels: #>   84, 28, 27, 30, 31, 32, ... #>  #> Rooted; includes branch lengths.  # extract the tree at a 42th basepair in the given tree sequence tree <- ts_phylo(ts, i = 42, mode = \"position\", quiet = TRUE)  # because the tree is a standard ape phylo object, we can plot it easily plot(tree, use.edge.length = FALSE) ape::nodelabels()"},{"path":"https://www.slendr.net/reference/ts_recapitate.html","id":null,"dir":"Reference","previous_headings":"","what":"Recapitate the tree sequence — ts_recapitate","title":"Recapitate the tree sequence — ts_recapitate","text":"Recapitate tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_recapitate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Recapitate the tree sequence — ts_recapitate","text":"","code":"ts_recapitate(   ts,   recombination_rate,   Ne = NULL,   demography = NULL,   random_seed = NULL )"},{"path":"https://www.slendr.net/reference/ts_recapitate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Recapitate the tree sequence — ts_recapitate","text":"ts Tree sequence object loaded ts_load recombination_rate constant value recombination rate Ne Effective population size recapitation process demography Ancestral demography passed internally msprime.sim_ancestry() (see msprime's documentation mode detail) random_seed Random seed passed pyslim's recapitate method (NULL, seed generated 0 maximum integer number available)","code":""},{"path":"https://www.slendr.net/reference/ts_recapitate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Recapitate the tree sequence — ts_recapitate","text":"Tree-sequence object class slendr_ts, serves interface point Python module tskit using slendr functions ts_ prefix.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_recapitate.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Recapitate the tree sequence — ts_recapitate","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  ts <- ts_load(slendr_ts, model) %>%   ts_recapitate(recombination_rate = 1e-8, Ne = 10000, random_seed = 42)  ts #> ╔════════════════════════╗ #> ║TreeSequence            ║ #> ╠═══════════════╤════════╣ #> ║Trees          │      70║ #> ╟───────────────┼────────╢ #> ║Sequence Length│  500000║ #> ╟───────────────┼────────╢ #> ║Time Units     │   ticks║ #> ╟───────────────┼────────╢ #> ║Sample Nodes   │      26║ #> ╟───────────────┼────────╢ #> ║Total Size     │70.8 KiB║ #> ╚═══════════════╧════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │ 298│ 9.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│  73│ 8.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │   0│ 1.2 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │  86│ 3.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   6│ 2.7 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   4│39.8 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │   0│16 Bytes│          No║ #> ╚═══════════╧════╧════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/ts_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","title":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","text":"Extract names times individuals interest current tree sequence (either sampled individuals user simplified )","code":""},{"path":"https://www.slendr.net/reference/ts_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","text":"","code":"ts_samples(ts)"},{"path":"https://www.slendr.net/reference/ts_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","text":"ts Tree sequence object class slendr_ts","code":""},{"path":"https://www.slendr.net/reference/ts_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","text":"Table individuals scheduled sampling across space time","code":""},{"path":"https://www.slendr.net/reference/ts_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract the table of individuals scheduled for simulation and sampling ts_samples(ts) #> # A tibble: 13 × 3 #>    name   time pop   #>    <chr> <int> <chr> #>  1 NEA_1 70000 NEA   #>  2 NEA_2 40000 NEA   #>  3 AFR_1     0 AFR   #>  4 AFR_2     0 AFR   #>  5 AFR_3     0 AFR   #>  6 AFR_4     0 AFR   #>  7 AFR_5     0 AFR   #>  8 CH_1      0 CH    #>  9 EUR_1     0 EUR   #> 10 EUR_2     0 EUR   #> 11 EUR_3     0 EUR   #> 12 EUR_4     0 EUR   #> 13 EUR_5     0 EUR"},{"path":"https://www.slendr.net/reference/ts_save.html","id":null,"dir":"Reference","previous_headings":"","what":"Save a tree sequence to a file — ts_save","title":"Save a tree sequence to a file — ts_save","text":"Save tree sequence file","code":""},{"path":"https://www.slendr.net/reference/ts_save.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Save a tree sequence to a file — ts_save","text":"","code":"ts_save(ts, file)"},{"path":"https://www.slendr.net/reference/ts_save.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Save a tree sequence to a file — ts_save","text":"ts Tree sequence object loaded ts_load file File tree sequence saved","code":""},{"path":"https://www.slendr.net/reference/ts_save.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Save a tree sequence to a file — ts_save","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/ts_save.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Save a tree sequence to a file — ts_save","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree sequence ts <- ts_load(slendr_ts, model)  # save the tree-sequence object to a different location another_file <- paste(tempfile(), \".trees\") ts_save(ts, another_file)"},{"path":"https://www.slendr.net/reference/ts_segregating.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","title":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","text":"Calculate density segregating sites given sets individuals","code":""},{"path":"https://www.slendr.net/reference/ts_segregating.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","text":"","code":"ts_segregating(   ts,   sample_sets,   mode = c(\"site\", \"branch\", \"node\"),   windows = NULL,   span_normalise = FALSE )"},{"path":"https://www.slendr.net/reference/ts_segregating.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set). simple vector provided, interpreted .list(sample_sets), meaning given statistic calculated individual separately. mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) added automatically) span_normalise Divide result span window? Default TRUE, see tskit documentation detail.","code":""},{"path":"https://www.slendr.net/reference/ts_segregating.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","text":"set individuals either single diversity value vector diversity values (one window)","code":""},{"path":"https://www.slendr.net/reference/ts_segregating.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # collect sampled individuals from all populations in a list sample_sets <- ts_samples(ts) %>%   split(., .$pop) %>%   lapply(function(pop) pop$name)  ts_segregating(ts, sample_sets) #> # A tibble: 4 × 2 #>   set   segsites #>   <chr>    <dbl> #> 1 AFR          1 #> 2 CH           0 #> 3 EUR        122 #> 4 NEA          7"},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":null,"dir":"Reference","previous_headings":"","what":"Simplify the tree sequence down to a given set of individuals — ts_simplify","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"function convenience wrapper around simplify method implemented tskit, designed work tree sequence data simulated SLiM using slendr R package.","code":""},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"","code":"ts_simplify(   ts,   simplify_to = NULL,   keep_input_roots = FALSE,   keep_unary = FALSE,   keep_unary_in_individuals = FALSE,   filter_nodes = TRUE )"},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"ts Tree sequence object class slendr_ts simplify_to character vector individual names. NULL, explicitly remembered individuals (.e. specified via schedule_sampling function left tree sequence simplification. keep_input_roots history ancestral MRCA samples retained tree sequence? Default FALSE. keep_unary unary nodes preserved simplification? Default FALSE. keep_unary_in_individuals unary nodes preserved simplification associated individual recorded table individuals? Default FALSE. set TRUE keep_unary also TRUE filter_nodes nodes reindexed simplification? Default TRUE. See tskit's documentation Python method simplify()","code":""},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"Tree-sequence object class slendr_ts, serves interface point Python module tskit using slendr functions ts_ prefix.","code":""},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"simplification process used remove redundant information tree sequence retains information necessary describe genealogical history set samples. information simplification works pyslim tskit, see official documentation https://tskit.dev/tskit/docs/stable/python-api.html#tskit.TreeSequence.simplify https://tskit.dev/pyslim/docs/latest/tutorial.html#simplification. clear description difference remembering retaining use techniques implement historical individuals (.e. ancient DNA samples) pyslim documentation https://tskit.dev/pyslim/docs/latest/tutorial.html#historical-individuals.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  ts <- ts_load(slendr_ts, model) ts #> ╔════════════════════════╗ #> ║TreeSequence            ║ #> ╠═══════════════╤════════╣ #> ║Trees          │      70║ #> ╟───────────────┼────────╢ #> ║Sequence Length│  500000║ #> ╟───────────────┼────────╢ #> ║Time Units     │   ticks║ #> ╟───────────────┼────────╢ #> ║Sample Nodes   │      26║ #> ╟───────────────┼────────╢ #> ║Total Size     │68.2 KiB║ #> ╚═══════════════╧════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │ 298│ 9.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│  73│ 8.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │   0│ 1.2 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │  86│ 3.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   5│ 2.6 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   3│37.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │   0│16 Bytes│          No║ #> ╚═══════════╧════╧════════╧════════════╝ #>   # simplify tree sequence to sampled individuals ts_simplified <- ts_simplify(ts)  # simplify to a subset of sampled individuals ts_small <- ts_simplify(ts, simplify_to = c(\"CH_1\", \"NEA_1\", \"NEA_2\", \"AFR_1\",                                             \"AFR_2\", \"EUR_1\", \"EUR_2\"))  ts_small #> ╔════════════════════════╗ #> ║TreeSequence            ║ #> ╠═══════════════╤════════╣ #> ║Trees          │      28║ #> ╟───────────────┼────────╢ #> ║Sequence Length│  500000║ #> ╟───────────────┼────────╢ #> ║Time Units     │   ticks║ #> ╟───────────────┼────────╢ #> ║Sample Nodes   │      14║ #> ╟───────────────┼────────╢ #> ║Total Size     │56.0 KiB║ #> ╚═══════════════╧════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │ 118│ 3.7 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│  32│ 4.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │   0│ 1.2 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │  39│ 2.1 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   5│ 2.6 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   4│37.8 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │   0│16 Bytes│          No║ #> ╚═══════════╧════╧════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/ts_table.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"function extracts data given tree sequence table. times converted model-specific time units tskit's \"generations backwards\" time direction.","code":""},{"path":"https://www.slendr.net/reference/ts_table.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"","code":"ts_table(ts, table = c(\"individuals\", \"edges\", \"nodes\", \"mutations\"))"},{"path":"https://www.slendr.net/reference/ts_table.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"ts Tree sequence object class slendr_ts table tree sequence table return","code":""},{"path":"https://www.slendr.net/reference/ts_table.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"Data frame information give tree-sequence table (can either table individuals, edges, nodes, mutations).","code":""},{"path":"https://www.slendr.net/reference/ts_table.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"processing analyses, output function ts_nodes might useful, merges information node individual tables one table annotates useful information model configuration data.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_table.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"","code":"# load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk and add mutations to it ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # get the 'raw' tskit table of individuals ts_table(ts, \"individuals\") #> # A tibble: 73 × 9 #>    ind_id  time pedigree_id  pop_id alive remembered retained sampled time_tskit #>     <dbl> <dbl>       <dbl> <int[1> <lgl> <lgl>      <lgl>    <lgl>    <dbl[1d]> #>  1      0   -10    16023346       0 TRUE  TRUE       TRUE     TRUE             0 #>  2      1   -10    16023350       1 TRUE  TRUE       TRUE     TRUE             0 #>  3      2   -10    16023352       1 TRUE  TRUE       TRUE     TRUE             0 #>  4      3   -10    16023353       1 TRUE  TRUE       TRUE     TRUE             0 #>  5      4   -10    16023355       1 TRUE  TRUE       TRUE     TRUE             0 #>  6      5   -10    16023359       1 TRUE  TRUE       TRUE     TRUE             0 #>  7      6   -10    16024625       3 TRUE  TRUE       TRUE     TRUE             0 #>  8      7   -10    16025081       3 TRUE  TRUE       TRUE     TRUE             0 #>  9      8   -10    16025520       3 TRUE  TRUE       TRUE     TRUE             0 #> 10      9   -10    16026684       3 TRUE  TRUE       TRUE     TRUE             0 #> # ℹ 63 more rows  # get the 'raw' tskit table of edges ts_table(ts, \"edges\") #> # A tibble: 298 × 5 #>       id child parent   left  right #>    <dbl> <int>  <int>  <dbl>  <dbl> #>  1     0     6     26      0 500000 #>  2     1    13     26      0 500000 #>  3     2     7     27      0 500000 #>  4     3     8     27      0 500000 #>  5     4    12     27      0 500000 #>  6     5     5     28      0 500000 #>  7     6    10     28      0 500000 #>  8     7    11     29 306056 500000 #>  9     8    26     29 306056 500000 #> 10     9    27     30      0 500000 #> # ℹ 288 more rows  # get the 'raw' tskit table of nodes ts_table(ts, \"nodes\") #> # A tibble: 86 × 5 #>    node_id    ind_id pop_id  time time_tskit #>      <int> <int[1d]>  <int> <dbl>      <dbl> #>  1       0        25      2 70010       2334 #>  2       1        25      2 70010       2334 #>  3       2        43      2 40010       1334 #>  4       3        43      2 40010       1334 #>  5       4         1      1   -10          0 #>  6       5         1      1   -10          0 #>  7       6         2      1   -10          0 #>  8       7         2      1   -10          0 #>  9       8         3      1   -10          0 #> 10       9         3      1   -10          0 #> # ℹ 76 more rows  # get the 'raw' tskit table of mutations ts_table(ts, \"mutations\") #> # A tibble: 2,246 × 5 #>       id  site  node     time time_tskit #>    <dbl> <int> <int>    <dbl>      <dbl> #>  1     0     0    83 3626483.    120883. #>  2     1     1    83 1766659.     58889. #>  3     2     2    83 2476958.     82566. #>  4     3     3    83 4356462.    145216. #>  5     4     4    83  876159.     29206. #>  6     5     5    83 3848814.    128294. #>  7     6     6    83 2554081.     85136. #>  8     7     7    83  976093.     32537. #>  9     8     8    32 3622571.    120753. #> 10     9     9    32  104626.      3488. #> # ℹ 2,236 more rows"},{"path":"https://www.slendr.net/reference/ts_tajima.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate Tajima's D for given sets of individuals — ts_tajima","title":"Calculate Tajima's D for given sets of individuals — ts_tajima","text":"discussion difference \"site\" \"branch\" options mode argument, please see tskit documentation https://tskit.dev/tskit/docs/stable/stats.html#sec-stats-mode","code":""},{"path":"https://www.slendr.net/reference/ts_tajima.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate Tajima's D for given sets of individuals — ts_tajima","text":"","code":"ts_tajima(ts, sample_sets, mode = c(\"site\", \"branch\", \"node\"), windows = NULL)"},{"path":"https://www.slendr.net/reference/ts_tajima.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate Tajima's D for given sets of individuals — ts_tajima","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set). simple vector provided, interpreted .list(sample_sets), meaning given statistic calculated individual separately. mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) added automatically)","code":""},{"path":"https://www.slendr.net/reference/ts_tajima.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate Tajima's D for given sets of individuals — ts_tajima","text":"set individuals either single Tajima's D value vector Tajima's D values (one window)","code":""},{"path":"https://www.slendr.net/reference/ts_tajima.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate Tajima's D for given sets of individuals — ts_tajima","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # calculate Tajima's D for given sets of individuals in a tree sequence ts ts_tajima(ts, list(eur = c(\"EUR_1\", \"EUR_2\", \"EUR_3\", \"EUR_4\", \"EUR_5\"),                    nea = c(\"NEA_1\", \"NEA_2\"))) #> # A tibble: 2 × 2 #>   set       D #>   <chr> <dbl> #> 1 eur   -1.16 #> 2 nea    2.18"},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"Extract data frame coordinates ancestry tracts given tree sequence.","code":""},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"","code":"ts_tracts(ts, census, squashed = TRUE, source = NULL, target = NULL)"},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"ts Tree sequence object class slendr_ts census Census time. See documentation linked Details information. slendr-specific tree sequence provided ts, census time expected given slendr model-specific time units, must correspond gene-flow event encoded model. squashed ancestry tracts squashed (.e., continuous tracts can traced different ancestral nodes merged)? Default TRUE. FALSE, effectively continuous ancestry tracts split individual segments, assigned specific ancestral node ID (recorded column ancestor_id). source source population extract tracts ? NULL (default), ancestry tracts populations contributing gene flow census time reported. Otherwise, ancestry tracts specified source populations extracted. Note option ignored non-slendr tree sequences! target Similar purpose source , except filters tracts discovered target population(s)","code":""},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"data frame containing coordinates ancestry tracts","code":""},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"function implements R-friendly interface algorithm extracting ancestry tracts provided Python module tspop https://tspop.readthedocs.io/en/latest/ developed Georgia Tsambos. Please make sure cite paper describes algorithm detail: https://academic.oup.com/bioinformaticsadvances/article/3/1/vbad163/7429395. technical details, see also tutorial : https://tspop.readthedocs.io/en/latest/basicusage.html. general, using function slendr-generated tree sequence, please aware output changes slightly get running pure tspop.get_pop_ancestry() Python. First, ts_tracts() populates output data frame additional metadata (names individuals populations). Additionally, slendr models, specifically designed return ancestry tracts originating ancestral population contributed ancestry gene-flow event started specific time (.e., scheduled model via gene_flow()) function. return every single ancestry tracts present tree sequence every single sample node (every single potential ancestry population) tspop.get_pop_ancestry() Python method. said, run tree sequence originate slendr simulation, behavior ts_tracts() identical underlying tspop.get_pop_ancestry(). current version slendr, ts_tracts() works slendr/msprime sequences slendr/SLiM tree sequences. Support slendr-generated SLiM tree sequences development. Tracts tree sequences originating non-slendr msprime SLiM simulations restricted way , mentioned previous paragraph, ts_tracts() situation effectively reduces standard tspop.get_pop_ancestry() call.","code":""},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env(quiet = TRUE)  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_msprime.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(file = slendr_ts, model = model)  # extract Neanderthal ancestry tracts (i.e. those corresponding to the # census event at the gene-flow time at 55000 kya as scheduled by # the simulation which produced the tree sequence) nea_tracts <- ts_tracts(ts, census = 55000, source = \"NEA\") #>  #> PopAncestry summary #> Number of ancestral populations: \t4 #> Number of sample chromosomes: \t\t26 #> Number of ancestors: \t\t\t763 #> Total length of genomes: \t\t130000000.000000 #> Ancestral coverage: \t\t\t120000000.000000 #>  nea_tracts #> # A tibble: 42 × 8 #>    name  node_id pop   source_pop    left   right length source_pop_id #>    <chr>   <dbl> <fct> <fct>        <dbl>   <dbl>  <dbl>         <dbl> #>  1 EUR_1      16 EUR   NEA          44049  101210  57161             2 #>  2 EUR_1      16 EUR   NEA         244041  280650  36609             2 #>  3 EUR_1      16 EUR   NEA        4636048 4736790 100742             2 #>  4 EUR_1      16 EUR   NEA        4756344 4850072  93728             2 #>  5 EUR_1      17 EUR   NEA         106906  268610 161704             2 #>  6 EUR_1      17 EUR   NEA        1420672 1432252  11580             2 #>  7 EUR_1      17 EUR   NEA        3438894 3511366  72472             2 #>  8 EUR_1      17 EUR   NEA        3624298 3682434  58136             2 #>  9 EUR_1      17 EUR   NEA        4139989 4239794  99805             2 #> 10 EUR_1      17 EUR   NEA        4318138 4351616  33478             2 #> # ℹ 32 more rows"},{"path":"https://www.slendr.net/reference/ts_tree.html","id":null,"dir":"Reference","previous_headings":"","what":"Get a tree from a given tree sequence — ts_tree","title":"Get a tree from a given tree sequence — ts_tree","text":"information optional keyword arguments see tskit documentation: https://tskit.dev/tskit/docs/stable/python-api.html#-treesequence-class","code":""},{"path":"https://www.slendr.net/reference/ts_tree.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get a tree from a given tree sequence — ts_tree","text":"","code":"ts_tree(ts, i, mode = c(\"index\", \"position\"), ...)"},{"path":"https://www.slendr.net/reference/ts_tree.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get a tree from a given tree sequence — ts_tree","text":"ts Tree sequence object class slendr_ts Position tree tree sequence. mode = \"index\", -th tree returned (zero-based indexing tskit), mode = \"position\", tree covering -th base simulated genome returned (, tskit's indexing). mode argument interpreted? Either \"index\" -th tree sequence genealogies, \"position\" along simulated genome. ... Additional keyword arguments accepted tskit.TreeSequence.tskit.TreeSequence.at_index methods","code":""},{"path":"https://www.slendr.net/reference/ts_tree.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get a tree from a given tree sequence — ts_tree","text":"Python-reticulate-based object class tskit.trees.Tree","code":""},{"path":"https://www.slendr.net/reference/ts_tree.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Get a tree from a given tree sequence — ts_tree","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract the zero-th tree in the tree sequence tree <- ts_tree(ts, i = 0)  # extract the tree at a position in the tree sequence tree <- ts_tree(ts, i = 100000, mode = \"position\")"},{"path":"https://www.slendr.net/reference/ts_vcf.html","id":null,"dir":"Reference","previous_headings":"","what":"Save genotypes from the tree sequence as a VCF file — ts_vcf","title":"Save genotypes from the tree sequence as a VCF file — ts_vcf","text":"Save genotypes tree sequence VCF file","code":""},{"path":"https://www.slendr.net/reference/ts_vcf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Save genotypes from the tree sequence as a VCF file — ts_vcf","text":"","code":"ts_vcf(ts, path, chrom = NULL, individuals = NULL)"},{"path":"https://www.slendr.net/reference/ts_vcf.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Save genotypes from the tree sequence as a VCF file — ts_vcf","text":"ts Tree sequence object class slendr_ts path Path VCF file chrom Chromosome name written CHROM column VCF individuals character vector individuals tree sequence. missing, individuals present tree sequence saved.","code":""},{"path":"https://www.slendr.net/reference/ts_vcf.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Save genotypes from the tree sequence as a VCF file — ts_vcf","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/world.html","id":null,"dir":"Reference","previous_headings":"","what":"Define a world map for all spatial operations — world","title":"Define a world map for all spatial operations — world","text":"Defines either abstract geographic landscape (blank containing user-defined landscape) using real Earth cartographic data Natural Earth project (https://www.naturalearthdata.com).","code":""},{"path":"https://www.slendr.net/reference/world.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Define a world map for all spatial operations — world","text":"","code":"world(   xrange,   yrange,   landscape = \"naturalearth\",   crs = NULL,   scale = c(\"small\", \"medium\", \"large\") )"},{"path":"https://www.slendr.net/reference/world.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Define a world map for all spatial operations — world","text":"xrange Two-dimensional vector specifying minimum maximum horizontal range (\"longitude\" using real Earth cartographic data) yrange Two-dimensional vector specifying minimum maximum vertical range (\"latitude\" using real Earth cartographic data) landscape Either \"blank\" (blank abstract geography), \"naturalearth\" (real Earth geography) object class sf defining abstract geographic features world crs EPSG code coordinate reference system use spatial operations. CRS assumed default (NULL), implying abstract landscape tied real-world geographic region (landscape = \"blank\" landscape custom-defined geographic landscape), implying WGS-84 (EPSG 4326) coordinate system real Earth landscape defined (landscape =   \"naturalearth\"). scale Natural Earth geographic data used (.e. landscape =   \"naturalearth\"), parameter determines resolution data used. value \"small\" corresponds 1:110m data provided package, values \"medium\" \"large\" correspond 1:50m 1:10m respectively downloaded internet. Default value \"small\".","code":""},{"path":"https://www.slendr.net/reference/world.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Define a world map for all spatial operations — world","text":"Object class slendr_map, encodes standard spatial object class sf additional slendr-specific attributes requested x-range y-range.","code":""},{"path":"https://www.slendr.net/reference/world.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Define a world map for all spatial operations — world","text":"","code":"# create a blank abstract world 1000x1000 distance units in size blank_map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # it is possible to construct custom landscapes (islands, corridors, etc.) island1 <- region(\"island1\", polygon = list(c(10, 30), c(50, 30), c(40, 50), c(0, 40))) island2 <- region(\"island2\", polygon = list(c(60, 60), c(80, 40), c(100, 60), c(80, 80))) island3 <- region(\"island3\", center = c(20, 80), radius = 10) archipelago <- island1 %>% join(island2) %>% join(island3)  custom_map <- world(xrange = c(1, 100), c(1, 100), landscape = archipelago)  # real Earth landscapes can be defined using freely-available Natural Earth # project data and with the possibility to specify an appropriate Coordinate # Reference System, such as this example of a map of Europe  real_map <- world(xrange = c(-15, 40), yrange = c(30, 60), crs = \"EPSG:3035\") #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpFhhY5X/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84"},{"path":"https://www.slendr.net/news/index.html","id":"slendr-090","dir":"Changelog","previous_headings":"","what":"slendr 0.9.0","title":"slendr 0.9.0","text":"full support running SLiM msprime simulations slendr analyzing tree sequences using tskit interface implemented. Please note Windows support still rather experimental – internal slendr test suite currently assumes SLiM installed using msys2 system described section 2.3.1 SLiM manual means installing SLiM (via conda) might require additional adjustments. fallback option form slim_path= argument slim() function can used non-standard SLiM installation circumstances. convenience, please add path directory containing slim.exe binary PATH variable editing C:/Users/<username>/Documents/.Renviron file accordingly. See relevant section Windows installation slendr documentation additional information. Feedback Windows functionality bug reports highly appreciated via GitHub issues! Many thanks @GKresearch @rdinnager huge help making Windows port happen! (PR #149) trivial change made slendr’s SLiM back-end script fixing issue introduced SLiM 4.1 upgrade (see changelog version 0.8.1 ). expected lead different simulation outputs two versions slendr (0.8.2 vs 0.8.1) SLiM (4.1 vs 4.0.1) used. (PR #148) msprime internal dependency slendr updated 1.3.0, Python 3.12. result, loading slendr, users prompted re-run setup_env() make sure dedicated slendr Python environment fully updated. time, prevents failing installation (least) M1 macOS using pip. (#5ce212, #a210d4)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-081","dir":"Changelog","previous_headings":"","what":"slendr 0.8.1","title":"slendr 0.8.1","text":"CRAN release: 2024-01-15 Fixed issue apparent contradiction time direction models range expansion scheduled within time interval together associated “locked-” changes population size time interval. (#d2a29e) introduction tspop installable via pip (see changelog previous version) caused GLIBCXX-related errors conda pip dependencies related pandas Python package. work around issue, setup_env() longer installs pandas conda regardless setting pip = TRUE|FALSE parameter. Instead, pandas installed via pip single step tspop installed. (#cbe960) WARNING: SLiM 4.1 just released includes couple backwards incompatible changes related implementation spatial maps prevent current version slendr’s slim() function working correctly. rely functionality provided slim() function, use SLiM 4.0. (Note want multiple versions SLiM system, can either use slim_path = argument slim() specify $PATH required version SLiM ~/.Renviron file just like normal circumstances). Porting slendr SLiM 4.1 worked .","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-080","dir":"Changelog","previous_headings":"","what":"slendr 0.8.0","title":"slendr 0.8.0","text":"CRAN release: 2023-12-07 order support new ts_tracts() function backed tspop module (see item ), new slendr Python environment required. , users run setup_env() get required Python dependencies now installed internal virtual environment named Python-3.11_msprime-1.2.0_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2. (#b5330c) Experimental support tspop link-ancestors algorithm detection ancestry tracts form new slendr function ts_tracts(). works slendr-generated msprime tree sequences “pure” msprime SLiM tree sequences (slendr-generated SLiM tree sequences), tested toy models. Note: tspop Python module published conda. order set new slendr Python environment, run setup_env(pip = TRUE) make sure Python dependencies installed pip instead conda. (PR #145) Updated Python dependencies (bugfix pyslim release v1.0.4 tskit v0.5.6, latter due broken jsonschema dependency tskit). (#001ee5) Experimental support manually created spatial tree sequences. (PR #144)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-072","dir":"Changelog","previous_headings":"","what":"slendr 0.7.2","title":"slendr 0.7.2","text":"CRAN release: 2023-08-08 new function ts_names() added, avoiding need extremely frequent (, unfortunately, cumbersome) trick getting named lists individual symbolic names ts_samples(ts) %>% split(., .[[split]]) %>% lapply([[, \"name\") confusing experienced R users. (#7db6ea) Fixed broken concatenation symbolic sample names tree-sequence statistic functions, provided unnamed single-element lists character vectors. (#b3c650) plot_model() now argument file =, making possible save visualization model without actually opening plotting device. can useful particularly working remote server, order avoid often slow X11 rendering. (#e60078) plot_model() now argument order = allowing override default -order ordering populations along x-axis. (#7a10ea)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-071","dir":"Changelog","previous_headings":"","what":"slendr 0.7.1","title":"slendr 0.7.1","text":"CRAN release: 2023-07-14 Starting release, *spatial* simulation data analysis functionality slendr conditional presence R geospatial packages sf, stars, rnaturalearth system. means users able install slendr (use non-spatial functionality) even without R packages installed. said, nothing really changes practice: spatial features slendr just one install.packages(c(\"sf\", \"stars\", \"rnaturalearth\")) away! difference slendr doesn’t try installation, users instructed (needed) package loaded. (#7a10ea) spatial dependencies present spatial slendr function called regardless (world(), move(), etc.), error message printed information install spatial dependencies via install.packages() . ? ’s true main reason slendr’s existence ability simulate spatio-temporal data realistic landscapes via SLiM. However, practice, “average” uses slendr wild (classrooms!) rely traditional, non-spatial interface, spatial features used comparatively rarely moment (except cutting-edge exploratory research). Given setting spatial dependencies can bit hurdle, decided make dependencies optional, rather force every user go process installation whether need spatial features . function check_dependencies() now exported can used check whether slendr Python environment () SLiM () present. useful software building upon slendr, normal users can freely ignore . (#6ae6ce) path file tree sequence loaded now tracked internally via attr(<tree sequence>, \"path\") attribute. Note implemented purposes clean large-scale simulation studies (facilitated demografr) mostly internal feature, considered experimental. (#f181a2) Attempts resize population right time split (led issues simulations) now prevented. (#f181a2) Fix minor issue preventing sampling msprime population right time creation. (#aea231)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-070","dir":"Changelog","previous_headings":"","what":"slendr 0.7.0","title":"slendr 0.7.0","text":"CRAN release: 2023-06-26 emergency upgrade match latest pyslim 1.0.3 due serious bug recapitation. See extensive discussion process identification bug eventual fix. brief summary practical consequences bug, see thread pyslim’s developer formal announcement . change require re-run setup_env() order update slendr’s Python internals creating new internal Python virtual environment. (#45539a) potential issue parent population scheduled removal daughter population splits now caught moment daughter population() call rather simulation slim() run. (#0791b5) function plot_model() new argument gene_flow=<TRUE|FALSE> determines whether gene-flow arrows visualized . (#104aa6) possibility perform recapitation, simplification, mutation tree sequence right inside call ts_load() (providing recapitate = TRUE, simplify = TRUE, mutate = TRUE, together arguments) now removed. motivation change realization benefit things like ts_load(\"<path>\", recapitate = TRUE, Ne = ..., recombination_rate = ...) ts_load(\"<path>\") %>% ts_recapitate(Ne = ..., recombination_rate = ...), frequent confusion recapitate = TRUE switches forgotten user. slendr teaching material actively used research codebases know use latter, explicit, pipeline approach anyway, one example reduncancy harm good. (#ad82ee) Note: Loading library(slendr) prompt message “legacy packages maptools, rgdal, rgeos, underpinning sp package, just loaded, retire October 2023. […].” internal business packages used slendr unfortunately silenced slendr’s side. ’s reason panic, can safely ignore . Apologies unnecessary noise.","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-060","dir":"Changelog","previous_headings":"","what":"slendr 0.6.0","title":"slendr 0.6.0","text":"CRAN release: 2023-06-07 relatively large update, unfortunately released haste due retirement rgdal package – significant dependency entire spatial R ecosystem phased effort move towards modern low-level geospatial architecture. Although slendr depend rgdal, many dependencies used (won’t short term, hence push remove rgdal dependency). significant update addition IBD functionality tskit, described . However, large part functionality extensively tested considered extremely experimental stage. like use , might safer either wait later release IBD functionality stable, use underlying, battle-tested Python implementation tskit. ts_ibd() now returns ID number MRCA node pair nodes sharing given IBD segment, well TMRCA node. (#7e2825) Trivial parameter errors caught population() calls rather simulation (solving minor issues discovered via big simulation runs development demografr). (#e33373) Fix error plotting exponential resizes last “present”. (#4c49a4) ts_ibd() longer gives obscure error = provided named list individuals’ names (instead expected unnamed list). names list elements used way, error happens somewhere deeply R->Python translation layer inside reticulate ’s need users concern . (#7965e4) Population size parameters times now explicitly converted integer numbers. internal, formal change (conversion happening implicitly inside SLiM engine anyway) now explicitly stated, also documentation relevant function. (#b7e89e) Population names now restricted strings also valid Python identifiers. Although restriction needed msprime back end slendr (SLiM), makes sense keep things tidy unified. fixes msprime crashing ValueError: population name must valid Python identifier. (#4ef518) layout algorithm plot_model() improved significantly. (PR #135). new optional argument run = added slim() msprime(). set TRUE (default), engines operate usual way. set FALSE, simulation run functions simply print command-line command execute engine question (returning CLI command invisibly). (#2e5b85) following start-note longer shown upon calling library(slendr): Users call init_env() manually activate Python environment slendr (see note version 0.5.0 extended explanation). ts_simplify() now accepts optional arguments keep_unary keep_unary_in_individuals (see official tskit docs detail) (#1b2112) Fix ts_load() failing load slendr-produced tree sequences simplified smaller set sampled individuals (reported ). issue caused incompatible sizes sampling table (always form used simulation) table individuals stored tree sequence simplification (potentially containing smaller set individuals original sampling table). fix , slendr tree sequence objects now track information individuals regarded “samples” (.e. symbolic names) maintained simplification, serialization loading, used slendr’s internal machinery join operations. (PR #137) Metadata summary ts_nodes() results longer printed whenever typed R console. Instead, summary can obtained explicit call summary() ts_nodes() tables. (#01af51 ts_tree() ts_phylo() now extract trees based tskit’s zero-based indexing #554e13. ts_simplify() now accepts filter_nodes = TRUE|FALSE, behavior tskit’s method #f07ffed.","code":"NOTE: Due to Python setup issues on some systems which have been     causing trouble particularly for novice users, calling library(slendr)     no longer activates slendr's Python environment automatically.     In order to use slendr's msprime back end or its tree-sequence     functionality, users must now activate slendr's Python environment     manually by executing init_env() after calling library(slendr).     (This note will be removed in the next major version of slendr.)"},{"path":"https://www.slendr.net/news/index.html","id":"slendr-051","dir":"Changelog","previous_headings":"","what":"slendr 0.5.1","title":"slendr 0.5.1","text":"CRAN release: 2023-03-09 minor release implements emergency fix CRAN warning suddenly popped latest CRAN checks. (#5600a4) new function ts_ibd() added, representing R interface tskit method TreeSequence.ibd_segments(). However, note ts_ibd() returns IBD results data frame (optionally, spatially annotated sf data frame). function operate around iteration, Python counterpart tskit. next major version slendr, function considered experimental. (PR #123)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-050","dir":"Changelog","previous_headings":"","what":"slendr 0.5.0","title":"slendr 0.5.0","text":"CRAN release: 2023-02-02 Minor breaking change! Python environments slendr longer automatically activated upon calling library(slendr)! Using coalescent msprime back end slendr’s tree-sequence functions now requires making explicit call new function init_env() library(slendr) executed. (PR #102) Motivation change: small proportion users experiencing issues broken conda environments various issues Python virtual environments general. ’s hard guess frequent , experience workshops courses suggests perhaps 1 20 users experiencing Python issues hindered ability use slendr .(Fun fact: first user-submitted GitHub issue upon releasing first version slendr R package … Python virtual environment issue). Explanation: Activating Python environments automatically upon calling library(slendr) popular feature hid away complexities R-Python interface powers slendr’s tree-sequence functionality. particularly convenient many slendr users, particularly experience Python . Unfortunately, cases Python virtual environments tskit/msprime/pyslim user’s system ended corrupted (anything else Python level got broken), automatic Python environment activation performed library(slendr) call failed slendr even loaded. Sadly, completely pulled rug slendr nothing done perspective (issue happened low-level layer embedded-Python slendr ’ve loaded R). Solving issues difficult experienced users, many slendr users experience Python , never used conda, don’t understand concept “Python virtual environments” R-Python interface works. ! , slendr R package. Splitting Python virtual environment activation step init_env() function means library(slendr) now always succeeds (regardless potential underlying Python issues user’s sytem), making much easier diagnose fix Python problems R package loaded. , recap: library(slendr) longer activates slendr’s isolated Python virtual environment. order simulate tree sequences analyse using interface tskit, necessary call init_env(). function performs Python-activation steps library(slendr) used call automagically earlier slendr versions. change scripts necessary. Related previous point: slendr now requires Python 3.11, msprime 1.2.0, tskit 0.5.4, pyslim 1.0.1, keep recent releases Python dependencies. , presents hassle user, thing required re-running setup_env(). (PR #112). named list provided sample_sets = argument oneway statistic function, names used set column resulting data frame even single samples used. (#2a6781) now possible non-spatial populations otherwise spatial model. course, plotting models map, spatial components model plotted slendr give warning. absolutely sure users intends , slendr also give warning running compile_model() models like . Please consider option experimental time-hard predict edge cases might break (unit tests documentation tests passing though). Feedback welcome. (PR #112). now possible label groups samples slendr’s tskit interface functions make data frames statistics results readable. example, running ts_f3(ts, = c(\"p1_1\", \"p1_2\", \"p1_3\"), B = c(\"p2_1\", \"p2_3\"), C = c(\"p3_1\", \"p3_2\", \"p3_\")) resulted following data-frame output: gets unwieldy rather quickly, especially dozens hundreds samples grouped together populations. new syntax allows following shortcut via customised group names leveraging standard named list functionality R: readable line tskit-interface functions slendr used functionality via sample_sets = argument (ts_divergence(), ts_diversity(), etc.). (#ac5e484) default state parent = argument population() now NULL instead \"ancestor\". prevents silly surprising clashes situation population’s name really “ancestor”. change internally populations ancestral, splits data frame element slendr model object includes population carries formal “ancestral parent population” \"__pop_is_ancestor\" instead just \"ancestor\". Note internal implementation detail something particularly involve user. Still, somehow using slendr’s internal data structures, keep mind. (#f8a39a2)","code":"> ts_f3(ts, A = c(\"p1_1\", \"p1_2\", \"p1_3\", \"p1_4\", \"p1_5\"),             B = c(\"p2_1\", \"p2_2\", \"p2_3\"),             C = c(\"p3_1\", \"p3_2\", \"p3_3\", \"p3_4\"))  # A tibble: 1 × 4   A                        B              C                         f3   <chr>                    <chr>          <chr>                  <dbl> 1 p1_1+p1_2+p1_3+p1_4+p1_5 p2_1+p2_2+p2_3 p3_1+p3_2+p3_3+p3_4 0.000130 > ts_f3(ts, A = list(group_one = c(\"p1_1\", \"p1_2\", \"p1_3\", \"p1_4\", \"p1_5\")),             B = list(group_two = c(\"p2_1\", \"p2_2\", \"p2_3\")),             C = list(group_three = c(\"p3_1\", \"p3_2\", \"p3_3\", \"p3_4\"))) # A tibble: 1 × 4   A         B         C                 f3   <chr>     <chr>     <chr>          <dbl> 1 group_one group_two group_three 0.000130"},{"path":"https://www.slendr.net/news/index.html","id":"slendr-040","dir":"Changelog","previous_headings":"","what":"slendr 0.4.0","title":"slendr 0.4.0","text":"CRAN release: 2022-09-30 msprime() function now makes sure given slendr model can fully coalesce single common ancestor population. Previously, multiple ancestral populations created parent = \"ancestor\" cause infinite simulation plugged msprime() backend. (#095b124) initial size population emerges split another population now printed population history summary R console. (#6525bf3) couple fixes support loading, processing, plotting “manually” created tree sequences implemented (see ). sure practically useful, ’s important able load even “pure” tree sequences simulators SLiM msprime. set unit tests added, making sure minimalist nodes & edges table can loaded, well nodes & edges & individuals, plus tables populations sites & mutations. PRs extensive unit tests bug reports tree sequences failing load appreciated! code handling cases “manually-created” tree sequences missing individual table, missing populations table, etc. seems especially brittle moment (#79adf14). -1 value missing value indicator used tskit now replaced R-like NA various tree-sequence tables (annotated slendr original tskit ) (#79adf14). Relative paths now expanded ts_save() (#382e0b7). slendr models can now optionally compiled without serialization disk. works msprime() coalescent back end much faster cases huge number simulations needs run non-serialized models, msprime() now calls back end engine directly R-Python interface (rather command line) output tree sequences saved disk, rather passed Python-R interface directly memory (PR #112). Deprecated argument sampling = functions slim() msprime() now permanently removed favour samples = argument (#0757b6e). Avoid unnecessary array type tskit results returned via reticulate. Numeric vectors (columns data frames numerical results) obtained way simple R numeric vector (#5101b39). One-way multi-way statistics results now returned simple numerical vectors. Previously, results returned type array despite “looking” vectors (values returned R reticulate-Python layer), caused unnecessary annoyances type-conversions R side things even intended (#403df3b). Computing population genetic statistics named samples present tree sequence (likely typos) now correctly caught reported error (#da7e0bb).","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-030","dir":"Changelog","previous_headings":"","what":"slendr 0.3.0","title":"slendr 0.3.0","text":"CRAN release: 2022-08-19 SLiM 4.0 now required running simulations slim() engine. want run slendr simulations SLiM (spatial non-spatial), need upgrade SLiM installation. SLiM 3.7.1 version longer supported upcoming new slendr spatial features depend SLiM 4.x maintaining two functionally identical yet syntactically different back ends feasible (PR #104). time SLiM 4.0 release, new versions Python modules msprime, tskit pyslim also released. fact, able work SLiM 4.0 tree sequences properly, Python modules must upgraded well. Next time load library(slendr), prompted setup new updated Python environment can easily running setup_env(). Experimental support running coalescent msprime simulations analysing tree-sequence data using tskit Windows platform now implemented (PR #102).","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-020","dir":"Changelog","previous_headings":"","what":"slendr 0.2.0","title":"slendr 0.2.0","text":"CRAN release: 2022-08-09 slendr now CRAN! Big changes way tree-sequence outputs handled slendr default. See comment extended description examples change. (PR #100). Briefly, simulation functions slim() msprime() now return tree-sequence object default (can switched setting load = FALSE), avoiding need always run ts <- ts_load(model) previously. time, parameter output = can now used slim() msprime() specify location tree-sequence file saved (temporary file default). slendr’s tree-sequence R interface tskit Python module generalized load, process, analyze tree sequences non-slendr models! means users can use slendr R package even analyzing tree sequences coming standard msprime SLiM scripts, including spatial capabilities available slendr tree sequences far. Please note generalization still rather experimental might corner cases tree sequence msprime SLiM script load properly leads errors. happens, please open GitHub issue script question attached. (PR #91) Removed functions function arguments originally deprecated renaming phase pre-preprint refactoring. affects compile, boundary, dispersal, expand, geneflow, plot.slendr, plot_graph, read, sampling, shrink. Similarly, deprecated dir argument compile_model now path, geneflow argument compile_model now gene_flow, _dist suffix removed competition_dist, mate_dist, dispersal_dist. get error missing function function argument code used work ancient version slendr, . (#985b451) setting isolated Python environment using setup_env(), slendr now makes decision whether install Python dependencies using pip (critical osx-arm64 conda msprime/tskit unfortunately currently broken) conda (every platform). can still influenced user using pip = <TRUE|FALSE> argument, now change default behavior ARM64 Mac. (#54a413d) name default slendr Python environment now shortened even , redundant _pandas prefix now dropped. Users notified upon calling library(slendr) new environment created. OK, ’s bug. (#54a413d) format default slendr Python environment now msprime-<version>_tskit-<version>_pyslim-<version>_pandas, dropping slendr_ prefix. paves way towards future non-slendr tskit R package, share Python environment slendr (R packages go hand hand). isn’t really user-facing change, except calling setup_env() suggests creating new Python environment library(slendr) appear slendr environment yet present. Calling setup_env() creating new Python environment scratch solve problem. (#eb05180) xrange yrange parameters world() now enforced two-dimensional numeric vectors, avoiding unnecessary issues misspecified longitude/latitude (#df95369) argument sampling = slim() msprime() now renamed samples = (#adf4e0d). automated setup_env() function creating dedicated mini Python environments slendr now installs packages using pip default. Reason: rate conda failures dependency conflicts (even trivial case installing nothing msprime + tskit + pyslim + pandas) high rely . option use conda package installations setup_env() still , users must explicitly call setup_env(pip = FALSE) get behavior. Note conda still used means install Python ! change affects way Python modules installed dedicated slendr Python environment, installation Python . (#81be1a7) name automatically created slendr-specific Python environment now composed names versions Python modules installed. makes possible naturally upgrade slendr Python dependencies case tskit / msprime / pyslim folks upgrade packages. case, slendr user upgrades slendr package (new version requires newer versions Python modules), slendr simply recommend create new Python environment without additional effort part. (#81be1a7) code setup_env() simplified bare essentials. Now serves way auto-setup dedicated, isolated Python installation slendr environment. interface install Python modules custom-defined Python environment created outside R removed functionality necessary – custom environments can easily activated calling reticulate::use_virtualenv reticulate::use_condaenv. (#30f24b9) Python users want use custom Python environments msprime, tskit, pyslim, can silence suggestion use setup_env() printed library(slendr) call setting options(slendr.custom_env = TRUE). (#30f24b9) argument sim_length = now renamed simulation_length =. accepted moment using old name simply inform user future deprecation. (#56491fb) Extensive set runnable examples including figures built-pre-compiled example model added documentation. (#395df62c)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-010","dir":"Changelog","previous_headings":"","what":"slendr 0.1.0","title":"slendr 0.1.0","text":"First numbered version slendr celebrate bioRxiv preprint. 🥳 🎉","code":""}]
+[{"path":"https://www.slendr.net/LICENSE.html","id":null,"dir":"","previous_headings":"","what":"MIT License","title":"MIT License","text":"Copyright (c) 2022 Martin Petr Permission hereby granted, free charge, person obtaining copy software associated documentation files (“Software”), deal Software without restriction, including without limitation rights use, copy, modify, merge, publish, distribute, sublicense, /sell copies Software, permit persons Software furnished , subject following conditions: copyright notice permission notice shall included copies substantial portions Software. SOFTWARE PROVIDED “”, WITHOUT WARRANTY KIND, EXPRESS IMPLIED, INCLUDING LIMITED WARRANTIES MERCHANTABILITY, FITNESS PARTICULAR PURPOSE NONINFRINGEMENT. EVENT SHALL AUTHORS COPYRIGHT HOLDERS LIABLE CLAIM, DAMAGES LIABILITY, WHETHER ACTION CONTRACT, TORT OTHERWISE, ARISING , CONNECTION SOFTWARE USE DEALINGS SOFTWARE.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"quick-installation","dir":"Articles","previous_headings":"","what":"Quick installation","title":"Installation instructions","text":"slendr available CRAN R package repository. , can install simply executing install.packages(\"slendr\") R console. want (need) get development version, can install directly GitHub executing devtools::install_github(\"bodkan/slendr\") via R package devtools (can gen devtools running install.packages(\"devtools\")). fact, decide try slendr, please make sure update regularly keep eye changelog regular basis! can find information latest bugfixes potential breaking changes. install slendr, calling library(slendr) check software dependencies available. , R package provide brief helpful guide resolve potential issues. rest vignette talks necessary software dependencies bit detail. Please note slendr extensively tested macOS Linux moment. However, experimental support runnig SLiM msprime simulations Windows analyzing tree-sequence outputs using tskit interface platform well. Feedback issues using slendr Windows highly appreciated.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"software-dependencies","dir":"Articles","previous_headings":"","what":"Software dependencies","title":"Installation instructions","text":"slendr relies three main software dependencies: geospatial data analysis R package sf (encoding spatial slendr models analysing spatial tree-sequence data), forward population genetic simulator SLiM (forward simulations), Python modules tskit, msprime, pyslim (coalescent simulations tree-sequence analysis), also pandas used internally simulation back ends. three widely used respective fields , , easily obtainable major operating systems (see information troubleshoot potential problems). Note depending use case, three sets dependencies necessarily needed. ’re going running forward spatial simulations, don’t need SLiM geospatial R packages sf, stars, rnaturalearth. applies also animation spatial models using gganimate R package. slendr install dependencies. need functionality, install respective R packages manually. vignette, briefly explain get slendr’s software dependencies installed. said, note normal circumstances (exception SLiM), manual installation individual dependencies required.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"sf-stars-rnaturalearth","dir":"Articles","previous_headings":"Software dependencies","what":"sf, stars, rnaturalearth","title":"Installation instructions","text":"R package sf heart geospatial data analysis R. available CRAN can installed major platforms executing install.packages(\"sf\") R session. applies stars rnaturalearth. first load slendr via library(slendr), ’re missing three geospatial R packages, notified instructed can easily obtain CRAN using single call install.packages(). said, sf depends number geospatial libraries depending exact setup Linux macOS machine, libraries missing. Luckily, easy install via Homebrew (macOS) via appropriate package manager Linux distribution (Ubuntu, Fedora, etc.). Detailed instructions operating system can found . ’re problems installation three packages, look help .","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"macos","dir":"Articles","previous_headings":"Software dependencies > sf, stars, rnaturalearth","what":"macOS","title":"Installation instructions","text":"One user recently installed slendr fresh macOS system reported needed install libgit2 order able install package devtools devtools::install_github(\"bodkan/slendr\") step described top page. Additionally, install couple C/C++ libraries well (dependencies sf package). end, able successfully install slendr running: Note assumes Homebrew package manager already setup Mac. ’re beginning computational scientist using Mac, strongly encourage install Homebrew. Sooner later need specific Linux/unix program anyway, Homebrew way get (Mac unix machine, without Homebrew poor one default).","code":"brew install libgit2 udunits gdal proj"},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"linux","dir":"Articles","previous_headings":"Software dependencies > sf, stars, rnaturalearth","what":"Linux","title":"Installation instructions","text":"Testing slendr installation fresh, pristine Debian installation dependencies previously installed, run following: ’s unlikely need (might need packages non-Debian distributions), got slendr dependencies running completely clean system. Might good start case trouble Linux machine. Windows special treatment necessary get slendr running. install slendr via install.packages(\"slendr\"), get binary version package dependencies without need compiling sources.","code":"sudo apt-get install libudunits2-dev libssl-dev libgdal-dev libgsl-dev libgit2-dev libfontconfig1-dev libharfbuzz-dev libfribidi-dev"},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"slim","dir":"Articles","previous_headings":"Software dependencies","what":"SLiM","title":"Installation instructions","text":"forward population genetic software SLiM available major software platforms. complete installation instructions can found . Mac, recommend installing SLiM via pkg installer available direct download website. Linux, can either install SLiM via appropriate package manager Linux distribution (see SLiM manual information), can easily compile . install SLiM Windows, please follow instructions section 2.3.1 SLiM manual. Note although SLiM also available conda, comes without SLiMgui! Note slendr requires SLiM 4.0 work earlier version. , running library(slendr) inform potential issues SLiM installation.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"adding-slim-to-the-path","dir":"Articles","previous_headings":"Software dependencies > SLiM","what":"Adding SLiM to the $PATH","title":"Installation instructions","text":"order slendr able find SLiM installation, R must able find slim binary (slim.exe case Windows) called $PATH variable. easiest way verify true call Sys.(\"slim\") (Sys.(\"slim.exe\") Windows) R session. instance, Mac, get : Windows, get (followed section 2.3.1 SLiM manual describing “official” way install SLiM recommend follow well): , hand, might get something like — empty string: means $PATH R configured properly R (slendr) won’t able find SLiM $PATH. add SLiM $PATH ? Probably convenient way editing .Renviron file. precise location file depends operation system can automatically get open text editor using command (might install.packages(\"usethis\") first): , can either add following (note ’s $ line!): Alternatively, already $PATH contents specified R session, can get calling Sys.getenv(\"PATH\") R console, grab entire string get way, append path installed SLiM using appropriate delimiter (: Linux/macOS, ; Windows) string. case, edit might look something like : instance, Mac, .Renviron file contains line (note last item /usr/local/bin matches path slim showed ): testing Windows machine, .Renviron file (, look path C:/msys64/mingw64/bin;$PATH\" matches slim.exe binary shown ): important check R can find SLiM using Sys.() command described . gives positive result, won’t able use slendr’s spatial SLiM simulations. ’re struggling , search advice related .Renviron $PATH online.","code":"> Sys.which(\"slim\")                  slim  \"/usr/local/bin/slim\" > Sys.which(\"slim.exe\")                             slim.exe  \"C:\\\\msys64\\\\mingw64\\\\bin\\\\slim.exe\" > Sys.which(\"slim\") slim    \"\" usethis::edit_r_environ() PATH=\"<path to directory with SLiM binaries>\" PATH=\"<original Sys.getenv(\"PATH\") content><delimiter><path to directory with SLiM binaries>\" PATH=\"/opt/homebrew/bin:/opt/homebrew/sbin:/Users/mp/.my_local/bin:/Users/mp/.my_local/AdmixTools/bin:/Library/TeX/texbin:/opt/homebrew/opt/gnu-sed/libexec/gnubin:/opt/homebrew/opt/coreutils/libexec/gnubin:/usr/local/bin\" PATH=\"C:\\\\msys64\\\\usr\\\\bin;C:\\\\rtools43\\\\x86_64-w64-mingw32.static.posix\\\\bin;C:\\\\rtools43\\\\usr\\\\bin;C:\\\\Program Files (x86)\\\\R\\\\R-4.3.2\\\\bin\\\\x64;C:/msys64/mingw64/bin\""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"fallback-options","dir":"Articles","previous_headings":"Software dependencies > SLiM > Adding SLiM to the $PATH","what":"Fallback options","title":"Installation instructions","text":"don’t want deal editing .Renviron file, able set path SLiM using command Sys.setenv(PATH = \"<path directory SLiM binaries>\"). beginning slendr R scripts though. Alternatively, slim() function argument slim_path= can specify full path slim slim.exe binaries directly (directory ’s case $PATH! full path slim/slim.exe binary ).","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"python","dir":"Articles","previous_headings":"Software dependencies","what":"Python","title":"Installation instructions","text":"order able run coalescent simulations process tree-sequence files, slendr needs Python modules tskit, msprime, pyslim (also needs pandas library). Setting isolated Python environment specific version Python packages (important avoid clashes among different Python programs needed system) can bit hassle users. especially true R users might use Python daily work. order make sure R package appropriate version Python available, correct versions Python module dependencies, slendr provides dedicated function setup_env() automatically downloads completely separate Python distribution installs required versions tskit, msprime, pyslim modules dedicated virtual environment. Moreover, Python installation virtual environment entirely isolated Python configurations already present user’s system, avoiding potential conflicts versions Python Python modules required slendr. Next time call library(slendr), need activate environment automatically calling init_env(). ’re comfortable Python don’t need worry beyond calling setup_env() init_env(), interaction Python necessary working slendr R.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"failing-conda","dir":"Articles","previous_headings":"Software dependencies > Python","what":"Failing conda?","title":"Installation instructions","text":"order support Windows, slendr uses conda download Python interpreter explained . Given fact, run setup_env(), slendr tries leverage conda present install Python dependencies (msprime, tskit, pyslim, pandas) via conda . Unfortunately, conda can break frustratingly many random reasons completely trips setup_env(). run issue, fallback option install Python dependencies msprime, tskit, pyslim, pandas via pip , unlike conda, works practically every time. setup_env() fails conda-related reason, first clear broken environment calling clear_env(), restart R session, call setup_env(pip = TRUE) instead default setup_env(). Note might require install GSL numerical library, ’s trivial issue macOS (brew install gsl) Linux (Ubuntu, instance, sudo apt-get install libgsl-dev).","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"information-for-python-experts","dir":"Articles","previous_headings":"Software dependencies > Python","what":"Information for Python experts","title":"Installation instructions","text":"case wondering slendr accomplish : slendr’s Python interface implemented using R package reticulate. embeds Python interpreter inside R session, enabling high-performance interoperability languages without need user intervention.","code":""},{"path":"https://www.slendr.net/articles/vignette-00-installation.html","id":"docker","dir":"Articles","previous_headings":"","what":"Docker","title":"Installation instructions","text":"currently official Docker image slendr one soon. current plan use geospatial image published Rocker project (already contains pre-compiled R, RStudio, necessary R package dependencies sf ) extend slendr SLiM. Stay tuned!","code":""},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"motivation","dir":"Articles","previous_headings":"","what":"Motivation","title":"Introduction and basic tutorial","text":"motivation starting project create programmable simulation framework add explicit spatial dimension population genetics models. Specifically, original idea able take models one —  representing simplified view history anatomically modern humans (AMH) West Eurasia last ~50 thousand years (comprehensive overview can found review Lazaridis)—design tool makes possible simulate models explicit geographical context capture processes similar following figure (taken study Haak et al. 2015):  reason probably clear. lot studying history humans species focused reconstructing population movements, expansions gene flow events, happen geographic context. fact, geographic component often interested (.e., “ancestors population come ?”, “route fast migrate?”, etc.). However, goes beyond just simulating demographic history. instance, selection pressure driving adaptation can often spatially heterogeneous: members population occupying one part continent exposed different environmental pressure individuals elsewhere, allele frequency distributions shaped adaptation process reflect spatial heterogeneity accordingly. framework enables simulation explicitly spatial genomic data situations allow us build realistic models test specific hypotheses, goals simply possible using non-spatial simulation methods. R package slendr introduced vignette presents framework. Internally, package two independent tightly interconnected units: R interface provides set functional primitives (“mini-language” sorts) encoding various features spatio-temporal models: population migrations, expansions gene flow, happening real geographic landscape defined freely available cartographic data. Populations represented simple R objects easily visualized spatial boundaries, making possible build complex models interactively set small simple building blocks. SLiM simulation back end represented built-generic SLiM script designed read spatio-temporal model configuration parameters objects established step 1. , tailor simulation run user-defined model. Alternatively, slendr also supports executing standard population genetics models random-mating setting. means models need explicit geographic map can simulated either built-SLiM back end script, efficient msprime back end also provided package. important design objective make integration parts 1. 2. appear completely seamless. Even extremely complex models, model building execution (.e., simulation) can performed without leaving convenience R interface RStudio. simulation complexities happen automatically hood knowledge SLiM required. fact, motto slendr package “Write complex spatiotemporal population genetics models simple R script.”","code":""},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"geospatial-data","dir":"Articles","previous_headings":"","what":"Geospatial data","title":"Introduction and basic tutorial","text":"Geospatial analysis deep complex topic, dozens libraries programs designed deal fact Earth three-dimensional object forced plot geographical objects (, case, simulate data) two-dimensional plane. Luckily, technical issues Coordinate Reference Systems, transformations manipulation geometric objects (shifting population boundaries, expansions, etc.) pretty much solved now. Unfortunately, dealing issues practice quite challenging requires non-trivial degree domain expertise. Programming even simple task geospatial data analysis also often requires lot code. R package designed provide collection primitives (“mini-language” sorts) programming population dynamics (splits, movement, gene flow, expansion spatial boundaries) across space time without explicitly deal challenges inherent geospatial analyses.","code":""},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"installation-and-setup","dir":"Articles","previous_headings":"","what":"Installation and setup","title":"Introduction and basic tutorial","text":"slendr R package available CRAN can installed simply running install.packages(\"slendr\"). need run latest development version (instance, need latest bugfixes), can get via R package devtools executing devtools::install_github(\"bodkan/slendr\") R terminal. can find detailed installation instructions vignette. get slendr installed, just need load : dependencies (SLiM necessary Python modules) missing, get informative message proceed.","code":"library(slendr)  # activate the internal Python environment needed for simulation and # tree-sequence processing init_env() #> The interface to all required Python modules has been activated."},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"defining-the-overall-world-map","dir":"Articles","previous_headings":"","what":"Defining the overall world map","title":"Introduction and basic tutorial","text":"anything else, need define section map world provide context downstream spatio-temporal manipulation population ranges. principle, source geospatial data can manipulated using simple features (sf) infrastructure used. now slendr package implicitly uses Natural Earth project data (vectorized form!), internally loads using rnaturalearth interface. first slendr function look map(). function load map entire world vectorized format zoom specified section world. Note call , specify coordinates zoom geographical Coordinate Reference System (CRS), longitude/latitude, also specified want perform downstream manipulation spatial population maps projected CRS (Lambert Azimuthal Equal-Area projection) appropriate representing wider European continent used tutorial. course, different CRS projections used based part world want simulate. Describing intricacies coordinate reference systems beyond scope tutorial, ’re interested learning encourage read freely available textbook dedicated topic. approach slendr: let user specify everything easy--understand longitude/latitude geographical CRS (can read map, making easy define spatial boundaries trajectories movement), internal data structures final exported spatial maps internally handled projected CRS, important make sure distances proportions distorted. Internally, map object currently normal sf class object without additional components. unlike slendr objects described , also sf objects carry additional internal components. Note summary object says “projected CRS: ETRS89-extended / LAEA Europe”. means world map indeed transformed projected CRS specified .","code":"map <- world(   xrange = c(-13, 70), # min-max longitude   yrange = c(18, 65),  # min-max latitude   crs = \"EPSG:3035\"    # coordinate reference system (CRS) for West Eurasia ) map #> slendr 'map' object  #> -------------------  #> map: internal coordinate reference system EPSG 3035  #> spatial limits (in degrees longitude and latitude): #>   - vertical -13 ... 70 #>   - horizontal 18 ... 65"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"plotting-geographic-features-and-population-ranges","dir":"Articles","previous_headings":"","what":"Plotting geographic features and population ranges","title":"Introduction and basic tutorial","text":"slendr package implements plotting function called plot_map(). order make easier convenient iteratively build complex models. function can intelligently decide (based given input arguments) right way present data user, helps define models quickly without relying lower-level mechanisms sf package. see examples plot_map() action .","code":""},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"defining-smaller-geographic-regions","dir":"Articles","previous_headings":"","what":"Defining smaller geographic regions","title":"Introduction and basic tutorial","text":"addition overall spatial map context, can also define smaller geographic boundaries. mostly useful whenever want restrict population’s movement (spatial population expansion) smaller region map intuitive geographic meaning (.e., Anatolia, West Eurasia, etc.). Note objects created population boundaries (yet anyway)! simply labels generic geographic boundaries can used later. attached population point. , object returned region() function actually sf object, carries additional annotation name region (“Anatolia”): However, object also carries additional class annotations purpose internal slendr machinery: Furthermore, note region() calls specified map object defined beginning. object added hidden attribute slendr object represents context geospatial transformations, expansions, plots. can use generic plot_map() function plot geographic regions context defined section world map:  Note map object longer explicitly specified. needed, class objects provided plot_map() function must carry “map” attribute. fact, object must carry map context — slendr complains whenever case. can check component really , although hidden, using built-attr function verify map object created beginning:","code":"africa <- region(   \"Africa\", map,   polygon = list(c(-18, 20), c(38, 20), c(30, 33),                  c(20, 33), c(10, 38), c(-6, 35)) ) europe <- region(   \"Europe\", map,   polygon = list(     c(-8, 35), c(-5, 36), c(10, 38), c(20, 35), c(25, 35),     c(33, 45), c(20, 58), c(-5, 60), c(-15, 50)   ) ) anatolia <- region(   \"Anatolia\", map,   polygon = list(c(28, 35), c(40, 35), c(42, 40),                  c(30, 43), c(27, 40), c(25, 38)) ) anatolia #> slendr 'region' object  #> ----------------------  #> name: Anatolia  #>  #> map: internal coordinate reference system EPSG 3035 class(anatolia) #> [1] \"slendr\"        \"slendr_region\" \"sf\"            \"data.frame\" plot_map(africa, europe, anatolia, title = \"Geographic regions\") all(attr(europe, \"map\") == map) #> [1] TRUE all(attr(anatolia, \"map\") == map) #> [1] TRUE"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"defining-spatial-population-boundaries","dir":"Articles","previous_headings":"","what":"Defining spatial population boundaries","title":"Introduction and basic tutorial","text":"One aims slendr package formalize specification spatial population boundaries changes time. core function population(), takes population name, time want enforce population’s boundary, effective population size population time, map object described . also specify existing population specified population split (explicitly say ’s ancestral population). specifying actual spatial boundaries, several options.","code":""},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"polygon-population-ranges","dir":"Articles","previous_headings":"Defining spatial population boundaries","what":"Polygon population ranges","title":"Introduction and basic tutorial","text":"can define detailed population boundaries using polygon geometry object region object created region() function , using polygon = argument population(). , reminder, note coordinates described context geographic CRS. First, let’s create African ancestors modern humans. restrict spatial boundary African population africa region defined :","code":"afr <- population(\"AFR\", time = 52000, N = 3000, map = map, polygon = africa)  plot_map(afr)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"circular-population-ranges","dir":"Articles","previous_headings":"Defining spatial population boundaries","what":"Circular population ranges","title":"Introduction and basic tutorial","text":"want simulate abstract simple population boundary, can specify circular range center radius arguments instead polygon. distance units slendr package specified coordinate system given “world creation”. instance, EPSG 3035 (’re using ) specifies distances meters. define location population non-Africans right split African ancestors: call plot_map() function returned object, option either plot population range “raw” form “intersected” form, case raw boundary intersected “background” landscape (removing large bodies water, etc.). intersected form ultimately exported serialized format (see ) loaded spatial map SLiM. plot_map() function renders intersected population ranges default.","code":"ooa <- population(   \"OOA\", parent = afr, time = 51000, N = 500, remove = 25000,   center = c(33, 30), radius = 400e3 ) plot_map(ooa, intersect = TRUE, title = \"'Intersected' population range\")"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"population-movement-across-a-landscape","dir":"Articles","previous_headings":"","what":"Population movement across a landscape","title":"Introduction and basic tutorial","text":"describe directional population movement, can use function move(). accepts coordinates destination points along way (trajectory) duration migration, automatically generates number intermediate spatial maps along trajectory movement produce reasonable degree spatial continuity (number can also specified manually). can inspect object returned move() function see now contains just first YAM population range 7000 years ago, also ranges intermediate locations: Checking result visually , see:  Let’s create population Eastern Hunter Gatherers (EHG), split first non-Africans 28000 years ago: ’re , let’s also create population Western Hunter Gatherers (WHG). people living region eventually became present day Europeans receiving gene flow groups time (see ), call “EUR” simplify modeling code little bit:","code":"ooa <- ooa %>% move(   trajectory = list(c(40, 30), c(50, 30), c(60, 40)),   start = 50000, end = 40000 ) ooa #> slendr 'population' object  #> --------------------------  #> name: OOA  #> habitat: terrestrial #>  #> number of spatial maps: 28  #> map: internal coordinate reference system EPSG 3035  #> scheduled removal at time  25000  #>  #> population history overview: #>   - time 51000: split from AFR (N = 500) #>   - time 50000-40000: movement across a landscape plot_map(ooa, title = \"Intermediate migration maps\") ehg <- population(   \"EHG\", parent = ooa, time = 28000, N = 1000, remove = 6000,   polygon = list(     c(26, 55), c(38, 53), c(48, 53), c(60, 53),     c(60, 60), c(48, 63), c(38, 63), c(26, 60)) ) eur <- population( # European population   name = \"EUR\", parent = ehg, time = 25000, N = 2000,   polygon = europe )"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"spatial-population-expansion","dir":"Articles","previous_headings":"","what":"Spatial population expansion","title":"Introduction and basic tutorial","text":"can simulate expanding range population using function expand_range(), accepts parameters specifying many kilometers boundary expand (argument), long expansion take (duration argument) many intermediate spatial map snapshots exported representing expansion (snapshots argument). instance, let’s represent expansion Anatolian farmers, also split OOA population 28000 years ago time split EHG population. Note use optional parameter, polygon, restricts expansion Europe, instead around Anatolia: Note , principle, specify entire spatio-temporal history population single pipeline using pipe operator %>%. , can inspect object returned expand_range() function see contains spatial maps (“snapshots”) expansion process across time: can () check results visually:  visually see really going behind scenes, can also plot raw, non-intersected form expansion : can see population Anatolian farmers point invades spatial boundary EUR population. , doesn’t imply gene flow. section gene flow , see slendr implements gene flow overlapping (non-overlapping) populations. Let’s add couple populations migrations move implementing gene flow . Yamnaya steppe herders:","code":"ana <- population( # Anatolian farmers   name = \"ANA\", time = 28000, N = 3000, parent = ooa, remove = 4000,   center = c(34, 38), radius = 500e3, polygon = anatolia ) %>%   expand_range( # expand the range by 2.500 km     by = 2500e3, start = 10000, end = 7000,     polygon = join(europe, anatolia)   ) ana #> slendr 'population' object  #> --------------------------  #> name: ANA  #> habitat: terrestrial #>  #> number of spatial maps: 16  #> map: internal coordinate reference system EPSG 3035  #> scheduled removal at time  4000  #>  #> population history overview: #>   - time 28000: split from OOA (N = 3000) #>   - time 10000-7000: range expansion plot_map(ana, title = \"Anatolian expansion into Europe\") plot_map(ana, title = \"Anatolian expansion into Europe (not intersected)\", intersect = FALSE) yam <- population( # Yamnaya steppe population   name = \"YAM\", time = 7000, N = 500, parent = ehg, remove = 2500,   polygon = list(c(26, 50), c(38, 49), c(48, 50),                  c(48, 56), c(38, 59), c(26, 56)) ) %>%   move(     trajectory = c(15, 50),     start = 5000, end = 3000, snapshots = 8   )  plot_map(yam)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"plotting-multiple-slendr-objects","dir":"Articles","previous_headings":"","what":"Plotting multiple slendr objects","title":"Introduction and basic tutorial","text":"addition plotting individual population ranges, generic function plot_map() can handle combination population ranges, can also partition individual facets. useful visual inspection specified model, looking potential issues export individual spatio-temporal maps. Obviously, lot multi-dimensional information:  see better way explore slendr model interactively.","code":"plot_map(afr, ooa, ehg, eur, ana, yam)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"defining-gene-flow-events","dir":"Articles","previous_headings":"","what":"Defining gene flow events","title":"Introduction and basic tutorial","text":"way slendr implements gene flow events calling gene_flow() function. function straightforward interface, shown . One thing note default, populations gene flow events must overlapping spatial ranges order simulate gene flow. probably rather obvious, populations can’t mix space-time don’t overlap given point space-time. example, look spatial boundaries plotted , ’ll see European African populations don’t overlap population ranges. try instruct slendr simulate geneflow , get error: error message instructs us visually verify case, can done slendr’s plot_map() function optional parameter pop_facets = F (set TRUE default). Many models include multiple gene flow events, can collect simple R list: gene_flow() function simply returns data frame collecting geneflow parameters compile_model() step :","code":"gf <- gene_flow(from = eur, to = afr, rate = 0.1, start = 20000, end = 15000) Error: No overlap between population ranges of EUR and AFR at time 20000.    Please check the spatial maps of both populations by running   `plot_map(eur, afr)` and adjust them accordingly. Alternatively, in case   this makes sense for your model, you can add `overlap = F` which   will instruct slendr to simulate gene flow without spatial overlap   between populations. gf <- list(   gene_flow(from = ana, to = yam, rate = 0.5, start = 6500, end = 6400, overlap = FALSE),   gene_flow(from = ana, to = eur, rate = 0.5, start = 8000, end = 6000),   gene_flow(from = yam, to = eur, rate = 0.75, start = 4000, end = 3000) ) gf #> [[1]] #>   from_name to_name tstart tend rate overlap #> 1       ANA     YAM   6500 6400  0.5   FALSE #>  #> [[2]] #>   from_name to_name tstart tend rate overlap #> 1       ANA     EUR   8000 6000  0.5    TRUE #>  #> [[3]] #>   from_name to_name tstart tend rate overlap #> 1       YAM     EUR   4000 3000 0.75    TRUE"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"compile-the-whole-model-and-load-it-in-slim","dir":"Articles","previous_headings":"","what":"Compile the whole model and load it in SLiM","title":"Introduction and basic tutorial","text":"crucial function slendr compile_model(). takes population ranges defined across space time, together list gene flow events (optional, since models won’t include gene flow), proceeds converting vectorized spatial ranges raster bitmaps. Furthermore, compiles information split times, \\(N_e\\) values, gene flow directions, times, rates series tables. saved automatically dedicated directory format understood back end SLiM script provided slendr (). files model directory look like? Ideally, user never worry ; fact, whole purpose slendr let work much higher level abstraction without worrying low-level details. said, might find useful see things look like behind curtain… First , can inspect contents directory see , indeed, contain defined spatial maps (now PNG files, SLiM requires). also contains series tab-delimited configuration tables. tables contain summaries model parameters defined graphically , namely: table population splits: table geneflow events: finally, table populations whose spatial maps updated throughout simulation, well times updates (table rather large, ’re showing ). object returned compile_model() function (called model ) binds information together. fact, easier debugging sanity checking, carries locations tables (well important information) inside , elements list: model$splits, model$geneflows, etc. case ’d want separate model specification running different scripts, slendr includes function read_model() just purpose:","code":"model_dir <- paste0(tempfile(), \"_tutorial-model\")  model <- compile_model(   populations = list(afr, ooa, ehg, eur, ana, yam), # populations defined above   gene_flow = gf, # gene-flow events defined above   generation_time = 30,   resolution = 10e3, # resolution in meters per pixel   competition = 130e3, mating = 100e3, # spatial interaction in SLiM   dispersal = 70e3, # how far will offspring end up from their parents   path = model_dir ) list.files(model_dir, pattern = \"*.jpg\") #> character(0) read.table(file.path(model_dir, \"populations.tsv\"), header = TRUE) #>   pop            parent    N tsplit_gen tsplit_orig tremove_gen tremove_orig #> 1 AFR __pop_is_ancestor 3000          1       52000          -1           -1 #> 2 OOA               AFR  500         34       51000         901        25000 #> 3 EHG               OOA 1000        801       28000        1534         6000 #> 4 ANA               OOA 3000        801       28000        1601         4000 #> 5 EUR               EHG 2000        901       25000          -1           -1 #> 6 YAM               EHG  500       1501        7000        1651         2500 #>   pop_id parent_id #> 1      0        -1 #> 2      1         0 #> 3      2         1 #> 4      3         1 #> 5      4         2 #> 6      5         2 read.table(file.path(model_dir, \"geneflow.tsv\"), header = TRUE) #>   from  to rate overlap tstart_gen tstart_orig tend_gen tend_orig from_id to_id #> 1  ANA YAM 0.50       0       1518        6500     1521      6400       3     5 #> 2  ANA EUR 0.50       1       1468        8000     1534      6000       3     4 #> 3  YAM EUR 0.75       1       1601        4000     1634      3000       5     4 loaded_model <- read_model(model_dir)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"visualize-the-entire-history-of-splits-and-gene-flow","dir":"Articles","previous_headings":"","what":"Visualize the entire history of splits and gene flow","title":"Introduction and basic tutorial","text":"code snippets , defined simple history European populations last 50000 years. history includes population splits gene flow events, well demographic changes. slendr tries make formal specification spatio-temporal population dynamics concise possible, hard really visualize everything happen SLiM side simulation starts just code alone. purpose, package includes function named plot_model() takes information relationships populations (.e., population gene flow objects defined ) plots form -called admixture graph (see discussion admixture graph concept). One important thing note unlike traditional admixture graphs node/population present , full slendr graph single population can participate many gene flow events course history. visualized assigning color population, different nodes color represent snapshots time demographic event affecting population happens.","code":"plot_model(model, proportions = TRUE)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"interactive-exploration-of-spatio-temporal-models","dir":"Articles","previous_headings":"","what":"Interactive exploration of spatio-temporal models","title":"Introduction and basic tutorial","text":"slightly fancier way visualize models implemented function explore_model(). function accepts compiled model parameter spawns R shiny-based browser app makes possible click time snapshots interactively visualize spatial maps time point.","code":"explore_model(model)"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"running-the-simulation","dir":"Articles","previous_headings":"","what":"Running the simulation","title":"Introduction and basic tutorial","text":"way feed entire serialized model SLiM slim() function, understands format model directory created compile_model() function generates SLiM script (using back end skeleton script part package can found calling system.file(\"scripts/script.slim\", package = \"slendr\"), case ’d like peek internals). output slendr simulation tree-sequence file (produced SLiM simulation behind scenes). tree sequence automatically loaded R returned user: lots slendr allows , terms simulation , also terms analyzing tree-sequence data. list provides list additional resources might want look .","code":"ts <- slim(model, sequence_length = 100000, recombination_rate = 1e-8) ts #> ╔═══════════════════════╗ #> ║TreeSequence           ║ #> ╠═══════════════╤═══════╣ #> ║Trees          │    134║ #> ╟───────────────┼───────╢ #> ║Sequence Length│ 100000║ #> ╟───────────────┼───────╢ #> ║Time Units     │  ticks║ #> ╟───────────────┼───────╢ #> ║Sample Nodes   │  10000║ #> ╟───────────────┼───────╢ #> ║Total Size     │2.7 MiB║ #> ╚═══════════════╧═══════╝ #> ╔═══════════╤═════╤═════════╤════════════╗ #> ║Table      │Rows │Size     │Has Metadata║ #> ╠═══════════╪═════╪═════════╪════════════╣ #> ║Edges      │18713│584.8 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Individuals│12836│  1.2 MiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Migrations │    0│  8 Bytes│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Mutations  │    0│  1.2 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Nodes      │18351│681.7 KiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Populations│    6│  2.6 KiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Provenances│    1│ 34.7 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Sites      │    0│ 16 Bytes│          No║ #> ╚═══════════╧═════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-01-tutorial.html","id":"more-information","dir":"Articles","previous_headings":"","what":"More information","title":"Introduction and basic tutorial","text":"vignette described basic features slendr package already quite long. much slendr demonstrated . instance: can tweak parameters influencing dispersal dynamics (“clumpy” populations , far offspring can migrate parents, etc.) change dynamics evolve time. See vignette information. can use slendr program non-spatial models, means standard, Wright-Fisher demographic model can simulated lines R code , instance, plugged Approximate Bayesian Computation pipeline analyses leveraging readily available R packages. can find vignette much detailed example vignette SLiM msprime back ends. can build complex spatial models still abstract (assuming real geographic location), including traditional simulations demes lattice structure. complete example shown vignette. SLiM saves data .trees tree-sequence file format, thanks R package reticulate interfacing Python code incredible power tskit pyslim process simulated data massive scale right fingertips, within convenient environment R. See much detailed example vignette vignette extensive tutorial feature.","code":""},{"path":"https://www.slendr.net/articles/vignette-02-grid-model.html","id":"simple-two-dimensional-grid-model","dir":"Articles","previous_headings":"","what":"Simple two-dimensional grid model","title":"Demes on a regular spatial grid","text":"First, let’s load slendr R package create two-dimensional abstract world map: Next, define helper function ) create single slendr population object, b) place population appropriate coordinate lattice world map based numeric identifier population (runs 1 \\(n \\times n\\) , n total number demes along one side regular grid): defined population construction function, let’s build model. Let’s say want create regular grid n × n populations, N individuals population: Let’s plot whole spatial population configuration, make sure set things correctly:  far, way model specified, population stuck circular “island”. can change programming gene flow events using slendr function gene_flow(). , let’s first program simple helper function generate gene flow events according neighborhood relationships two-dimensional grid, allowing population exchange migrants neighbors (making sure coordinates population stay within grid using simple modulo arithmetic population index ). Let’s test function. gene flow events population lower left corner grid (, first population series)? everything works, population allowed exchange migrants neighbor right (population number 2) neighbor . Looks right! Let’s generate entire set continuous gene flow events: total number individual gene flow events : Finally, can compile whole model: familiar SLiM manual recognize model described section 5.3.3. Finally, can run simulation using slim() function.","code":"library(slendr)  init_env() #> The interface to all required Python modules has been activated. map <- world(   xrange = c(0, 1000),   yrange = c(0, 1000),   landscape = \"blank\" ) create_pop <- function(i, n_side, map, N, radius) {   # get dimensions of the world map   dim <- c(diff(attr(map, \"xrange\")), diff(attr(map, \"yrange\")))    # position of the i-th population on the two-dimensional lattice grid   coords <- c((i - 1) %% n_side, (i - 1) %/% n_side)   center <- coords / n_side * dim + dim / (2 * n_side)    pop <- tryCatch({     population(       name = sprintf(\"pop%d\", i),       N = N,       time = 1,       map = map,       center = center + c(attr(map, \"xrange\")[1], attr(map, \"yrange\")[1]),       radius = radius     )   }, error = function(e) NULL)    pop } n <- 5  populations <-   seq(1, n * n) %>%   lapply(create_pop, n_side = n, map = map, N = 100, radius = 40) do.call(plot_map, populations) + ggplot2::theme(legend.position = \"none\") set_geneflow <- function(i, n_side, rate, start, end, populations) {   pop <- populations[[i]]    # get the position of the i-th population on the n*n grid   coords <- c((i - 1) %% n_side, (i - 1) %/% n_side)     # get coordinates of the i-th population's neighbors on the grid   neighbor_pos <- list(     c(coords[1] - 1, coords[2]), c(coords[1] + 1, coords[2]),     c(coords[1], coords[2] + 1), c(coords[1], coords[2] - 1)   )    # generate geneflow events for population coordinates inside the grid   geneflows <- lapply(neighbor_pos, function(pos) {     if (any(pos < 0 | pos >= n_side)) return(NULL)     neighbor <- populations[[pos[2] * n_side + pos[1] + 1]]     if (is.null(neighbor)) return(NULL)      rbind(       gene_flow(from = pop, to = neighbor, rate = rate, start = start, end = end, overlap = FALSE),       gene_flow(from = neighbor, to = pop, rate = rate, start = start, end = end, overlap = FALSE)     )   }) %>%     do.call(rbind, .)    geneflows } set_geneflow(1, n, rate = 0.1, start = 2, end = 1000, populations) #>   from_name to_name tstart tend rate overlap #> 1      pop1    pop2      2 1000  0.1   FALSE #> 2      pop2    pop1      2 1000  0.1   FALSE #> 3      pop1    pop6      2 1000  0.1   FALSE #> 4      pop6    pop1      2 1000  0.1   FALSE geneflows <-   seq(1, n * n) %>%   lapply(set_geneflow, n, rate = 0.05, start = 2, end = 1000, populations) %>%   do.call(rbind, .) %>%   unique # filter out duplicate events due to symmetries nrow(geneflows) #> [1] 80 model <- compile_model(   populations = populations, gene_flow = geneflows,   generation_time = 1, resolution = 10,   competition = 10, mating = 10, dispersal = 10,   simulation_length = 1000 ) ts <- slim(model, sequence_length = 10000, recombination_rate = 0) # simulate a single 10kb locus ts #> ╔═══════════════════════╗ #> ║TreeSequence           ║ #> ╠═══════════════╤═══════╣ #> ║Trees          │      1║ #> ╟───────────────┼───────╢ #> ║Sequence Length│  10000║ #> ╟───────────────┼───────╢ #> ║Time Units     │  ticks║ #> ╟───────────────┼───────╢ #> ║Sample Nodes   │   5000║ #> ╟───────────────┼───────╢ #> ║Total Size     │1.4 MiB║ #> ╚═══════════════╧═══════╝ #> ╔═══════════╤════╤═════════╤════════════╗ #> ║Table      │Rows│Size     │Has Metadata║ #> ╠═══════════╪════╪═════════╪════════════╣ #> ║Edges      │9118│284.9 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Individuals│6357│622.4 KiB│         Yes║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Migrations │   0│  8 Bytes│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Mutations  │   0│  1.2 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Nodes      │9143│340.0 KiB│         Yes║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Populations│  25│  5.7 KiB│         Yes║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Provenances│   1│ 34.6 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Sites      │   0│ 16 Bytes│          No║ #> ╚═══════════╧════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-02-grid-model.html","id":"population-grid-on-a-real-geographic-landscape","dir":"Articles","previous_headings":"","what":"Population grid on a real geographic landscape","title":"Demes on a regular spatial grid","text":"can take things one step . wanted similar thing (.e. simulate regularly spaced demes) real geographic context? Let’s zoom interesting part world create grid demes using helper function create_pop defined (population boundary 300 km diameter): course, lay regular grid across map world, population boundaries fall outside African continent. solve issue, go list populations filter least 50% area land, using another helper function: Let’s plot layout population grid real geographic background:  Next, probably set scenario gene flow subpopulations; perhaps interested studying selected allele spreads continent based factors interest. , simulate data spatial model, first compile_model() run SLiM via slim() function. Given process described example , won’t repeating .","code":"map <- world(   xrange = c(-25, 55),   yrange = c(-32, 35),   crs = 4326 )  n <- 20  populations <-   seq(1, n * n) %>%   lapply(create_pop, n_side = n, map = map, N = 100, radius = 1.5) continent <- region(   map = map, polygon = list(     c(-10, 35), c(-20, 20), c(-15, 8), c(-10, 5), c(0, 2),     c(20, -40), c(35, -32), c(50, -25), c(55, -10), c(50, 0),     c(53, 13), c(45, 10), c(37, 20), c(32, 30), c(16, 38), c(0, 38)   ) )  check_area <- function(pop, map, continent) {   if (is.null(pop)) return(NULL)    # total population area   pop_area <- area(pop)$area   # population area overlapping a map   map_area <- area(overlap(pop, map))   # population area overlapping African continent   continent_area <- area(overlap(pop, continent))    # at least 50% of population's boundary be on land, and it must fall   # on to the African continent itself   if (continent_area == 0 || (map_area / pop_area) < 0.5)     return(NULL)   else     return(pop) }  filtered <- lapply(populations, check_area, map, continent) %>%   Filter(Negate(is.null), .) do.call(plot_map, filtered) + ggplot2::theme(legend.position = \"none\")"},{"path":"https://www.slendr.net/articles/vignette-02-grid-model.html","id":"more-customized-spatial-model","dir":"Articles","previous_headings":"","what":"More customized spatial model","title":"Demes on a regular spatial grid","text":"want introduce spatiality model need manual control position subpopulation, can course customize spatial layout much detail. Consider followin map South America:  Let’s lay demes scattered, irregular fashion. illustration purposes, define population’s geographic range circle radius 200 km. Importantly, note introduced one ancestral population population associated location world map! want simulate data coalescent backend, formally encode population genealogies eventually coalesce. words, models specified run msprime() back end contain isolated demes formally descend single ancestor run models msprime, get error infinite coalescent times. formally introduce ancestral population : Furthermore, let’s say fairly good idea complex interaction/gene-flow network deme, can encode like (instead forcing regular arrangement demes gene-flow interactions previous examples): Now can compile model&nbsp`twist, however: try simulate data non-spatial way using msprime, skip serialization model data disk, skipping spatial interaction dispersal parameters. words, care simulating demographic model traditional, Wright-Fisher, random-mating demes gene flow , dispersal within individual deme. ? Mostly just demonstrate possible! However, certainly situations “half-spatial” model useful – instance, situation within-deme individual dynamics interest, location subpopulation . can make sure demographic model specified correctly quickly visualizing :  can also plot spatial organization demes map glory, including gene flow events (one arrow per unidirectional gene flow). Note get warning message stating (non-spatial) ancestral population won’t visualized map:  Finally, can simulate tree sequence model! Notice simulating data using msprime(), effectively dropping continuous space dimension model. situation, map serves visual aid, making easier set complex “spatial” position demes map.","code":"xrange <- c(-90, -20) yrange <- c(-58, 15)  map <- world(xrange = xrange, yrange = yrange, crs = \"EPSG:31970\") #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpmEC5ek/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84 plot_map(map) # non-spatial ancestral population p_anc <- population(\"p_anc\", N = 1000, time = 1)  # spatial populations p1 <- population(\"p1\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-75, 0), radius = 200e3) p2 <- population(\"p2\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-60, 5), radius = 200e3) p3 <- population(\"p3\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-65, -5), radius = 200e3) p4 <- population(\"p4\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-60, -20), radius = 200e3) p5 <- population(\"p5\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-65, -35), radius = 200e3) p6 <- population(\"p6\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-69, -42), radius = 200e3) p7 <- population(\"p7\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-51, -10), radius = 200e3) p8 <- population(\"p8\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-45, -15), radius = 200e3) p9 <- population(\"p9\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-71, -12), radius = 200e3) p10 <- population(\"p10\", N = 1000, time = 500, parent = p_anc, map = map, center = c(-55, -25), radius = 200e3) gf <- list(   gene_flow(p1, p2, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p2, p1, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p1, p3, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p3, p1, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p2, p3, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p3, p2, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p2, p7, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p7, p2, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p3, p7, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p7, p3, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p7, p8, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p8, p7, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p4, p7, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p7, p4, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p4, p5, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p5, p4, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p5, p6, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p6, p5, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p3, p4, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p4, p3, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p1, p9, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p9, p1, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p3, p9, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p9, p3, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p4, p9, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p9, p4, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p10, p4, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p4, p10, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p10, p8, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p8, p10, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p10, p5, 0.1, 1000, 2000, overlap = FALSE),   gene_flow(p5, p10, 0.1, 1000, 2000, overlap = FALSE) ) model <- compile_model(   populations = list(p_anc, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10), gene_flow = gf,   generation_time = 1, simulation_length = 5000,   serialize = FALSE ) #> Warning: Model containing a mix of spatial and non-spatial populations will be compiled. #> Although this is definitely supported, make sure this is really what you want. #> Warning: Spatial models must be serialized to disk for SLiM to simulate data from. #> Compiled like this, your model can only be simulated with msprime. plot_model(model) plot_map(model, gene_flow = TRUE) #> Warning: All gene-flow event will be visualized at once. If you wish to visualize #> gene flows at a particular point in time, use the `time` argument. #> Warning: Non-spatial populations in your model won't be visualized ts <- msprime(model, sequence_length = 10e6, recombination_rate = 1e-8, random_seed = 42)  ts #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │      34057║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│   10000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │      22000║ #> ╟───────────────┼───────────╢ #> ║Total Size     │    9.4 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │182742│  5.6 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│ 11000│300.8 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │     0│ 16 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │ 74344│  2.0 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│    11│605 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     1│ 10.5 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │     0│ 16 Bytes│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"setting-up-python-environment","dir":"Articles","previous_headings":"","what":"Setting up Python environment","title":"Tree-sequence processing and statistics","text":"First, order able interface tskit pyslim using reticulate package (run simulations using msprime, ), need working Python environment required Python modules pyslim, tskit msprime already installed. setting Python environments can quite hassle, slendr provides single function setup_env() make things easier. call without arguments, slendr automatically download, install, setup completely separate Python environment (based “miniconda” distribution) just slendr activate background. important stress setup_env() interfere way Python installations might already computer. Python installation environment entirely isolated used just purpose slendr workflows. Python environment set , can activate calling: can use another built-function check_env() make sure slendr installed configured correct environment us: Now ’re good go ready simulate analyse tree sequence outputs R!","code":"setup_env() init_env() #> The interface to all required Python modules has been activated. check_env() #> Summary of the currently active Python environment: #>  #> Python binary: /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.1_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python  #> Python version: 3.12.2 | packaged by conda-forge | (main, Feb 16 2024, 20:54:21) [Clang 16.0.6 ]  #>  #> slendr requirements: #>  - tskit: version 0.5.6 ✓  #>  - msprime: version 1.3.1 ✓  #>  - pyslim: version 1.0.4 ✓  #>  - tspop: present ✓"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"model-of-neanderthal-introgression-into-eurasians","dir":"Articles","previous_headings":"Setting up Python environment","what":"Model of Neanderthal introgression into Eurasians","title":"Tree-sequence processing and statistics","text":"First, let’s set simple non-spatial model Neanderthal introgression using slendr. essentially procedure shown another vignette introducing non-spatial slendr models. different spatial model, except left map argument calling population(). ’s toy model visualized “demographic graph” sorts (.e., tree-like structure specifying population splits additional edges representing gene flow events). particularly illuminating simple example, ’s always worth keeping mind graph embedded within every slendr model can always invoked make sure model ’re setting correct:","code":"library(ggplot2) library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union set.seed(314159)  # create the ancestor of everyone and a chimpanzee outgroup # (we set both N = 1 to reduce the computational time for this model) chimp <- population(\"CH\", time = 6.5e6, N = 1000)  # two populations of anatomically modern humans: Africans and Europeans afr <- population(\"AFR\", parent = chimp, time = 6e6, N = 10000) eur <- population(\"EUR\", parent = afr, time = 70e3, N = 5000)  # Neanderthal population splitting at 600 ky ago from modern humans # (becomes extinct by 40 ky ago) nea <- population(\"NEA\", parent = afr, time = 600e3, N = 1000, remove = 40e3)  # 3% Neanderthal introgression into Europeans between 55-50 ky ago gf <- gene_flow(from = nea, to = eur, rate = 0.03, start = 55000, end = 45000)  model <- compile_model(   populations = list(chimp, nea, afr, eur), gene_flow = gf,   generation_time = 30,   path = paste0(tempfile(), \"_introgression\") ) cowplot::plot_grid(   plot_model(model, sizes = FALSE),   plot_model(model, sizes = FALSE, log = TRUE),   nrow = 1 )"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"scheduling-of-sampling-events","dir":"Articles","previous_headings":"","what":"Scheduling of sampling events","title":"Tree-sequence processing and statistics","text":"Now defined model, sample data ? Ideally, like schedule sampling events given time, sampling defined number individuals given population. slendr provides function schedule_sampling() serves define sampling schedule automatically enforces populations already (.e. appearance simulation) still (removed simulation) present sampled . example, want sample two Neanderthal individuals (older one Altai Neanderthal published Pruefer et al. 2014, younger one Vindija Neanderthal published Pruefer et al., 2017). two genomes need estimate Neanderthal ancestry proportion using -called \\(f_4\\)-ratio statistic (, also see Petr et al., PNAS 2019): can see, schedule_sampling() function simply accepts vector times remembering schedule, list pairs (<slendr population>, <number individuals>) encoding populations many individuals remembered time points given times vector. Next, want sample present-day individuals: outgroup representing chimpanzee, couple Africans Europeans: can see , schedule_sampling() function returns plain old data frame simple structure three columns: time, population name, number individuals. means can define sampling events using whatever input data might already available (radiocarbon-dated ancient DNA samples Excel sheet publication). instance, lot interest estimate trajectory Neanderthal ancestry Europe time using ancient DNA data anatomically modern human individuals (also called early modern humans, EMH) across last couple tens thousands years. can simulate something close available EMH ancient DNA data set last 50 thousand years running : samples single ancient European individuals randomly chosen times 40 10 ky ago. One nice feature schedule_sampling() function schedules sampling events population, population present simulation given time. makes possible simply take wide time range sampling, specify populations sizes samples, let function generate sampling events populations present time. reason stricter control sampling required, behavior can switched setting strict = TRUE like : Now already model object ready, can simulate data , sampling individuals according sampling schedule. Although use slim() function shown previous vignettes, case run simulation msprime() coalescent back end. , model non-spatial using coalescent simulator much efficient forward simulation. Switching msprime SLiM back ends slendr demonstrated much detail dedicated vignette. simulation back end utilized msprime() function (well slim() function) produces tree-sequence output immediately loaded ready downstream analysis. Note bind individual sampling schedule data frames using rbind function provided base R (show , sampling schedule really just data frame can manipulate ).","code":"nea_samples <- schedule_sampling(model, times = c(70000, 40000), list(nea, 1)) nea_samples #> # A tibble: 2 × 7 #>    time pop       n y_orig x_orig y     x     #>   <int> <chr> <int> <lgl>  <lgl>  <lgl> <lgl> #> 1 40000 NEA       1 NA     NA     NA    NA    #> 2 70000 NEA       1 NA     NA     NA    NA present_samples <- schedule_sampling(model, times = 0, list(chimp, 1), list(afr, 5), list(eur, 10)) present_samples #> # A tibble: 3 × 7 #>    time pop       n y_orig x_orig y     x     #>   <int> <chr> <int> <lgl>  <lgl>  <lgl> <lgl> #> 1     0 CH        1 NA     NA     NA    NA    #> 2     0 AFR       5 NA     NA     NA    NA    #> 3     0 EUR      10 NA     NA     NA    NA emh_samples <- schedule_sampling(model, times = runif(n = 40, min = 10000, max = 40000), list(eur, 1)) emh_samples #> # A tibble: 40 × 7 #>     time pop       n y_orig x_orig y     x     #>    <int> <chr> <int> <lgl>  <lgl>  <lgl> <lgl> #>  1 10320 EUR       1 NA     NA     NA    NA    #>  2 11188 EUR       1 NA     NA     NA    NA    #>  3 11396 EUR       1 NA     NA     NA    NA    #>  4 11529 EUR       1 NA     NA     NA    NA    #>  5 11927 EUR       1 NA     NA     NA    NA    #>  6 12675 EUR       1 NA     NA     NA    NA    #>  7 13689 EUR       1 NA     NA     NA    NA    #>  8 13744 EUR       1 NA     NA     NA    NA    #>  9 14775 EUR       1 NA     NA     NA    NA    #> 10 16362 EUR       1 NA     NA     NA    NA    #> # ℹ 30 more rows # this attempts to sample a Neanderthal individual at a point when Neanderthals # are already extinct, resulting in an error schedule_sampling(model, times = 10000, list(nea, 1), strict = TRUE) Error: Cannot schedule sampling for 'NEA' at time 10000 because the population will not be present in the simulation at that point. Consider running this function with `strict = FALSE` which will automatically retain only valid sampling events. ts <- msprime(   model, sequence_length = 100e6, recombination_rate = 1e-8,   samples = rbind(nea_samples, present_samples, emh_samples),   random_seed = 314159, verbose = TRUE ) #> -------------------------------------------------- #> msprime command: #>  #> /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.1_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python /private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/Rtmp1N5UNL/file10d9023ac675d_introgression/script.py --seed 314159 --model /private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/Rtmp1N5UNL/file10d9023ac675d_introgression --output /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmp1N5UNL/file10d9041197df2.trees --sequence-length 100000000 --recombination-rate 1e-08 --sampling-schedule /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmp1N5UNL/file10d906f4b6ddb --verbose   #> -------------------------------------------------- #>  #> Tree sequence was saved to: #>  /var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmp1N5UNL/file10d9041197df2.trees  #> Loading the tree-sequence file... ts #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │     240041║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        116║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   40.0 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │916588│ 28.0 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│    58│  1.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │     0│ 16 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │189654│  5.1 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     4│338 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     1│  7.0 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │     0│ 16 Bytes│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"r-interface-for-tskit-and-pyslim","dir":"Articles","previous_headings":"","what":"R interface for tskit and pyslim","title":"Tree-sequence processing and statistics","text":"Tree-sequences one revolutionary developments population genetics last couple decades number reasons. One possibility store extremely large data sets succinctly encoding entire evolutionary history sample individuals series correlated tree genealogies along genome. Going much detail topic clearly beyond scope tutorial, especially everything explain much better elsewhere. Instead, demonstrate rest vignette can access manipulate tree-sequence outputs generated slendr models perform various statistics using Python modules tskit pyslim directly slendr, without leave R! key magical R package reticulate creates seamless binding Python modules R. means even don’t know Python, slendr allows quite lot tree-sequences R. course, proficient Python user, needs said tree-sequence file generated slendr & SLiM, can easily perform every conceivable analysis directly using tskit. intention show can continue working tree-sequence files R even run entire slendr simulation.","code":""},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"loading-and-processing-tree-sequence-output-files","dir":"Articles","previous_headings":"","what":"Loading and processing tree-sequence output files","title":"Tree-sequence processing and statistics","text":"default, msprime() slim() run produce tree-sequence object (saved temporary file) immediately load . Thus, use-cases, explicit loading simulated tree sequence needed. said, sake completeness, let’s run simulation specify custom path tree-sequence file . case tree-sequence output saved custom location disk, can load tree sequence using slendr function ts_load(). ’re dealing tree sequence produced SLiM back end (case ), can also instruct function simplify tree-sequence individuals explicitly sampled (recall sampling schedule set schedule_sampling() function ). Note provide model object generated compile_model() order model annotation information simulated tree-sequence data (, loading): surprisingly, get output got printed tree sequence returned msprime() function. shows normal circumstances, loading output manually via ts_load() needed. try simplify msprime-generated tree sequence, get warning. tree sequence already simplified form, definition coming coalescent simulator. default, simplification trims tree sequence remembered individuals (.e. explicitly scheduled sampling), true every msprime tree sequence. Alternatively, can also narrow simplification defined set individuals using simplify_to = argument. Internally, simplification implemented dedicated function ts_simplify() can always call explicitly, like : Similarly, slendr provides function ts_recapitate() performs [recapitation]https://tskit.dev/pyslim/docs/latest/tutorial.html#recapitation). , needed msprime tree sequence, fully coalesced (recapitated) definition. current tree sequence object, simply get warning informing us ’re attempting something effect: can make sure tree sequence fully coalesced calling another slendr function ts_coalesced(). useful dealing slim()-produced tree sequences: might noticed simulate mutations SLiM run. computational efficiency. Luckily, tree-sequence contains complete history sample individuals makes easy sprinkle mutations genealogies simulation . can add mutations given rate running: processed simulated tree sequence, can calculate basic statistics simulated data. However, , first like note everything rest vignette (.e. whenever call function prefix ts_*() slendr), interfacing tskit Python module hood. goal capture analyses one might want perform tree-sequences R wrap neat interface indistinguishable R function—, , reason reticulate created first place (making various Python data science modules appear regular R packages).","code":"output_file <- tempfile()  ts <- msprime(   model, sequence_length = 100e6, recombination_rate = 1e-8,   samples = rbind(nea_samples, present_samples, emh_samples),   output = output_file, random_seed = 314159 )  output_file #> [1] \"/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmp1N5UNL/file10d90648929\" ts <- ts_load(output_file, model) ts #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │     240041║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        116║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   40.0 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │916588│ 28.0 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│    58│  1.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │     0│ 16 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │189654│  5.1 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     4│338 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     1│  7.0 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │     0│ 16 Bytes│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝ ts_simplify(ts) #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │     238681║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        116║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   36.7 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │864772│ 26.4 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│    58│  1.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │     0│ 16 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │139203│  3.7 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     4│338 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     2│  7.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │     0│ 16 Bytes│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝ ts_small <- ts_simplify(ts, simplify_to = c(\"CH_1\", \"NEA_1\", \"NEA_2\", \"AFR_1\", \"AFR_2\", \"EUR_20\", \"EUR_50\")) ts_small #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │     132496║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │         14║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   19.0 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │441206│ 13.5 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│     7│220 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │     0│ 16 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │ 79249│  2.1 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     4│338 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     2│  7.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │     0│ 16 Bytes│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝ ts <- ts_recapitate(ts, recombination_rate = 1e-8, Ne = 10000) ts_coalesced(ts) #> [1] TRUE ts <- ts_mutate(ts, mutation_rate = 1e-8, random_seed = 314159) ts #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │     240041║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        116║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   76.8 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │916588│ 28.0 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│    58│  1.6 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │623249│ 22.0 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │189654│  5.1 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     4│338 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     2│  7.8 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │621331│ 14.8 MiB│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"visualisation-of-trees-and-tree-sequences","dir":"Articles","previous_headings":"","what":"Visualisation of trees and tree-sequences","title":"Tree-sequence processing and statistics","text":"Now introduce function ts_phylo() can used extract one tree tree-sequence (either -th tree sequence, tree overlapping -th position simulated genome, depending value mode argument) convert phylo class, standard format phylogenetic trees R world. phylo format, see packages ape, phangorn, phytools. type tree object R console, can verify got ordinary phylo object: means whole R phylogenetic ecosystem disposal analyze trees. instance can use powerful package ggtree plot tree just extracted:","code":"# extract the 42nd tree in the tree sequence tree <- ts_phylo(ts_small, 42 - 1) #> Starting checking the validity of tree... #> Found number of tips: n = 14  #> Found number of nodes: m = 13  #> Done. tree #>  #> Phylogenetic tree with 14 tips and 13 internal nodes. #>  #> Tip labels: #>   13 (EUR_50), 12 (EUR_50), 11 (CH_1), 10 (CH_1), 9 (AFR_2), 8 (AFR_2), ... #> Node labels: #>   77386, 70, 2699, 5175, 5199, 6642, ... #>  #> Rooted; includes branch lengths. library(ggtree) #> ggtree v3.10.0 For help: https://yulab-smu.top/treedata-book/ #>  #> If you use the ggtree package suite in published research, please cite #> the appropriate paper(s): #>  #> Guangchuang Yu, David Smith, Huachen Zhu, Yi Guan, Tommy Tsan-Yuk Lam. #> ggtree: an R package for visualization and annotation of phylogenetic #> trees with their covariates and other associated data. Methods in #> Ecology and Evolution. 2017, 8(1):28-36. doi:10.1111/2041-210X.12628 #>  #> Guangchuang Yu.  Data Integration, Manipulation and Visualization of #> Phylogenetic Trees (1st edition). Chapman and Hall/CRC. 2022, #> doi:10.1201/9781003279242 #>  #> G Yu. Data Integration, Manipulation and Visualization of Phylogenetic #> Trees (1st ed.). Chapman and Hall/CRC. 2022. ISBN: 9781032233574 ggtree(tree) +   geom_point2(aes(subset = !isTip)) + # points for internal nodes   geom_tiplab() + # sample labels for tips   hexpand(0.1)    # make more space for the tip labels library(ape) plot(tree) nodelabels()"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"accessing-tskit-functionality-directly","dir":"Articles","previous_headings":"","what":"Accessing tskit functionality directly","title":"Tree-sequence processing and statistics","text":"mentioned previous section, goal vignette show use slendr perform main tree-sequence operations using convenient R interface tskit. However, always keep mind restricted subset tskit functionality slendr translated R (.e. functions prefix ts_). Thanks incredible R package reticulate, can access Python methods object variables directly, using $ operator. example, instead calling function ts_coalesced() tree-sequence , check trees coalesced running following snippet instead (note inefficient ’re operation first one hundred trees): believe makes sense use R interface whenever possible (even makes many operations little bit convenient). However, functionality slendr missing, can always resort accessing Python objects directly just demonstrated. can verify methods attributes Python tree-sequence object still accessible R: fact, recognize elements output examples involving ts_ functions vignette! short, blackbox—slendr provides slightly convenient layer tskit R users.","code":"# iterate over all trees in the tree-sequence and check if each # has only one root (i.e. is fully coalesced) - note that Python # lists are 0-based, which is something we need to take care of all(sapply(seq_len(ts$num_trees)[1:100],            function(i) ts$at_index(i - 1)$num_roots == 1)) names(ts) #>   [1] \"alignments\"                      \"allele_frequency_spectrum\"       #>   [3] \"as_fasta\"                        \"as_nexus\"                        #>   [5] \"as_vcf\"                          \"aslist\"                          #>   [7] \"at\"                              \"at_index\"                        #>   [9] \"breakpoints\"                     \"check_index\"                     #>  [11] \"coalescence_time_distribution\"   \"coiterate\"                       #>  [13] \"count_topologies\"                \"decapitate\"                      #>  [15] \"delete_intervals\"                \"delete_sites\"                    #>  [17] \"diffs\"                           \"discrete_genome\"                 #>  [19] \"discrete_time\"                   \"divergence\"                      #>  [21] \"divergence_matrix\"               \"diversity\"                       #>  [23] \"draw_svg\"                        \"draw_text\"                       #>  [25] \"dump\"                            \"dump_tables\"                     #>  [27] \"dump_text\"                       \"edge\"                            #>  [29] \"edge_diffs\"                      \"edges\"                           #>  [31] \"edges_child\"                     \"edges_left\"                      #>  [33] \"edges_parent\"                    \"edges_right\"                     #>  [35] \"edgesets\"                        \"equals\"                          #>  [37] \"f2\"                              \"f3\"                              #>  [39] \"f4\"                              \"file_uuid\"                       #>  [41] \"first\"                           \"Fst\"                             #>  [43] \"genealogical_nearest_neighbours\" \"general_stat\"                    #>  [45] \"genetic_relatedness\"             \"genetic_relatedness_weighted\"    #>  [47] \"genotype_matrix\"                 \"get_ll_tree_sequence\"            #>  [49] \"get_num_mutations\"               \"get_num_nodes\"                   #>  [51] \"get_num_records\"                 \"get_num_sites\"                   #>  [53] \"get_num_trees\"                   \"get_pairwise_diversity\"          #>  [55] \"get_population\"                  \"get_sample_size\"                 #>  [57] \"get_samples\"                     \"get_sequence_length\"             #>  [59] \"get_time\"                        \"haplotypes\"                      #>  [61] \"has_reference_sequence\"          \"ibd_segments\"                    #>  [63] \"impute_unknown_mutations_time\"   \"indexes_edge_insertion_order\"    #>  [65] \"indexes_edge_removal_order\"      \"individual\"                      #>  [67] \"individual_locations\"            \"individual_populations\"          #>  [69] \"individual_times\"                \"individuals\"                     #>  [71] \"individuals_flags\"               \"individuals_location\"            #>  [73] \"individuals_population\"          \"individuals_time\"                #>  [75] \"kc_distance\"                     \"keep_intervals\"                  #>  [77] \"last\"                            \"ll_tree_sequence\"                #>  [79] \"load\"                            \"load_tables\"                     #>  [81] \"ltrim\"                           \"max_root_time\"                   #>  [83] \"max_time\"                        \"mean_descendants\"                #>  [85] \"metadata\"                        \"metadata_schema\"                 #>  [87] \"migration\"                       \"migrations\"                      #>  [89] \"migrations_dest\"                 \"migrations_left\"                 #>  [91] \"migrations_node\"                 \"migrations_right\"                #>  [93] \"migrations_source\"               \"migrations_time\"                 #>  [95] \"min_time\"                        \"mutation\"                        #>  [97] \"mutations\"                       \"mutations_node\"                  #>  [99] \"mutations_parent\"                \"mutations_site\"                  #> [101] \"mutations_time\"                  \"nbytes\"                          #> [103] \"newick_trees\"                    \"node\"                            #> [105] \"nodes\"                           \"nodes_flags\"                     #> [107] \"nodes_individual\"                \"nodes_population\"                #> [109] \"nodes_time\"                      \"num_edges\"                       #> [111] \"num_individuals\"                 \"num_migrations\"                  #> [113] \"num_mutations\"                   \"num_nodes\"                       #> [115] \"num_populations\"                 \"num_provenances\"                 #> [117] \"num_samples\"                     \"num_sites\"                       #> [119] \"num_trees\"                       \"pairwise_diversity\"              #> [121] \"parse_windows\"                   \"population\"                      #> [123] \"populations\"                     \"provenance\"                      #> [125] \"provenances\"                     \"records\"                         #> [127] \"reference_sequence\"              \"rtrim\"                           #> [129] \"sample_count_stat\"               \"sample_size\"                     #> [131] \"samples\"                         \"segregating_sites\"               #> [133] \"sequence_length\"                 \"simplify\"                        #> [135] \"site\"                            \"sites\"                           #> [137] \"sites_position\"                  \"split_edges\"                     #> [139] \"subset\"                          \"table_metadata_schemas\"          #> [141] \"tables\"                          \"tables_dict\"                     #> [143] \"Tajimas_D\"                       \"time_units\"                      #> [145] \"to_macs\"                         \"to_nexus\"                        #> [147] \"trait_correlation\"               \"trait_covariance\"                #> [149] \"trait_linear_model\"              \"trait_regression\"                #> [151] \"trees\"                           \"trim\"                            #> [153] \"union\"                           \"variants\"                        #> [155] \"write_fasta\"                     \"write_nexus\"                     #> [157] \"write_vcf\"                       \"Y1\"                              #> [159] \"Y2\"                              \"Y3\""},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"calculating-f-statistics","dir":"Articles","previous_headings":"","what":"Calculating f-statistics","title":"Tree-sequence processing and statistics","text":"addition revolutionary breakthrough terms computation efficiency, many statistics often interested population genetics natural consequence direct access tree sequence genealogies, simply genealogies capture true demographic history sample. , can’t go much detail encourage take look paper Ralph et al. duality statistics expressed terms branch lengths traditional summaries based samples genetic variation. instance, functions ts_f2(), ts_f3(), ts_f4() ts_f4ratio() calculate well-known set Patterson’s \\(f\\)-statistics: functions accept mode = argument, specifying whether statistics calculated using mutation site patterns (mode = \"site\", default), branch lengths (mode = \"branch\"), node (mode = \"node\"), well windows argument, similarly “multiway” statistics implemented tskit. See relevant sections official tskit documentation topic. Note previous chunk referred individuals names (numeric IDs nodes tskit Python). allow readability make easier see individuals based specified sampling schedule (names assigned individuals based order sampling). can get overview individuals scheduled sampling (.e. permanently remembered) names helper function ts_samples(): said, like run statistics nodes rather individuals, can simply using integer IDs instead character names function’s interface.","code":"# f2 is a measure of the branch length connecting A and B ts_f2(ts, A = \"EUR_1\", B = \"AFR_1\") #> # A tibble: 1 × 3 #>   A     B            f2 #>   <chr> <chr>     <dbl> #> 1 EUR_1 AFR_1 0.0000448 # f4 is a measure of the drift shared between A and B after their split from C ts_f3(ts, A = \"EUR_1\", B = \"AFR_1\", C = \"CH_1\") #> # A tibble: 1 × 4 #>   A     B     C            f3 #>   <chr> <chr> <chr>     <dbl> #> 1 EUR_1 AFR_1 CH_1  0.0000130 # this value should be very close to zero (no introgression in Africans) ts_f4(ts, \"AFR_1\", \"AFR_2\", \"NEA_1\", \"CH_1\", mode = \"branch\") #> # A tibble: 1 × 5 #>   W     X     Y     Z        f4 #>   <chr> <chr> <chr> <chr> <dbl> #> 1 AFR_1 AFR_2 NEA_1 CH_1  -118. # this value should be significantly negative (many more ABBA sites # compared to BABA site due to the introgression into Europeans) ts_f4(ts, \"AFR_1\", \"EUR_1\", \"NEA_1\", \"CH_1\", mode = \"branch\") #> # A tibble: 1 × 5 #>   W     X     Y     Z        f4 #>   <chr> <chr> <chr> <chr> <dbl> #> 1 AFR_1 EUR_1 NEA_1 CH_1  -741. ts_samples(ts) #> # A tibble: 58 × 3 #>    name   time pop   #>    <chr> <dbl> <chr> #>  1 NEA_1 70000 NEA   #>  2 NEA_2 40000 NEA   #>  3 EUR_1 39171 EUR   #>  4 EUR_2 39134 EUR   #>  5 EUR_3 37739 EUR   #>  6 EUR_4 36531 EUR   #>  7 EUR_5 36362 EUR   #>  8 EUR_6 35945 EUR   #>  9 EUR_7 33900 EUR   #> 10 EUR_8 33854 EUR   #> # ℹ 48 more rows"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"estimating-neanderthal-ancestry-proportions","dir":"Articles","previous_headings":"","what":"Estimating Neanderthal ancestry proportions","title":"Tree-sequence processing and statistics","text":"Let’s try put new tools practice estimate proportion Neanderthal ancestry Africans Europeans simulated data. can using Patterson’s \\(f_4\\)-ratio statistic implemented ts_f4ratio() function slendr (can find information particular version statistic Petr et al., PNAS 2019): now summarise inferred Neanderthal distribution populations, see Neanderthal ancestry Africans (expected model–Africans receive Neanderthal introgression pulse) small proportion Neanderthal ancestry Europeans (consistent 3% introgression pulse simulated ):  exactly specified model configuration , suggesting simulations work . can see quite bit noise ’s simulated small amount sequence. can also plot trajectory Neanderthal ancestry Europe time-window simulated ancient present-day DNA samples:  , result consistent empirical estimates Neanderthal ancestry using ancient DNA data (see Petr et al., PNAS 2019).","code":"# first get a table of simulated African and European individuals in the tree-sequence inds <- ts_samples(ts) %>% dplyr::filter(pop %in% c(\"AFR\", \"EUR\"))  # estimate the amounts of Neanderthal ancestry in these individuals and add # these values to the table inds$ancestry <- ts_f4ratio(ts, X = inds$name, \"NEA_1\", \"NEA_2\", \"AFR_1\", \"CH_1\")$alpha ggplot(inds, aes(pop, ancestry, fill = pop)) +   geom_boxplot() +   geom_jitter() +   labs(y = \"Neanderthal ancestry proportion\", x = \"\") +   theme(legend.position = \"none\") +   coord_cartesian(ylim = c(0, 0.1)) dplyr::filter(inds, pop == \"EUR\") %>%   ggplot(aes(time, ancestry)) +   geom_point() +   geom_smooth(method = \"lm\", linetype = 2, color = \"red\", linewidth = 0.5) +   xlim(40000, 0) + coord_cartesian(ylim = c(0, 0.1)) +   labs(x = \"time [years ago]\", y = \"Neanderthal ancestry proportion\") #> `geom_smooth()` using formula = 'y ~ x'"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"admixtools-analyses","dir":"Articles","previous_headings":"","what":"ADMIXTOOLS analyses","title":"Tree-sequence processing and statistics","text":"case like verify f-statistics results using venerable ADMIXTOOLS software (see linked paper formally introduced statistics first place), can convert tree-sequence data file format called EIGENSTRAT using ts_eigenstrat() function. file conversion internally handled R package admixr returns EIGENSTRAT object ties individual EIGENSTRAT file components together (see tutorial admixr extensive overview). admixr R package running automated ADMIXTOOLS analyses entirely R makes types analyses convenient. Running admixr analysis easy plugging object admixr function. instance, can estimate proportion Neanderthal ancestry couple individuals \\(X\\) like (admixr calls proportion alpha): fact, lets compare values obtained tskit admixr/ADMIXTOOLS individuals:  correspondence two looks good! 🎉 , note large amount variance around expected value 3% ancestry due extremely small amount sequence data simulated .","code":"snps <- ts_eigenstrat(ts, prefix = file.path(tempdir(), \"eigenstrat\", \"data\")) #> 1282 multiallelic sites (0.206% out of 621331 total) detected and removed library(admixr)  f4ratio(data = snps, X = c(\"EUR_1\", \"EUR_2\", \"AFR_2\"),         A = \"NEA_1\", B = \"NEA_2\", C = \"AFR_1\", O = \"CH_1\") #> Rows: 3 Columns: 12 #> ── Column specification ──────────────────────────────────────────────────────── #> Delimiter: \" \" #> chr (8): X1, X2, X3, X4, X5, X6, X7, X8 #> dbl (4): X9, X10, X11, X12 #>  #> ℹ Use `spec()` to retrieve the full column specification for this data. #> ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message. #> # A tibble: 3 × 8 #>   A     B     X     C     O       alpha  stderr Zscore #>   <chr> <chr> <chr> <chr> <chr>   <dbl>   <dbl>  <dbl> #> 1 NEA_1 NEA_2 EUR_1 AFR_1 CH_1  0.0218  0.00506   4.30 #> 2 NEA_1 NEA_2 EUR_2 AFR_1 CH_1  0.0328  0.00788   4.16 #> 3 NEA_1 NEA_2 AFR_2 AFR_1 CH_1  0.00408 0.00208   1.96 europeans <- inds[inds$pop == \"EUR\", ]$name  # tskit result result_ts <- ts_f4ratio(ts, X = europeans, A = \"NEA_1\", B = \"NEA_2\", C = \"AFR_1\", O = \"CH_1\") %>% select(alpha_ts = alpha)  # result obtained by admixr/ADMIXTOOLS result_admixr <- f4ratio(snps, X = europeans, A = \"NEA_1\", B = \"NEA_2\", C = \"AFR_1\", O = \"CH_1\") %>% select(alpha_admixr = alpha) #> Rows: 50 Columns: 12 #> ── Column specification ──────────────────────────────────────────────────────── #> Delimiter: \" \" #> chr (8): X1, X2, X3, X4, X5, X6, X7, X8 #> dbl (4): X9, X10, X11, X12 #>  #> ℹ Use `spec()` to retrieve the full column specification for this data. #> ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message. bind_cols(result_admixr, result_ts) %>%   ggplot(aes(alpha_ts, alpha_admixr)) +   geom_point() +   geom_abline(slope = 1, linetype = 2, color = \"red\", linewidth = 0.5) +   labs(x = \"f4-ratio statistic calculated with admixr/ADMIXTOOLS\",        y = \"f4-ratio statistic calculated with tskit\")"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"vcf-output","dir":"Articles","previous_headings":"","what":"VCF output","title":"Tree-sequence processing and statistics","text":"case need process simulated data software, can use function ts_vcf() save simulated genotypes VCF format: can also specify subset individuals saved VCF:","code":"ts_vcf(ts, path = file.path(tempdir(), \"output.vcf.gz\")) ts_vcf(ts, path = file.path(tempdir(), \"output_subset.vcf.gz\"),        individuals = c(\"CH_1\", \"NEA_1\", \"EUR_1\", \"AFR_1\"))"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"other-statistics","dir":"Articles","previous_headings":"","what":"Other statistics","title":"Tree-sequence processing and statistics","text":"follows brief overview statistics implemented tskit slendr provides easy--use R interface. see, goal functions get result using single function call, making convenient quick interactive exploratory analyses simulated data right R console. continue use simulated Neanderthal introgression tree-sequence data examples.","code":""},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"f_st","dir":"Articles","previous_headings":"Other statistics","what":"\\(F_{st}\\)","title":"Tree-sequence processing and statistics","text":"\\(F_{st}\\) statistic implemented function ts_fst(). single genome-wide \\(F_{st}\\) calculated (.e. window-based calculation), ts_fst() returns simple three-column data frame case non-named list sample sets provided, set names generated automatically: course, much less readable encourage name sample sets appropriately. case two sample sets specified, pairwise statistics computed: many statistics implemented tskit, ts_fst() accepts windows argument, specifying breakpoints windows. case, Fst column resulting data frame called “list-column”, item column vector \\(F_{st}\\) values, one per window. List-columns can little confusing new R users, highly encourage get used allow extremely concise elegant handling structured data within normal data frames (can start introduction). instance, window-based \\(F_st\\) values afr-vs-eur calculation (first row table ):","code":"ts_fst(ts, sample_sets = list(afr = c(\"AFR_1\", \"AFR_2\", \"AFR_3\"), eur = c(\"EUR_1\", \"EUR_2\"))) #> # A tibble: 1 × 3 #>   x     y        Fst #>   <chr> <chr>  <dbl> #> 1 afr   eur   0.0495 ts_fst(ts, sample_sets = list(c(\"AFR_1\", \"AFR_2\", \"AFR_3\"), c(\"EUR_1\", \"EUR_2\"))) #> # A tibble: 1 × 3 #>   x     y        Fst #>   <chr> <chr>  <dbl> #> 1 set_1 set_2 0.0495 ts_fst(ts, sample_sets = list(afr = c(\"AFR_1\", \"AFR_2\", \"AFR_3\"),                               eur = c(\"EUR_1\", \"EUR_2\"),                               nea = c(\"NEA_1\", \"NEA_2\"))) #> # A tibble: 3 × 3 #>   x     y        Fst #>   <chr> <chr>  <dbl> #> 1 afr   eur   0.0495 #> 2 afr   nea   0.555  #> 3 eur   nea   0.541 # define breakpoints between 20 windows breakpoints <- seq(0, ts$sequence_length, length.out = 21)  # calculate window-based Fst statistic win_fst <- ts_fst(   ts, windows = breakpoints,   sample_sets = list(afr = c(\"AFR_1\", \"AFR_2\", \"AFR_3\"),                      eur = c(\"EUR_1\", \"EUR_2\"),                      nea = c(\"NEA_1\", \"NEA_2\")) )  # we get 20 values for each parwise calculation win_fst #> # A tibble: 3 × 3 #>   x     y     Fst          #>   <chr> <chr> <named list> #> 1 afr   eur   <dbl [20]>   #> 2 afr   nea   <dbl [20]>   #> 3 eur   nea   <dbl [20]> win_fst[1, ]$Fst #> $`1` #>  [1] 0.04241163 0.06803338 0.04129167 0.06031210 0.05516414 0.03059425 #>  [7] 0.03320629 0.05755971 0.05271680 0.04346970 0.03659568 0.05432308 #> [13] 0.05913699 0.05905198 0.04119386 0.04389669 0.05086942 0.06181126 #> [19] 0.03929837 0.06373172"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"tajimas-d","dir":"Articles","previous_headings":"Other statistics","what":"Tajima’s \\(D\\)","title":"Tree-sequence processing and statistics","text":"function ts_tajima() nearly interface ts_fst() shown . non-window version calculated, get single genome-wide values sample set (named non-named list character vectors individual names): window-based version, function returns D column list column vectors \\(\\)-th element Tajima’s D value \\(\\)-th window:","code":"ts_tajima(ts, list(afr = c(\"AFR_1\", \"AFR_2\", \"AFR_3\"), eur = c(\"EUR_1\", \"EUR_2\"))) #> # A tibble: 2 × 2 #>   set           D #>   <chr>     <dbl> #> 1 afr   -0.000701 #> 2 eur   -0.0140 ts_tajima(ts, list(afr = c(\"AFR_1\", \"AFR_2\"), eur = c(\"EUR_1\", \"EUR_2\")), windows = breakpoints) #> # A tibble: 2 × 2 #>   set   D            #>   <chr> <named list> #> 1 afr   <dbl [20]>   #> 2 eur   <dbl [20]>"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"diversity","dir":"Articles","previous_headings":"Other statistics","what":"Diversity","title":"Tree-sequence processing and statistics","text":"can calculate diversity within given groups individuals function ts_diversity(). instance, even extremely simplified example, expect highest levels diversity Africans, followed Europeans, Neanderthals “degenerate” single individual outgroup “chimpanzee”. true? Let’s find . First extract individuals populations, creating list character vectors group (functions ts_diversity() expects input): Now can calculate diversity population sort results increasing order diversity: Great! matches expectations. simulated chimp “population” one individual, expect essentially diversity millions years evolution.","code":"# get sampled individuals from all populations sample_sets <- ts_samples(ts) %>%   split(., .$pop) %>%   lapply(function(pop) pop$name)  sample_sets #> $AFR #> [1] \"AFR_1\" \"AFR_2\" \"AFR_3\" \"AFR_4\" \"AFR_5\" #>  #> $CH #> [1] \"CH_1\" #>  #> $EUR #>  [1] \"EUR_1\"  \"EUR_2\"  \"EUR_3\"  \"EUR_4\"  \"EUR_5\"  \"EUR_6\"  \"EUR_7\"  \"EUR_8\"  #>  [9] \"EUR_9\"  \"EUR_10\" \"EUR_11\" \"EUR_12\" \"EUR_13\" \"EUR_14\" \"EUR_15\" \"EUR_16\" #> [17] \"EUR_17\" \"EUR_18\" \"EUR_19\" \"EUR_20\" \"EUR_21\" \"EUR_22\" \"EUR_23\" \"EUR_24\" #> [25] \"EUR_25\" \"EUR_27\" \"EUR_26\" \"EUR_28\" \"EUR_29\" \"EUR_30\" \"EUR_31\" \"EUR_32\" #> [33] \"EUR_33\" \"EUR_34\" \"EUR_35\" \"EUR_36\" \"EUR_37\" \"EUR_38\" \"EUR_39\" \"EUR_40\" #> [41] \"EUR_41\" \"EUR_42\" \"EUR_43\" \"EUR_44\" \"EUR_45\" \"EUR_46\" \"EUR_47\" \"EUR_48\" #> [49] \"EUR_49\" \"EUR_50\" #>  #> $NEA #> [1] \"NEA_1\" \"NEA_2\" ts_diversity(ts, sample_sets) %>% dplyr::arrange(diversity) #> # A tibble: 4 × 2 #>   set   diversity #>   <chr>     <dbl> #> 1 CH    0.0000431 #> 2 NEA   0.0000455 #> 3 EUR   0.000394  #> 4 AFR   0.000400"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"divergence","dir":"Articles","previous_headings":"Other statistics","what":"Divergence","title":"Tree-sequence processing and statistics","text":"can calculate pairwise divergence groups individuals using function ts_divergence(). Given model, expect lowest divergence two modern human groups AFR EUR, Neanderthals two modern humans, three groups (AFR, EUR NEA) equal, much deeper divergence outgroup chimpanzee CH. sorting table based value divergence column, can see results fit expectations.","code":"ts_divergence(ts, sample_sets) %>% arrange(divergence) #> # A tibble: 6 × 3 #>   x     y     divergence #>   <chr> <chr>      <dbl> #> 1 AFR   EUR     0.000449 #> 2 EUR   NEA     0.000757 #> 3 AFR   NEA     0.000780 #> 4 CH    NEA     0.00401  #> 5 CH    EUR     0.00402  #> 6 AFR   CH      0.00402"},{"path":"https://www.slendr.net/articles/vignette-05-tree-sequences.html","id":"more-information","dir":"Articles","previous_headings":"","what":"More information","title":"Tree-sequence processing and statistics","text":"couple examples statistical functions implemented tskit provide native R interface slendr. can find tree-sequence statistics reference manual project website. statistics tskit library implemented, intend expand selection provided slendr near future. functionality like use project missing slendr, please don’t hesitate let us now creating issue GitHub page. Finally, like see examples tskit interface action, take look vignette describes switching SLiM msprime back ends slendr package.","code":""},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"introduction","dir":"Articles","previous_headings":"","what":"Introduction","title":"Spatially annotated tree sequences","text":"main selling point slendr R package programming complex spatially explicit population genetic models. use SLiM simulation engine, can store simulated data efficiently tree sequence format allows us run large population-scale simulations. previous vignettes, described can specify spatial population dynamics can access tree sequence data calculate population genetic statistics (focusing non-spatial models simplicity). Now ’s time show work simulated tree sequence spatial context.","code":""},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"model-specification","dir":"Articles","previous_headings":"","what":"Model specification","title":"Spatially annotated tree sequences","text":"Let’s first load required R libraries: begin specifying spatial model. use demographic model modern human history West Eurasia, extensively discussed introductory tutorial main landing page. complete model definition script, without comments: sanity check defined demography correctly, can plot graph summarizing population divergences geneflow events calling plot_model(model):  completeness, (slightly busy) overview spatial population ranges defined :","code":"library(slendr)  library(dplyr) library(ggplot2)  init_env()  seed <- 314159 set.seed(seed) # simulated world map map <- world(   xrange = c(-13, 70), # min-max longitude   yrange = c(18, 65),  # min-max latitude   crs = \"EPSG:3035\"    # coordinate reference system (CRS) for West Eurasia )  # couple of broad geographic regions africa <- region(   \"Africa\", map,   polygon = list(c(-18, 20), c(38, 20), c(30, 33),                  c(20, 33), c(10, 38), c(-6, 35)) ) europe <- region(   \"Europe\", map,   polygon = list(     c(-8, 35), c(-5, 36), c(10, 38), c(20, 35), c(25, 35),     c(33, 45), c(20, 58), c(-5, 60), c(-15, 50)   ) ) anatolia <- region(   \"Anatolia\", map,   polygon = list(c(28, 35), c(40, 35), c(42, 40),                  c(30, 43), c(27, 40), c(25, 38)) )  # define population histories  # African ancestral population afr <- population(   \"AFR\", time = 52000, N = 3000,   map = map, polygon = africa )  # population of the first migrants out of Africa ooa <- population(   \"OOA\", parent = afr, time = 51000, N = 500, remove = 25000,   center = c(33, 30), radius = 400e3 ) %>%   move(     trajectory = list(c(40, 30), c(50, 30), c(60, 40)),     start = 50000, end = 40000, snapshots = 20   )  # Eastern hunter-gatherers ehg <- population(   \"EHG\", parent = ooa, time = 28000, N = 1000, remove = 6000,   polygon = list(     c(26, 55), c(38, 53), c(48, 53), c(60, 53),     c(60, 60), c(48, 63), c(38, 63), c(26, 60)) )  # European population eur <- population(name = \"EUR\", parent = ehg, time = 25000, N = 2000, polygon = europe)  # Anatolian farmers ana <- population(   name = \"ANA\", time = 28000, N = 3000, parent = ooa, remove = 4000,   center = c(34, 38), radius = 500e3, polygon = anatolia ) %>%   expand_range(     by = 2500e3, start = 10000, end = 7000,     polygon = join(europe, anatolia), snapshots = 20   ) # expand the range by 2.500 km  # Yamnaya steppe population yam <- population(   name = \"YAM\", time = 7000, N = 500, parent = ehg, remove = 2500,   polygon = list(c(26, 50), c(38, 49), c(48, 50),                  c(48, 56), c(38, 59), c(26, 56)) ) %>%   move(trajectory = list(c(15, 50)), start = 5000, end = 3000, snapshots = 10)  # geneflow events gf <- list(   gene_flow(from = ana, to = yam, rate = 0.5, start = 6000, end = 5000, overlap = FALSE),   gene_flow(from = ana, to = eur, rate = 0.5, start = 8000, end = 6000),   gene_flow(from = yam, to = eur, rate = 0.75, start = 4000, end = 3000) )  # compile the spatial model model <- compile_model(   populations = list(afr, ooa, ehg, eur, ana, yam),   gene_flow = gf,   generation_time = 30, resolution = 10e3,   competition = 150e3, mating = 120e3, dispersal = 90e3 ) plot_model(model) plot_map(afr, ooa, ehg, eur, ana, yam)"},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"scheduling-sampling-events-and-simulation","dir":"Articles","previous_headings":"","what":"Scheduling sampling events and simulation","title":"Spatially annotated tree sequences","text":"Now schedule sampling single individual population every two thousand years, starting 40 thousand years ago way present (feature discussed basic tree sequence overview): Finally, can simulate data model process output tree sequence (recapitate simplify ):","code":"# one ancient individual every two thousand years ancient <- schedule_sampling(model,                     times = seq(40000, 1, by = -500),                     list(ooa, 1), list(ehg, 1), list(eur, 1),                     list(ana, 1), list(yam, 1))  # present-day Africans and Europeans present <- schedule_sampling(model, times = 0, list(afr, 5), list(eur, 30))  samples <- rbind(ancient, present) ts <- slim(   model, sequence_length = 100e3, recombination_rate = 1e-8, burnin = 200e3,   samples = samples, method = \"batch\", random_seed = 314159, max_attempts = 1 ) %>%   ts_recapitate(recombination_rate = 1e-8, Ne = 10000, random_seed = seed) %>%   ts_simplify()  ts #> ╔═════════════════════════╗ #> ║TreeSequence             ║ #> ╠═══════════════╤═════════╣ #> ║Trees          │       87║ #> ╟───────────────┼─────────╢ #> ║Sequence Length│   100000║ #> ╟───────────────┼─────────╢ #> ║Time Units     │    ticks║ #> ╟───────────────┼─────────╢ #> ║Sample Nodes   │      432║ #> ╟───────────────┼─────────╢ #> ║Total Size     │205.8 KiB║ #> ╚═══════════════╧═════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │1173│36.7 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│ 711│71.2 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │   0│ 1.2 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │ 939│35.4 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   7│ 2.7 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   3│37.8 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │   0│16 Bytes│          No║ #> ╚═══════════╧════╧════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"extracting-spatial-tree-sequence-information","dir":"Articles","previous_headings":"","what":"Extracting spatial tree sequence information","title":"Spatially annotated tree sequences","text":"showed basic tutorial, important function data exploration ts_nodes(). function extracts information individuals nodes recorded tree sequence object loaded annotated slendr : completeness, also functions ts_individuals(), ts_nodes() ts_edges() extract tree sequence tables “raw” unprocessed form, ts_nodes() much convenient data exploration analyses. First, combined information low-level tables individuals nodes single table importantly, model generated data spatial model, ts_nodes() automatically annotates node/individual tables position node space (real projected coordinates) time. means can spatial data analysis directly table returned ts_nodes(). Even better, although can see returned object belongs slendr’s class slendr_ts_nodes, internally stored spatial sf object. means can use functionality powerful R package sf well many packages geospatial analyses directly data: Typing object R console presents user-friendly summary spatio-temporal data extracted tree sequence: first part summary, see many individuals (sampled retained) nodes present tree sequence together additional useful information, including section internally stored sf object. crucial point—**can always use internal sf object spatial data directly*. data returned ts_nodes() internally transformed projected CRS used model, can use returned object data class sf. instance, beginning vignette, specified world map model represented projected CRS (EPSG 3035) can verify typing: fact ts_nodes() result just another sf object makes easy visualize overlay contents map, see .","code":"data <- ts_nodes(ts) class(data) #> [1] \"slendr\"       \"slendr_nodes\" \"sf\"           \"tbl_df\"       \"tbl\"          #> [6] \"data.frame\" data #> Simple feature collection with 939 features and 12 fields (with 11 geometries empty) #> Geometry type: POINT #> Dimension:     XY #> Bounding box:  xmin: 2426565 ymin: 257945.3 xmax: 8576618 ymax: 4878988 #> Projected CRS: ETRS89-extended / LAEA Europe #> # A tibble: 939 × 13 #>    name  pop   node_id  time time_tskit           location sampled #>    <chr> <fct>   <int> <dbl>      <dbl>        <POINT [m]> <lgl>   #>  1 AFR_1 AFR       362     0          0  (3716841 1078277) TRUE    #>  2 AFR_1 AFR       363     0          0  (3716841 1078277) TRUE    #>  3 AFR_2 AFR       364     0          0 (3541863 257945.3) TRUE    #>  4 AFR_2 AFR       365     0          0 (3541863 257945.3) TRUE    #>  5 AFR_3 AFR       366     0          0 (3366331 688429.9) TRUE    #>  6 AFR_3 AFR       367     0          0 (3366331 688429.9) TRUE    #>  7 AFR_4 AFR       368     0          0  (3329518 1360326) TRUE    #>  8 AFR_4 AFR       369     0          0  (3329518 1360326) TRUE    #>  9 AFR_5 AFR       370     0          0  (3263687 1123829) TRUE    #> 10 AFR_5 AFR       371     0          0  (3263687 1123829) TRUE    #> # ℹ 929 more rows #> # ℹ 6 more variables: remembered <lgl>, retained <lgl>, alive <lgl>, #> #   pedigree_id <dbl>, ind_id <dbl>, pop_id <int> map #> slendr 'map' object  #> -------------------  #> map: internal coordinate reference system EPSG 3035  #> spatial limits (in degrees longitude and latitude): #>   - vertical -13 ... 70 #>   - horizontal 18 ... 65"},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"using-the-simple-features-interface","dir":"Articles","previous_headings":"","what":"Using the simple features interface","title":"Spatially annotated tree sequences","text":"’s hard overstate powerful R ecosystem around sf package . However, getting familiar package geospatial analysis general can bit hurdle, especially novice users takes time get familiar many new concepts. Although many slendr features encoding programming spatial models handling simulated tree sequence data discussed far designed abstract away complexities underlying low-level details let focus problem hand, spatial data analysis unfortunately whole another matter. Luckily, data generated slendr different source spatial data great free resources disposal. bottom line : spatio-temporal data extracted tree sequences slendr different normal sf object. resource find manipulating, plotting, analysing sf data can applied slendr results well. remainder vignette look couple examples.","code":""},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"plotting-locations-of-simulated-sampled-individuals","dir":"Articles","previous_headings":"","what":"Plotting locations of simulated sampled individuals","title":"Spatially annotated tree sequences","text":"Every spatial object slendr internally class sf. flexibility ggplot2 sf packages means can overlay locations sampled individuals (saved sf format ts_nodes()) top world map (also sf object):  sf simple features objects (, extension, even slendr_spatial objects) internally stored normal data frames couple bells whistles top , powerful tools manipulating tabular data disposal. example, let’s say wanted split sampled individuals tree sequence epochs plot individually using standard ggplot2 features. simply first , adding new column specifying epoch simulated individual belong: chunk code simply adds new column epoch sf spatial data frame object called epochs . can use ggplot2 function geom_sf plot locations sampled individuals map, facet corresponding one epoch (warning can safely ignored):  hope little excursion handling slendr spatial objects (, extension, sf objects) standard data frame manipulation functions ggplot2 visualisation convinced great flexibility analysing spatial slendr data. best introduction -called “tidy” data analysis, encourage read freely-available book R Data Science.","code":"sampled_data <- ts_nodes(ts) %>% filter(sampled)  ggplot() +   geom_sf(data = map, fill = \"lightgray\", color = NA) +   geom_sf(data = sampled_data, aes(shape = pop, color = time)) +   ggtitle(\"Locations of simulated sampled individuals\") +   scale_color_continuous(type = \"viridis\") +   theme_bw() epochs <- sampled_data %>%   mutate(epoch = cut(time, breaks = c(40000, 30000, 10000, 4000, 0)),          epoch = ifelse(is.na(epoch), 0, epoch),          epoch = factor(epoch, labels = c(\"present\", \"(present, 4 ky]\", \"(4 ky, 10 ky]\",                                           \"(10 ky, 30 y]\", \"(30 ky, 40 ky]\"))) ggplot() +   geom_sf(data = map, fill = \"lightgray\", color = NA) +   geom_sf(data = epochs, aes(shape = pop, color = pop)) +   facet_wrap(~ epoch) +   ggtitle(\"Locations of simulated sampled individuals in different epochs\") +   theme_bw()"},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"extracting-spatio-temporal-ancestral-relationships","dir":"Articles","previous_headings":"","what":"Extracting spatio-temporal ancestral relationships","title":"Spatially annotated tree sequences","text":"Perhaps even useful plotting locations simulated individuals accessing locations (times) ancestors particular tree sequence node (“focal node”). Starting focal node individual, can trace geographical location nodes lineage going back way root function ts_ancestors(). record time location every individual happens ancestor least one sampled individual, means know true location every node tree sequence. simplest use case determining locations times every single node genealogical history individual along tree sequence (possible recover ancestral relationships multiple samples ): function starts given node (, name sampled diploid individual provided, two nodes), extracts information parent nodes node entire tree sequence, records locations times, proceeds one level “higher” genealogical history gather information parents parent nodes, etc., reaches root node. result process another sf object row table encodes information single branch genealogy “focal” node individual (example, \"EUR_25\"): row table, two columns location parent_location carry spatial location node (node_id) parent node (parent_id), respectively, columns time parent_time (times nodes) pop parent_pop (populations nodes belong). column connection contains sf geometry object line connecting two nodes coordinate reference system “model world”. column focal_id tells us focal node’s genealogy rows table belong , level column shows deep genealogical past branch (.e. row table) belong . table contains complete information spatio-temporal relationships nodes genealogy given focal sample. spirit demonstrating slendr tree sequence tables interact sf ggplot2 environments, let’s look immediate parent nodes two nodes sampled individual (.e. nodes level 1) using filter function R package dplyr: mentioned , three columns encoding spatial information: location parent_location carry information location child parent node (POINT class), connection object (LINESTRING class) contains line connecting two nodes (branch tree sequence also spatial connection). can plot three spatial features (two points line) individually map:  figure can see red focal node immediate parents tree sequence genealogy (coalescent sense, immediate parents individual!). case ’re surprised see two parents, recall recombination events make history encoded sample complicated can involve ancestors “move ” tree sample, just two ancestors. Looking example detail, can see one node (chromosome) individual “EUR_67” two ancestors, covering portion individuals chromosome, chromosome covered single ancestor (columns left_pos right_pos): convenient way analysis companion function ts_ancestors() called plot_ancestors(). function accepts sf object spatial branching data created ts_ancestors() plots paths nodes map leading focal node root(s) tree sequence (instead just paths immediate parents shown previous figure). case, working single diploid individual, get two sets paths nodes (chromosomes) plot two facets:  can compare result animation recapitulates simulation, presented first vignette. comparing spatial tree sequence figure animation, can immediately notice several things: spatial tree sequence paths trace ancestry single European individual back Africa. fact, also see cluster past ancestral nodes (.e. concentrated coalescent events) place Africa (OOA) migrant population settled around 40,000 thousand years ago (yellow population animation). One chromosome traces ancestry EHG population indicated green square expected programmed Yamnaya migration (descending EHG) east central Europe contribute significant part ancestry present-day Europeans (compare demographic graph top vignette). also see chromosome traces ancestry Anatolia (blue crosses). makes sense, simulated European ancestry part Anatolian. Let’s look spatial ancestry another sample. instance, know simulated history Anatolian population model much simpler. According demographic graph , Anatolians split ancestral population Eurasians Anatolia expanded wave Europe. sampled following individuals: Can see hint spatial dynamics Anatolians spatio-temporal distribution ancestral node locations one sampled individuals? Let’s pick last individual immediately plot spatial ancestry tidyverse-style using pipe operator %>%:  might expect given late age sample, position map (red crossed circle) Anatolia Europe represents one descendants migrants moved Anatolia Europe. can clearly seen position parental nodes tree sequence: nodes represent real individuals lived point past, can see , indeed, lived Anatolia.","code":"ind <- \"EUR_67\"  lineages <- ts_ancestors(ts, ind, verbose = TRUE) #> Collecting ancestors of EUR_67 [1/1]... #>  #> Generating data about spatial relationships of nodes... lineages #> Simple feature collection with 112 features and 12 fields #> Active geometry column: connection #> Geometry type: LINESTRING #> Dimension:     XY #> Bounding box:  xmin: 2736603 ymin: 607014.9 xmax: 8374222 ymax: 4762949 #> Projected CRS: ETRS89-extended / LAEA Europe #> # A tibble: 112 × 15 #>    name  pop   node_id level child_id parent_id child_time parent_time child_pop #>  * <chr> <fct>   <int> <fct>    <int>     <int>      <dbl>       <dbl> <fct>     #>  1 EUR_… EUR       406 1          406       456          0        1900 EUR       #>  2 EUR_… EUR       406 2          456       460       1900        2020 EUR       #>  3 EUR_… EUR       406 3          460       469       2020        3010 EUR       #>  4 EUR_… EUR       406 4          469       488       3010        5080 EUR       #>  5 EUR_… EUR       406 5          488       495       5080        5740 EUR       #>  6 EUR_… EUR       406 6          495       543       5740        9760 EUR       #>  7 EUR_… EUR       406 7          543       608       9760       16090 EUR       #>  8 EUR_… EUR       406 7          543       695       9760       22960 EUR       #>  9 EUR_… EUR       406 8          608       695      16090       22960 EUR       #> 10 EUR_… EUR       406 8          695       704      22960       23560 EUR       #> # ℹ 102 more rows #> # ℹ 6 more variables: parent_pop <fct>, child_location <POINT [m]>, #> #   parent_location <POINT [m]>, connection <LINESTRING [m]>, left_pos <dbl>, #> #   right_pos <dbl> filter(lineages, level == 1) #> Simple feature collection with 2 features and 12 fields #> Active geometry column: connection #> Geometry type: LINESTRING #> Dimension:     XY #> Bounding box:  xmin: 3965813 ymin: 2689630 xmax: 4683571 ymax: 2918506 #> Projected CRS: ETRS89-extended / LAEA Europe #> # A tibble: 2 × 15 #>   name   pop   node_id level child_id parent_id child_time parent_time child_pop #> * <chr>  <fct>   <int> <fct>    <int>     <int>      <dbl>       <dbl> <fct>     #> 1 EUR_67 EUR       406 1          406       456          0        1900 EUR       #> 2 EUR_67 EUR       407 1          407       451          0        1390 EUR       #> # ℹ 6 more variables: parent_pop <fct>, child_location <POINT [m]>, #> #   parent_location <POINT [m]>, connection <LINESTRING [m]>, left_pos <dbl>, #> #   right_pos <dbl> level1_branches <- ts_ancestors(ts, \"EUR_67\") %>% filter(level == 1)  ggplot() +   geom_sf(data = map, fill = \"lightgray\", color = NA) +   geom_sf(data = level1_branches[, ]$child_location, shape = 13, size = 3, color = \"red\") +   geom_sf(data = level1_branches[, ]$connection, linetype = 3) +   geom_sf(data = level1_branches[, ]$parent_location, shape = 20, color = \"blue\") +   theme_bw() +   ggtitle(\"Parent nodes (blue) of a focal individual (red)\") as_tibble(level1_branches)[, c(\"name\", \"node_id\", \"child_id\", \"parent_id\", \"left_pos\", \"right_pos\")] #> # A tibble: 2 × 6 #>   name   node_id child_id parent_id left_pos right_pos #>   <chr>    <int>    <int>     <int>    <dbl>     <dbl> #> 1 EUR_67     406      406       456        0    100000 #> 2 EUR_67     407      407       451        0    100000 ggplot() +   geom_sf(data = map) +   geom_sf(data = lineages, size = 0.5, alpha = 0.2) +   geom_sf(data = sf::st_set_geometry(lineages, \"parent_location\"),           aes(shape = parent_pop, color = parent_pop)) +   geom_sf(data = filter(ts_nodes(ts), name == ind), size = 3) +   guides(alpha = \"none\") +   coord_sf(expand = 0) +   labs(x = \"longitude\", y = \"latitude\") +   facet_grid(. ~ node_id) +   ggtitle(\"Ancestry encoded by two nodes (chromosomes) of EUR_67\") ts_samples(ts) %>% filter(pop == \"ANA\") #> # A tibble: 48 × 3 #>    name    time pop   #>    <chr>  <int> <chr> #>  1 ANA_1  27500 ANA   #>  2 ANA_2  27000 ANA   #>  3 ANA_3  26500 ANA   #>  4 ANA_4  26000 ANA   #>  5 ANA_5  25500 ANA   #>  6 ANA_6  25000 ANA   #>  7 ANA_7  24500 ANA   #>  8 ANA_8  24000 ANA   #>  9 ANA_9  23500 ANA   #> 10 ANA_10 23000 ANA   #> # ℹ 38 more rows lineages <- ts_ancestors(ts, \"ANA_45\")  ggplot() +   geom_sf(data = map) +   geom_sf(data = lineages, size = 0.5, alpha = 0.2) +   geom_sf(data = sf::st_set_geometry(lineages, \"parent_location\"),           aes(shape = parent_pop, color = parent_pop)) +   geom_sf(data = filter(ts_nodes(ts), name == \"ANA_45\"), size = 3) +   guides(alpha = \"none\") +   coord_sf(expand = 0) +   labs(x = \"longitude\", y = \"latitude\") +   facet_grid(. ~ node_id) +   ggtitle(\"Ancestry encoded by two nodes (chromosomes) of ANA_45\")"},{"path":"https://www.slendr.net/articles/vignette-06-locations.html","id":"calculating-distances-and-other-statistics-using-the-sf-package","dir":"Articles","previous_headings":"","what":"Calculating distances and other statistics using the sf package","title":"Spatially annotated tree sequences","text":"can summarise spatial ancestral dynamics figures using statistics? Lets take one look sf object locations times ancestral nodes sampled individuals, focusing following subset columns: can use standard dplyr table manipulation functions compute distances connected notes times separate (.e. branch lengths traditional phylogenetic sense). can use two quantities compute fast () movement ancestral individuals different time periods history sample: Let’s also convert data (absolute distances distance per generation – .e., “speed”) long format easier plotting side side: Let’s try summarise information distances “traveled” nodes different time period fitting spline (rather plotting raw data individual nodes):","code":"lineages <-   ts_samples(ts) %>%   pull(name) %>%   ts_ancestors(ts, x = .)  select(lineages, connection, child_time, parent_time) #> Simple feature collection with 20798 features and 2 fields #> Geometry type: LINESTRING #> Dimension:     XY #> Bounding box:  xmin: 2426565 ymin: 257945.3 xmax: 8576618 ymax: 4878988 #> Projected CRS: ETRS89-extended / LAEA Europe #> # A tibble: 20,798 × 3 #>                              connection child_time parent_time #>                        <LINESTRING [m]>      <dbl>       <dbl> #>  1   (8481422 2958176, 8289917 2751867)      40000       40690 #>  2   (8289917 2751867, 8110419 2813524)      40690       41710 #>  3    (8110419 2813524, 3348636 939123)      41710       67030 #>  4   (8110419 2813524, 2736603 1136489)      41710       83530 #>  5   (8110419 2813524, 3202720 1100178)      41710      203980 #>  6  (8110419 2813524, 3475143 607014.9)      41710      205270 #>  7  (8110419 2813524, 4086614 772146.9)      41710      243100 #>  8    (3348636 939123, 3202720 1100178)      67030      203980 #>  9  (2736603 1136489, 3969941 671890.4)      83530       88810 #> 10 (3969941 671890.4, 3991115 816590.1)      88810       96730 #> # ℹ 20,788 more rows distances <- lineages %>%   mutate(branch_length = abs(parent_time - child_time) / model$generation_time,          distance = sf::st_length(connection) %>% units::set_units(km) %>% as.numeric(),          speed = distance / branch_length,          epoch = cut(parent_time, breaks = c(Inf, seq(60000, 0, by = -3000)), dig.lab = 10, include.lowest = TRUE)) %>%   as_tibble() %>% # strip away the spatial annotation   select(name, pop, node_id, branch_length, distance, speed, parent_pop, parent_time, child_pop, child_time, epoch) distances_long <- distances %>%   filter(child_time < 60000) %>%   filter(!pop %in% c(\"AFR\", \"OOA\")) %>%   tidyr::pivot_longer(cols = c(distance, speed),                       names_to = \"stat\",                       values_to = \"value\") %>%   mutate(facet = case_when(     stat == \"distance\" ~ \"absolute distance of a node from parent\",     stat == \"speed\" ~ \"distance traveled by a node per generation\")) distances_long %>%   ggplot(aes(child_time, value, color = child_pop)) +   geom_smooth(method = \"loess\", aes(group = child_pop)) +   geom_hline(yintercept = 0, linetype = 2, linewidth = 0.5) +   labs(y = \"kilometers\", x = \"time [years ago]\") +   theme(axis.text.x = element_text(hjust = 1, angle = 45),         legend.position = \"bottom\") +   facet_wrap(~ facet, scales = \"free_y\") +   guides(color = guide_legend(\"ancestral node population\")) #> `geom_smooth()` using formula = 'y ~ x'"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"defining-a-model","dir":"Articles","previous_headings":"Detecting gene flow from msprime and SLiM tree sequences","what":"Defining a model","title":"Simulating data with SLiM and msprime backends","text":"now define simple non-spatial model gene flow populations (also known population admixture introgression). involve essentially procedure shown another vignette introducing non-spatial slendr models. Note different normally specify spatial model, except left map argument making population() calls. demonstrate additional features slendr interface tskit Python library, perform gene flow detection test using -called \\(f_4\\) \\(f_4\\)-ratio statistics (briefly introduces vignette). make things little interesting, define two population models: one model without gene flow, another includes gene flow. defining model expressing \\(f\\)-statistics use nomenclature used first study Patterson et al. described \\(f_4-ratio\\) statistic first place. ’re familiar statistical tests gene flow, recommend take look relevant sections linked paper. shiny_graph Let’s start first defining populations splits established figure : Note Ne populations x1 x2 set much higher rest. set \\(N_e\\) populations lower values speed forward SLiM simulations (won’t affect results ’re interested anyway). Higher values \\(N_e\\) x1 x2 populations ensure effect drift acting two populations much smaller. simulate later measure proportion ancestry population b x1, ensure ancestry proportion drift far away expectation make interesting patterns stand clearly. (course, done demonstration purposes speed SLiM simulations making \\(N_e\\) populations smaller. practice, running kinds simulations using msprime back end.)","code":"seq_len <- 100e6 # amount of sequence to simulate rec_rate <- 1e-8 # uniform recombination rate mut_rate <- 1e-8 # mutation rate  o <- population(\"outgroup\", time = 1, N = 100) c <- population(\"c\", time = 2500, N = 100, parent = o) a <- population(\"a\", time = 3000, N = 100, parent = c) b <- population(\"b\", time = 3500, N = 100, parent = a) x1 <- population(\"x1\", time = 3800, N = 5000, parent = c) x2 <- population(\"x2\", time = 4000, N = 5000, parent = x1)"},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"compiling-the-model-and-simulating-data","dir":"Articles","previous_headings":"","what":"Compiling the model and simulating data","title":"Simulating data with SLiM and msprime backends","text":"now use populations compile two models: first model without gene flow (left panel figure ), second model include 10% gene flow b x1 (right panel figure ). also schedule sampling couple individuals end simulation: 50 individuals populations x1 x2 capture bit natural variation b ancestry x1, one individual rest. running simulations, let’s first make sure models set correctly:   Now run models (without gene flow) two slendr backends, SLiM msprime: Note using exactly model configuration object simulation runs! fact, even function interface looks nearly exactly . doesn’t matter specific details demographic models , slendr interpret correctly regardless back end simulation engine choose use.","code":"# no gene flow model model_nogf <- compile_model(populations = list(a, b, x1, x2, c, o), generation_time = 1, simulation_length = 4500)  samples <- schedule_sampling(   model_nogf, times = 4500,   list(a, 1), list(b, 1), list(x1, 50), list(x2, 50), list(c, 1), list(o, 1) )  # model with gene flow gf <- gene_flow(from = b, to = x1, start = 4100, end = 4400, rate = 0.1)  model_gf <- compile_model(populations = list(a, b, x1, x2, c, o), gene_flow = gf, generation_time = 1, simulation_length = 4500)  samples <- schedule_sampling(   model_gf, times = 4500,   list(a, 1), list(b, 1), list(x1, 50), list(x2, 50), list(c, 1), list(o, 1) ) plot_model(model_nogf, sizes = FALSE) plot_model(model_gf, sizes = FALSE, proportions = TRUE) # model without gene flow slim_nogf <- slim(model_nogf, sequence_length = seq_len, recombination_rate = rec_rate, samples = samples, random_seed = seed) msprime_nogf <- msprime(model_nogf, sequence_length = seq_len, recombination_rate = rec_rate, samples = samples, random_seed = seed)  # model with b -> x1 gene flow slim_gf <- slim(model_gf, sequence_length = seq_len, recombination_rate = rec_rate, samples = samples, random_seed = seed) msprime_gf <- msprime(model_gf, sequence_length = seq_len, recombination_rate = rec_rate, samples = samples, random_seed = seed)"},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"comparing-the-outputs-of-msprime-and-slim-runs-of-a-slendr-model","dir":"Articles","previous_headings":"Compiling the model and simulating data","what":"Comparing the outputs of msprime and SLiM runs of a slendr model","title":"Simulating data with SLiM and msprime backends","text":"run introgression models two simulation back ends, let’s load SLiM msprime tree sequence outputs compare contents (hoping !). can see , tree sequence summary data loaded processed slendr’s ts_load() function (data produced SLiM backend) similar got “manually” msprime-produced tree sequence using custom defined functions. somehow obvious nice cheap sanity check indicating tree sequence data structure produced two backends demographic model almost .","code":"# SLiM outputs -- we can use built-in slendr functions for those slim_nogf <-   slim_nogf %>%   ts_recapitate(Ne = 10, recombination_rate = rec_rate, random_seed = seed) %>%   ts_mutate(mut_rate, random_seed = seed)  slim_gf <-   slim_gf %>%   ts_recapitate(Ne = 10, recombination_rate = rec_rate, random_seed = seed) %>%   ts_mutate(mut_rate, random_seed = seed)  # msprime outputs (note that recapitation and simplification doesn't make # sense here because we already have fully coalesced genealogies for our # individuals  of interest msprime_nogf <- ts_mutate(msprime_nogf, mut_rate, random_seed = seed) msprime_gf <- ts_mutate(msprime_gf, mut_rate, random_seed = seed) slim_nogf #> ╔═════════════════════════╗ #> ║TreeSequence             ║ #> ╠═══════════════╤═════════╣ #> ║Trees          │   248158║ #> ╟───────────────┼─────────╢ #> ║Sequence Length│100000000║ #> ╟───────────────┼─────────╢ #> ║Time Units     │    ticks║ #> ╟───────────────┼─────────╢ #> ║Sample Nodes   │    21008║ #> ╟───────────────┼─────────╢ #> ║Total Size     │ 70.6 MiB║ #> ╚═══════════════╧═════════╝ #> ╔═══════════╤══════╤════════╤════════════╗ #> ║Table      │Rows  │Size    │Has Metadata║ #> ╠═══════════╪══════╪════════╪════════════╣ #> ║Edges      │972815│29.7 MiB│          No║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Individuals│131750│12.6 MiB│         Yes║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Migrations │     0│ 8 Bytes│          No║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Mutations  │257348│ 9.1 MiB│          No║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Nodes      │157230│ 5.7 MiB│         Yes║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Populations│     7│ 2.8 KiB│         Yes║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Provenances│     3│38.0 KiB│          No║ #> ╟───────────┼──────┼────────┼────────────╢ #> ║Sites      │257040│ 6.1 MiB│          No║ #> ╚═══════════╧══════╧════════╧════════════╝ msprime_nogf #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │      62745║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│  100000000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        208║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   14.4 MiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤══════╤═════════╤════════════╗ #> ║Table      │Rows  │Size     │Has Metadata║ #> ╠═══════════╪══════╪═════════╪════════════╣ #> ║Edges      │247454│  7.6 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Individuals│   104│  2.9 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Migrations │     0│  8 Bytes│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Mutations  │ 69172│  2.4 MiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Nodes      │ 32685│893.7 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Populations│     6│414 Bytes│         Yes║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Provenances│     2│  3.9 KiB│          No║ #> ╟───────────┼──────┼─────────┼────────────╢ #> ║Sites      │ 69144│  1.6 MiB│          No║ #> ╚═══════════╧══════╧═════════╧════════════╝"},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"comparing-results-of-population-genetics-statistics","dir":"Articles","previous_headings":"Compiling the model and simulating data","what":"Comparing results of population genetics statistics","title":"Simulating data with SLiM and msprime backends","text":"loaded msprime SLiM tree sequences generated model, let’s see can get comparable results calculate statistics interest. Let’s first extract individuals populations x1 x2 estimate values \\(f_4(c, x1 \\textrm{ } x2; b, o)\\) (significantly negative individuals population x1 due ’s ancestry coming b introgression models consistent zero x2) \\(f_4\\textrm{-ratio}(, b, \\textrm{ } x2; c, o)\\), estimates proportion b-like ancestry x1 x2: Now, let’s repeat analysis tree sequences produced msprime backend two slendr demographic models: Finally can proceed plotting SLiM msprime backend results two models compare together. noted , despite fact SLiM forward simulator msprime coalescent simulator, population genetic expectations gene flow gene flow models match closely (barring level uncertainty expected due randomness population genetic processes play).   Things seem looking good behave expected! couple observations: model without gene flow, \\(f_4\\) statistic value zero x1 x2 populations. expected, \\(f_4\\) ~0 consistent relationship lineages conforming standard phylogenetic tree (admixture edges). Essentially, \\(f_4\\) hypothesis test “tree-ness” data. model includes gene flow b x1, \\(f_4\\) value significantly negative. means simple tree hypothesis can rejected gene flow must occurred b x1. Note test reveal direction gene flow, ’s presence. Consistently \\(f_4\\) results, \\(f_4-ratio\\) estimates match expectations. Specifically, statistic estimates around 10% ancestry b x1 gene flow model 0% ancestry cases. importantly, results obtained SLiM msprime back ends nearly exactly . means (non-spatial models) can use SLiM msprime backend interchangeably.","code":"# extract vector of names of the \"test individuals\" in populations `x1` and `x2` X <- ts_samples(slim_gf) %>% filter(pop %in% c(\"x1\", \"x2\")) %>% pull(name) X #>   [1] \"x1_1\"  \"x1_2\"  \"x1_3\"  \"x1_4\"  \"x1_5\"  \"x1_6\"  \"x1_7\"  \"x1_8\"  \"x1_9\"  #>  [10] \"x1_10\" \"x1_11\" \"x1_12\" \"x1_13\" \"x1_14\" \"x1_15\" \"x1_16\" \"x1_17\" \"x1_18\" #>  [19] \"x1_19\" \"x1_20\" \"x1_21\" \"x1_22\" \"x1_23\" \"x1_24\" \"x1_25\" \"x1_26\" \"x1_27\" #>  [28] \"x1_28\" \"x1_29\" \"x1_30\" \"x1_31\" \"x1_32\" \"x1_33\" \"x1_34\" \"x1_35\" \"x1_36\" #>  [37] \"x1_37\" \"x1_38\" \"x1_39\" \"x1_40\" \"x1_41\" \"x1_42\" \"x1_43\" \"x1_44\" \"x1_45\" #>  [46] \"x1_46\" \"x1_47\" \"x1_48\" \"x1_49\" \"x1_50\" \"x2_1\"  \"x2_2\"  \"x2_3\"  \"x2_4\"  #>  [55] \"x2_5\"  \"x2_6\"  \"x2_7\"  \"x2_8\"  \"x2_9\"  \"x2_10\" \"x2_11\" \"x2_12\" \"x2_13\" #>  [64] \"x2_14\" \"x2_15\" \"x2_16\" \"x2_17\" \"x2_18\" \"x2_19\" \"x2_20\" \"x2_21\" \"x2_22\" #>  [73] \"x2_23\" \"x2_24\" \"x2_25\" \"x2_26\" \"x2_27\" \"x2_28\" \"x2_29\" \"x2_30\" \"x2_31\" #>  [82] \"x2_32\" \"x2_33\" \"x2_34\" \"x2_35\" \"x2_36\" \"x2_37\" \"x2_38\" \"x2_39\" \"x2_40\" #>  [91] \"x2_41\" \"x2_42\" \"x2_43\" \"x2_44\" \"x2_45\" \"x2_46\" \"x2_47\" \"x2_48\" \"x2_49\" #> [100] \"x2_50\" # calculate f4-statistics on individuals of `x1` and `x2` populations using data # from the two models (a model with no gene flow and a gene flow model) -- we use # map_dfr to iterate across all individuals from `X_individuals` and binding all # resulting data frames into a single data frame df_slim_f4 <- rbind(   map_dfr(X, ~ ts_f4(slim_nogf, \"c_1\", .x, \"b_1\", \"outgroup_1\")) %>% mutate(model = \"no gene flow\"),   map_dfr(X, ~ ts_f4(slim_gf, \"c_1\", .x, \"b_1\", \"outgroup_1\")) %>% mutate(model = \"gene flow\") ) %>%   select(X, f4, model) %>%   mutate(simulator = \"SLiM backend\")  # compute the proportions of `b` ancestry in `x1` (expected 10%) and `x2` # (expected 0% because this population did not receive any gene flow from `b`) df_slim_f4ratio <- rbind(   ts_f4ratio(slim_nogf, X, \"a_1\", \"b_1\", \"c_1\", \"outgroup_1\") %>% mutate(model = \"no gene flow\"),   ts_f4ratio(slim_gf, X, \"a_1\", \"b_1\", \"c_1\", \"outgroup_1\") %>% mutate(model = \"gene flow\") ) %>%   select(X, alpha, model) %>%   mutate(simulator = \"SLiM backend\") df_msprime_f4 <- rbind(   map_dfr(X, ~ ts_f4(msprime_nogf, \"c_1\", .x, \"b_1\", \"outgroup_1\")) %>% mutate(model = \"no gene flow\"),   map_dfr(X, ~ ts_f4(msprime_gf, \"c_1\", .x, \"b_1\", \"outgroup_1\")) %>% mutate(model = \"gene flow\") ) %>%   select(X, f4, model) %>%   mutate(simulator = \"msprime backend\")  # compute the proportions of `b` ancestry in `x1` (expected 10%) and `x2` # (expected 0% because this population did not receive any gene flow from `b`) df_msprime_f4ratio <- rbind(   ts_f4ratio(msprime_nogf, X, \"a_1\", \"b_1\", \"c_1\", \"outgroup_1\") %>% mutate(model = \"no gene flow\"),   ts_f4ratio(msprime_gf, X, \"a_1\", \"b_1\", \"c_1\", \"outgroup_1\") %>% mutate(model = \"gene flow\") ) %>%   select(X, alpha, model) %>%   mutate(simulator = \"msprime backend\") df_f4 <- rbind(df_slim_f4, df_msprime_f4) %>%   mutate(population = ifelse(grepl(\"x1_\", X),                              \"x1 (received gene flow)\",                              \"x2 (no gene flow)\"))  ggplot(df_f4, aes(f4, fill = population)) +   geom_histogram(bins = 50) +   facet_grid(simulator ~ model) +   geom_vline(xintercept = 0, linetype = 2) +   labs(y = \"number of individuals\", x = \"f4 statistic\",        title = \"f4(c, x1 or x2; b, outgroup)\",        subtitle = \"f4 ~0 is consistent with no gene flow, negative value indicates gene flow with 'b'\") +   theme(legend.position = \"bottom\") df_f4ratio <- rbind(df_slim_f4ratio, df_msprime_f4ratio) %>%   mutate(population = ifelse(grepl(\"x1_\", X),                              \"x1 (received gene flow)\",                              \"x2 (no gene flow)\"))  ggplot(df_f4ratio, aes(alpha, fill = population)) +   geom_histogram(bins = 30) +   facet_grid(simulator ~ model) +   geom_vline(xintercept = 0.1, linetype = 2) +   labs(y = \"number of individuals\", x = \"ancestry proportion (f4-ratio statistic)\",        title = \"f4-ratio estimate of 'b' ancestry calculated from simulated data\",        subtitle = \"f4-ratio = f4(a, outgroup; x1 or x2, c) / f4(a, outgroup; b, c)\") +   theme(legend.position = \"bottom\")"},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"computing-allele-frequency-spectra","dir":"Articles","previous_headings":"","what":"Computing allele frequency spectra","title":"Simulating data with SLiM and msprime backends","text":"Let’s look one example. Imagine following five demographic models describing changes population size abstract population:  models population size change events , depending specifics project, defined using different time units. work radiocarbon-dated samples ancient DNA might used thinking population history units “years present”—case, “present” time (bottom figure) time 0 earlier events specified “5000 years present”, etc. hand, interested modelling theoretical population, might want concern quite happy starting simulation “generation 1” continue specified “generation X”. slendr package makes extremely easy, can specify times whatever direction units want, long kept consistent populations events model. rather unique feature among popgen simulation software programs either make express model “generations forward direction”, explicitly convert time generations going backwards. previous section vignette shown SLiM msprime can used simulation backend. remainder vignette present additional analysis, demonstrating fact can express time models forward backward direction still run models slendr’s SLiM (forward simulator) msprime (backward simulator). Specifically, define five demographic models figure twice, different orientations time – forward backward. compute allele frequency spectrum simulation (five models, forward backward) verify forward backward pairs give result regardless whether use SLiM back end msprime backend.","code":""},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"model-definition","dir":"Articles","previous_headings":"Computing allele frequency spectra","what":"Model definition","title":"Simulating data with SLiM and msprime backends","text":"First, let’s define five population histories forward time direction: Now let’s flip flow time around define population histories units “time ago”: convenience, write helper function , given population history object (ten objects just ), compile slendr model, run SLiM msprime back ends, load tree sequence, compute allele frequency spectrum (AFS) . save us lots code repetition (always good thing), minimize chance copy-pasting bugs (definitely good thing!) compare AFS patterns across different models two slendr simulation back ends (according population genetics theory). Now function doesn’t anything special. simply wraps standard slendr simulation pipeline (1. create population(s); 2. compile model object; 3. simulate model; 4. load results; 5. analyse data) single function. Now whole pipeline, can generate simulated data bind result single data frame: Finally, can plot allele frequency spectra models, two time directions, two simulation back ends:  , couple things note: two models involve population contraction (single step exponential collapse), see decrease low frequency variants. hand, models population expansion show high proportion low frequency variants. model constant population size shows intermediate pattern. nice validation simulation works expected exactly predicted population genetics theory. forward backward time models give exactly result. unexpected anything else bug slendr. However, although unsurprising, hope makes clear free use whatever time unit specification want encode slendr models. non-spatial models, results obtained msprime SLiM back ends consistent one another. , makes sense although completely different software, population genetic theory governing shape allele frequency spectra applies equally . fact, example just implemented important part slendr unit test suite verifies simulations created slendr correct.","code":"N <- 1000 N_factor <- 5 # by what factor should Ne change  seq_len <- 50e6 rec_rate <- 1e-8 mut_rate <- 1e-8 # constant Ne model forward_const <- population(\"const\", time = 1, N = N)  # decreasing step Ne model forward_decr <- population(\"decr\", time = 1, N = N, map = FALSE) %>%   resize(time = 2000, N = N / N_factor, how = \"step\")  # increasing step Ne model forward_incr <- population(\"inc\", time = 1, N = N) %>%   resize(time = 2000, N = N * N_factor, how = \"step\")  # exponential increase in size forward_exp_incr <- population(\"exp_inc\", time = 1, N = N) %>%   resize(time = 2000, end = 3000, N = N * N_factor, how = \"exponential\")  # exponential decrease in size forward_exp_decr <- population(\"exp_decr\", time = 1, N = N) %>%   resize(time = 2000, end = 3000, N = N / N_factor, how = \"exponential\") # constant Ne model backward_const <- population(\"const\", time = 5000, N = N)  # decreasing step Ne model backward_decr <- population(\"decr\", time = 5000, N = N) %>%   resize(time = 3000, N = N / N_factor, how = \"step\")  # increasing step Ne model backward_incr <- population(\"inc\", time = 5000, N = N) %>%   resize(time = 3000, N = N * N_factor, how = \"step\")  # exponential increase in size backward_exp_incr <- population(\"exp_inc\", time = 5000, N = N) %>%   resize(time = 3000, end = 2000, N = N * N_factor, how = \"exponential\")  # exponential decrease in size backward_exp_decr <- population(\"exp_decr\", time = 5000, N = N) %>%   resize(time = 3000, end = 2000, N = N / N_factor, how = \"exponential\") compile_run_afs <- function(model_name, pop, seed = 42) {   # maximum length of the simulation (necessary for forward models which start   # in generation 1)   simulation_length <- 5000    # define sampling times given the direction of time   if (attr(pop, \"history\")[[1]]$time == 1) {     sampling_time <- simulation_length     direction <- \"forward\"   } else {     sampling_time <- 0     direction <- \"backward\"   }    # compile model   model <- compile_model(pop, generation_time = 15, direction = direction, simulation_length = simulation_length)    samples <- schedule_sampling(model, times = sampling_time, list(pop, 50))    # run the model in SLiM   ts_slim <- slim(model, sequence_length = seq_len, recombination_rate = rec_rate,                   samples = samples, random_seed = seed, verbose = FALSE)    # run the same model in msprim   ts_msprime <- msprime(model, sequence_length = seq_len, recombination_rate = rec_rate,                         samples = samples, random_seed = seed, verbose = FALSE)    # load the SLiM tree sequence   ts_slim <- ts_recapitate(ts_slim, Ne = N, recombination_rate = rec_rate, random_seed = seed) %>%     ts_mutate(mut_rate, random_seed = seed)    # load the msprime tree sequence   ts_msprime <- ts_mutate(ts_msprime, mut_rate, random_seed = seed)    # compute the AFS from the SLiM and msprime tree sequences and bind the   # results (derived allele counts per frequency bin) in a data frame   msprime_afs <- ts_afs(ts_msprime, polarised = TRUE)[-1]   slim_afs <- ts_afs(ts_slim, polarised = TRUE)[-1]    rbind(     data.frame(simulator = \"msprime\", model = model_name, f = msprime_afs),     data.frame(simulator = \"SLiM\", model = model_name, f = slim_afs)   ) %>%     group_by(simulator, model) %>%     mutate(n = 1:n(), direction = direction) %>%     ungroup() } afs <- bind_rows(   compile_run_afs(\"constant\", forward_const),   compile_run_afs(\"constant\", backward_const),   compile_run_afs(\"step contraction\", forward_decr),   compile_run_afs(\"step contraction\", backward_decr),   compile_run_afs(\"step increase\", forward_incr),   compile_run_afs(\"step increase\", backward_incr),   compile_run_afs(\"exponential decrease\", forward_exp_decr),   compile_run_afs(\"exponential decrease\", backward_exp_decr),   compile_run_afs(\"exponential increase\", forward_exp_incr),   compile_run_afs(\"exponential increase\", backward_exp_incr) ) %>%   mutate(model = factor(     model, levels = c(\"step contraction\", \"constant\", \"step increase\",                       \"exponential decrease\", \"exponential increase\"))   ) ggplot(afs, aes(n, f, color = direction, linetype = simulator)) +   geom_line(stat = \"identity\") +   facet_wrap(~ model) +   labs(x = \"number of derived alleles\", y = \"frequency\",        title = \"Site frequency spectra obtained from five demographic models\",        subtitle = \"Each model was specified in forward or backward direction of time and executed by two different backend scripts in slendr (SLiM and msprime)\") +   guides(color = guide_legend(\"direction of\\ntime in slendr\"),          linetype = guide_legend(\"slendr backend\\nengine used\")) +   scale_linetype_manual(values = c(3, 2)) +   scale_x_continuous(breaks = c(1, seq(20, 100, 20)), limits = c(1, 100)) +   theme(legend.position = \"bottom\")"},{"path":"https://www.slendr.net/articles/vignette-07-backends.html","id":"conclusion","dir":"Articles","previous_headings":"","what":"Conclusion","title":"Simulating data with SLiM and msprime backends","text":"example shows standard, non-spatial demographic model like implement simulate R, slendr gives way efficiently using ’s msprime back end addition SLiM back end used throughout documentation.","code":""},{"path":"https://www.slendr.net/articles/vignette-08-nonslendr-tskit.html","id":"non-spatial-slim-tree-sequences","dir":"Articles","previous_headings":"","what":"Non-spatial SLiM tree sequences","title":"Analyzing non-slendr tree sequences","text":"Consider following SLiM script, creates couple populations (different \\(N_e\\)) splitting ancestral population p1 (lets save /tmp/nonspatial.slim): run script SLiM, can use slendr load output tree sequence (saved /tmp/nonspatial-slim.trees), simplify , overlay mutations using standard functionality originally developed slendr tree sequences. Note command use loading slendr tree sequences, except direct ts_load() function straight tree-sequence output file rather using ts_load(<model-object>) format used working standard slendr simulations. way, slendr can extract information individual’s names, nodes, population assignments, etc. just slendr tree sequence function ts_nodes(). standard slendr models, function loads “raw” node individual tree-sequences tables, performs couple join operations, presents whole thing nice unified form interactive data analysis (can also include spatial information—see ): Moving tskit statistics, can use data table extract list nodes belonging population (various tskit tree-sequence statistics operate , slendr follows design). computing nucleotide diversity four populations using ts_diversity() function, first creating list lists node IDs (.e. chromosomes) individuals population: Just slendr tree sequences (demonstrated paper) can get individual trees , extracted phylogenetic format provided ape R package. first simplify tree sequence even just 10 nodes make things manageable: R tree object, can use packages like ggtree visualize tree (phylogenetic package work ). Note nodes ‘ape phylo’ trees must conform strict format (must labelled 1...N), extract information node IDs tskit tree-sequence data able plot tree.","code":"initialize() {     setSeed(42);     initializeTreeSeq();     initializeMutationRate(0);     initializeMutationType(\"m1\", 0.5, \"f\", 0.0);     initializeGenomicElementType(\"g1\", m1, 1.0);     initializeGenomicElement(g1, 0, 1e6);     initializeRecombinationRate(1e-8); }  1 early() {     sim.addSubpop(\"p1\", 10); }  1000 early() {     sim.addSubpopSplit(\"p2\", 500, p1); }  3000 early() {     sim.addSubpopSplit(\"p3\", 2500, p1); }  5000 early() {     sim.addSubpopSplit(\"p4\", 10000, p1); }  6000 late() {     sim.treeSeqOutput(\"/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmpxdy6r9/file1112e7d444348\"); } ts <- ts_load(nonspatial_trees_file) %>%   ts_simplify() %>%   ts_mutate(mutation_rate = 1e-7, random_seed = SEED) data <- ts_nodes(ts) %>% dplyr::filter(sampled) data #> # A tibble: 26,020 × 11 #>    pop   node_id  time time_tskit sampled remembered retained alive pedigree_id #>    <chr>   <int> <dbl>      <dbl> <lgl>   <lgl>      <lgl>    <lgl>       <dbl> #>  1 p1          0     0          0 TRUE    FALSE      FALSE    TRUE     20073000 #>  2 p1          1     0          0 TRUE    FALSE      FALSE    TRUE     20073000 #>  3 p1          2     0          0 TRUE    FALSE      FALSE    TRUE     20073001 #>  4 p1          3     0          0 TRUE    FALSE      FALSE    TRUE     20073001 #>  5 p1          4     0          0 TRUE    FALSE      FALSE    TRUE     20073002 #>  6 p1          5     0          0 TRUE    FALSE      FALSE    TRUE     20073002 #>  7 p1          6     0          0 TRUE    FALSE      FALSE    TRUE     20073003 #>  8 p1          7     0          0 TRUE    FALSE      FALSE    TRUE     20073003 #>  9 p1          8     0          0 TRUE    FALSE      FALSE    TRUE     20073004 #> 10 p1          9     0          0 TRUE    FALSE      FALSE    TRUE     20073004 #> # ℹ 26,010 more rows #> # ℹ 2 more variables: ind_id <dbl>, pop_id <int> sample_sets <- split(data$node_id, data$pop)  # compute nucleotide diversity in each population # (any other ts_*() tskit R interface function should work) ts_diversity(ts, sample_sets) #> # A tibble: 4 × 2 #>   set    diversity #>   <chr>      <dbl> #> 1 p1    0.00000755 #> 2 p2    0.000241   #> 3 p3    0.000463   #> 4 p4    0.000201 samples <- sample(data$node_id, 10) ts_small <- ts_simplify(ts, simplify_to = samples)  # extract the 42nd tree in the genealogy to an R 'phylo' format tree <- ts_phylo(ts_small, 42 - 1) #> Starting checking the validity of tree... #> Found number of tips: n = 10  #> Found number of nodes: m = 9  #> Done. tree #>  #> Phylogenetic tree with 10 tips and 9 internal nodes. #>  #> Tip labels: #>   1, 0, 9, 4, 8, 7, ... #> Node labels: #>   45, 28, 22, 31, 35, 36, ... #>  #> Rooted; includes branch lengths. library(ggtree) #> ggtree v3.10.0 For help: https://yulab-smu.top/treedata-book/ #>  #> If you use the ggtree package suite in published research, please cite #> the appropriate paper(s): #>  #> Guangchuang Yu, David Smith, Huachen Zhu, Yi Guan, Tommy Tsan-Yuk Lam. #> ggtree: an R package for visualization and annotation of phylogenetic #> trees with their covariates and other associated data. Methods in #> Ecology and Evolution. 2017, 8(1):28-36. doi:10.1111/2041-210X.12628 #>  #> Guangchuang Yu.  Data Integration, Manipulation and Visualization of #> Phylogenetic Trees (1st edition). Chapman and Hall/CRC. 2022, #> doi:10.1201/9781003279242 #>  #> LG Wang, TTY Lam, S Xu, Z Dai, L Zhou, T Feng, P Guo, CW Dunn, BR #> Jones, T Bradley, H Zhu, Y Guan, Y Jiang, G Yu. treeio: an R package #> for phylogenetic tree input and output with richly annotated and #> associated data. Molecular Biology and Evolution. 2020, 37(2):599-603. #> doi: 10.1093/molbev/msz240  #>  #> Attaching package: 'ggtree' #> The following object is masked from 'package:magrittr': #>  #>     inset  labels <- ts_nodes(tree) %>% select(node = phylo_id, tskit_id = node_id)  ggtree(tree, branch.length = \"none\") %<+% labels +   geom_label(aes(label = tskit_id)) library(ape) plot(tree, show.tip.label = FALSE) nodelabels() tiplabels()"},{"path":"https://www.slendr.net/articles/vignette-08-nonslendr-tskit.html","id":"msprime-non-slendr-tree-sequences","dir":"Articles","previous_headings":"","what":"msprime (non-slendr) tree sequences","title":"Analyzing non-slendr tree sequences","text":"applies also msprime tree sequences (really surprising, given ’s tskit hood). can start Python: can proceed loading msprime tree sequence R analyze slendr functionality:","code":"import msprime ts = msprime.sim_ancestry(100) ts.dump(<trees file>) ts <- ts_load(msprime_trees_file)  ts_nodes(ts) #> # A tibble: 199 × 7 #>    pop   ind_id node_id  time time_tskit sampled pop_id #>    <chr>  <dbl>   <int> <dbl>      <dbl> <lgl>    <int> #>  1 0         NA       0     0          0 TRUE         0 #>  2 0         NA       1     0          0 TRUE         0 #>  3 0         NA       2     0          0 TRUE         0 #>  4 0         NA       3     0          0 TRUE         0 #>  5 0         NA       4     0          0 TRUE         0 #>  6 0         NA       5     0          0 TRUE         0 #>  7 0         NA       6     0          0 TRUE         0 #>  8 0         NA       7     0          0 TRUE         0 #>  9 0         NA       8     0          0 TRUE         0 #> 10 0         NA       9     0          0 TRUE         0 #> # ℹ 189 more rows"},{"path":"https://www.slendr.net/articles/vignette-08-nonslendr-tskit.html","id":"spatial-slim-non-slendr-tree-sequences","dir":"Articles","previous_headings":"","what":"Spatial SLiM (non-slendr) tree sequences","title":"Analyzing non-slendr tree sequences","text":"Furthermore, generalized interface also supports slendr’s spatial tree-sequence features, bells whistles. instance, lets take following spatial SLiM script (modified SLiM manual) execute SLiM usual way: can load simplify output tree sequence just vignette (anywhere slendr documentation): Finally, can access spatio-temporal data embedded output tree sequence standard slendr way (note spatial sf column location POINT data type): get tree sequence converted spatial sf data format, can use standard geospatial packages use spatial data analysis methods packages provide. briefly demonstrate means, can trivially plot location recorded node:  can also collect spatio-temporal ancestry information particular node (.e. times locations ancestors way root, “link” plot signifying parent-child edge somewhere along tree sequence) plot 2D surface (x y dimensions [0, 1]). plot little chaotic, hopefully conveys idea (“focal node” 0 highlighted red). essentially plot last figure paper.","code":"initialize() {     setSeed(42);       initializeSLiMOptions(keepPedigrees=T, dimensionality=\"xy\");       initializeTreeSeq();       initializeMutationRate(1e-7);       initializeMutationType(\"m1\", 0.5, \"f\", 0.0);       initializeGenomicElementType(\"g1\", m1, 1.0);       initializeGenomicElement(g1, 0, 1e6);       initializeRecombinationRate(1e-8); } 1 early() {     sim.addSubpop(\"p1\", 500);      // initial positions are random in ([0,1], [0,1])     p1.individuals.x = runif(p1.individualCount);       p1.individuals.y = runif(p1.individualCount); } modifyChild() {       // draw a child position near the first parent, within bounds       do child.x = parent1.x + rnorm(1, 0, 0.02);       while ((child.x < 0.0) | (child.x > 1.0));        do child.y = parent1.y + rnorm(1, 0, 0.02);       while ((child.y < 0.0) | (child.y > 1.0));        return T; } 1: late() {     sim.treeSeqRememberIndividuals(sim.subpopulations.individuals, permanent = F); }  10000 late() {     sim.treeSeqOutput(\"/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T//Rtmpxdy6r9/file1112e332bd700\"); } ts <- ts_load(spatial_trees_file) %>% ts_simplify() data <- ts_nodes(ts) data #> Simple feature collection with 1941 features and 11 fields #> Geometry type: POINT #> Dimension:     XY #> Bounding box:  xmin: 0.02032986 ymin: 0.001180405 xmax: 0.9990375 ymax: 0.9979969 #> CRS:           NA #> # A tibble: 1,941 × 12 #>    pop   node_id  time time_tskit               location sampled remembered #>    <chr>   <int> <dbl>      <dbl>                <POINT> <lgl>   <lgl>      #>  1 p1          0     0          0  (0.3158008 0.8489413) TRUE    FALSE      #>  2 p1          1     0          0  (0.3158008 0.8489413) TRUE    FALSE      #>  3 p1          2     0          0  (0.5946915 0.5792152) TRUE    FALSE      #>  4 p1          3     0          0  (0.5946915 0.5792152) TRUE    FALSE      #>  5 p1          4     0          0  (0.8320243 0.5595862) TRUE    FALSE      #>  6 p1          5     0          0  (0.8320243 0.5595862) TRUE    FALSE      #>  7 p1          6     0          0  (0.5532792 0.1833623) TRUE    FALSE      #>  8 p1          7     0          0  (0.5532792 0.1833623) TRUE    FALSE      #>  9 p1          8     0          0 (0.7333698 0.01834573) TRUE    FALSE      #> 10 p1          9     0          0 (0.7333698 0.01834573) TRUE    FALSE      #> # ℹ 1,931 more rows #> # ℹ 5 more variables: retained <lgl>, alive <lgl>, pedigree_id <dbl>, #> #   ind_id <dbl>, pop_id <int> ggplot() + geom_sf(data = data, aes(color = time), alpha = 0.5) ancestral_links <- ts_ancestors(ts, 0)  ggplot() +     geom_sf(data = ancestral_links, size = 0.5, aes(alpha = parent_time)) +     geom_sf(data = sf::st_set_geometry(ancestral_links, \"parent_location\"), aes(color = parent_time)) +     geom_sf(data = data[data$node_id == 0, ], size = 3, color = \"red\")"},{"path":"https://www.slendr.net/articles/vignette-08-nonslendr-tskit.html","id":"conclusion","dir":"Articles","previous_headings":"","what":"Conclusion","title":"Analyzing non-slendr tree sequences","text":"vignette gave brief overview using slendr’s R-tskit interface loading, processing, analyzing “pure” non-slendr tree sequences produced msprime SLiM scripts. Although touched upon basic features R-tskit interface standard tree sequences, important note far slendr concerned, matter tree sequence produced, long conforms tskit specification. means regardless source tree sequence data, able use slendr’s tskit functionality run analyses.","code":""},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"note","dir":"Articles","previous_headings":"","what":"Note","title":"Examples from the slendr paper","text":"Since publication slendr paper, new versions SLiM slendr released. Specifically, results paper based SLiM version 4.0.1 slendr version 0.5. Whenever new version SLiM released, ’s always chance aspect randomness simulation changes slightly, affects results even though random seed kept value. , figures paper (specifically figure Example 4 paper) look slightly different. However, results still relevant applicable, just presentation changed.","code":"# workaround for GitHub actions getting \"Killed\" due to out of memory issues # -- when running unit tests on GitHub, smaller amount of sequence will be # simulated if (Sys.getenv(\"RUNNER_OS\") != \"\") {   scaling <- 10 # use 10X less sequence on GitHub actions } else {   scaling <- 1 # but simulate normal amount of data otherwise }  library(slendr)  library(dplyr) #>  #> Attaching package: 'dplyr' #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union library(ggplot2) library(purrr) library(tidyr) library(cowplot) library(forcats)  init_env() #> The interface to all required Python modules has been activated.  SEED <- 42 set.seed(SEED) # placeholder \"figure\" where a code chunk will be pasted p_code <- ggplot() +   geom_text(aes(x = 1, y = 1), label = \"code will be here\") +   theme_void()"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"script-from-panel-a","dir":"Articles","previous_headings":"Example 1 (Figure 2)","what":"Script from panel A","title":"Examples from the slendr paper","text":"","code":"o <- population(\"o\", time = 1, N = 100) c <- population(\"c\", time = 2500, N = 100, parent = o) a <- population(\"a\", time = 2800, N = 100, parent = c) b <- population(\"b\", time = 3700, N = 100, parent = a) x1 <- population(\"x1\", time = 4000, N = 15000, parent = c) x2 <- population(\"x2\", time = 4300, N = 15000, parent = x1)  gf <- gene_flow(from = b, to = x1, start = 5400, end = 5800, 0.1)  model <- compile_model(   populations = list(o, a, b, c, x1, x2), gene_flow = gf,   generation_time = 1, simulation_length = 6000 )  plot_model(model, sizes = FALSE, proportions = TRUE) # panel B  ts <- msprime(model, sequence_length = 100e6 / scaling, recombination_rate = 1e-8, random_seed = SEED) %>%   ts_mutate(mutation_rate = 1e-8, random_seed = SEED)  samples <- ts_samples(ts) %>% group_by(pop) %>% sample_n(100)  # panel C divergence <- ts_divergence(ts, split(samples$name, samples$pop))  # panel D f4ratio <- ts_f4ratio(   ts, X = filter(samples, pop %in% c(\"x1\", \"x2\"))$name,   A = \"a_1\", B = \"b_1\", C = \"c_1\", O = \"o_1\" )"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"plotting-code","dir":"Articles","previous_headings":"Example 1 (Figure 2)","what":"Plotting code","title":"Examples from the slendr paper","text":"","code":"divergence <- divergence %>% mutate(pair = paste(x, \"-\", y))  f4ratio <- f4ratio %>% mutate(population = gsub(\"^(.*)_.*$\", \"\\\\1\", X), alpha = alpha * 100)  p_ex1_divergence <- divergence %>%   ggplot(aes(fct_reorder(pair, divergence), divergence)) +   geom_point(size = 2.5) +   xlab(\"population pair\") + ylab(\"pairwise divergence\") +   theme_minimal() +   theme(legend.position = \"bottom\",         legend.text = element_text(size = 10),         axis.text.x = element_text(hjust = 1, angle = 45, size = 8),         axis.title.x = element_blank())  p_ex1_f4ratio <- f4ratio %>%   ggplot(aes(population, alpha)) +   geom_hline(yintercept = 0, linetype = 2) +   geom_jitter(alpha = 0.5) +   geom_boxplot(outlier.shape = NA, alpha = 0.7) +   ylab(base::expression(italic(\"f\")[4]~\"-ratio ancestry proportion [%]\")) +   theme_minimal() +   coord_cartesian(ylim = c(0, 20)) +   theme(legend.position = \"none\",         axis.text.x = element_text(size = 11),         axis.title.x = element_blank(),         panel.grid.major.x = element_blank())#; p_ex3_f4ratio  # let's avoid ggpubr as another dependency: # https://github.com/kassambara/ggpubr/blob/master/R/as_ggplot.R#L27 p_ex1_legend <- ggdraw() + draw_grob(grid::grobTree(get_legend(p_ex1_divergence)))  p_ex1_model <- plot_model(model, sizes = FALSE, proportions = TRUE)  p_ex1 <- plot_grid(   p_code,   plot_grid(     p_ex1_model,     plot_grid(       p_ex1_divergence + theme(legend.position = \"none\"),       p_ex1_f4ratio,       ncol = 2, rel_widths = c(1, 0.8), labels = c(\"C\", \"D\")     ),     p_ex1_legend, nrow = 3, rel_heights = c(1, 1, 0.1),     labels = \"B\"   ),   nrow = 1, labels = c(\"A\", \"\") )  p_ex1"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"script-from-panel-a-1","dir":"Articles","previous_headings":"Example 2 (Figure 3)","what":"Script from panel A","title":"Examples from the slendr paper","text":"","code":"map <- world(xrange = c(0, 10), yrange = c(0, 10),              landscape = region(center = c(5, 5), radius = 5))  p1 <- population(\"pop1\", time = 1, N = 2000, map = map, competition = 0) p2 <- population(\"pop2\", time = 1, N = 2000, map = map, competition = 9) p3 <- population(\"pop3\", time = 1, N = 2000, map = map, competition = 6) p4 <- population(\"pop4\", time = 1, N = 2000, map = map, competition = 5) p5 <- population(\"pop5\", time = 1, N = 2000, map = map, competition = 4) p6 <- population(\"pop6\", time = 1, N = 2000, map = map, competition = 3) p7 <- population(\"pop7\", time = 1, N = 2000, map = map, competition = 2) p8 <- population(\"pop8\", time = 1, N = 2000, map = map, competition = 1)  model <- compile_model(   populations = list(p1, p2, p3, p4, p5, p6, p7, p8),   generation_time = 1, simulation_length = 5000, resolution = 0.1,   mating = 0.1, dispersal = 0.05 )  ts <-   slim(model, sequence_length = 10e6 / scaling, recombination_rate = 1e-8, random_seed = SEED) %>%   ts_simplify() %>%   ts_mutate(mutation_rate = 1e-7, random_seed = SEED) #> Warning: Simplifying a non-recapitated tree sequence. Make sure this is what #> you really want  locations <- ts_nodes(ts) %>% filter(time == max(time))  heterozygosity <- ts_samples(ts) %>%   group_by(pop) %>%   sample_n(100) %>%   mutate(pi = ts_diversity(ts, name)$diversity)"},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"plotting-code-1","dir":"Articles","previous_headings":"Example 2 (Figure 3)","what":"Plotting code","title":"Examples from the slendr paper","text":"","code":"p_ex2_clustering <- ggplot() +   geom_sf(data = map) +   geom_sf(data = locations, aes(color = pop), size = 0.05, alpha = 0.25) +   facet_grid(. ~ pop, switch = \"x\") +   xlab(\"spatial distributions emerged in the simulation\") +   theme(     strip.background = element_blank(),     strip.text = element_text(size = 11),     panel.grid = element_blank(),     axis.ticks = element_blank(),     axis.text = element_blank(),     panel.background = element_blank()   ) +   guides(color = \"none\")  p_ex2_diversity <- ggplot(heterozygosity, aes(pop, pi, color = pop)) +   geom_violin(color = \"black\") +   geom_jitter(alpha = 0.5) +   labs(y = \"individual heterozygosity\") +   guides(color = \"none\") +   theme_minimal() +   theme(axis.title.x = element_blank(),         axis.text.x = element_blank(), panel.grid.major.x = element_blank(),         plot.margin = margin(t = 0.2, r = 0.2, b = -0.1, l = 0.2, \"cm\"))  p_ex2 <- plot_grid(   p_code,   plot_grid(     p_ex2_diversity,     p_ex2_clustering +       theme(plot.margin = margin(t = 0, r = 0.4, b = 0, l = 1.8, \"cm\")),     nrow = 2,     rel_heights = c(1, 0.5),     labels = c(\"B\", \"C\")   ),   nrow = 2, labels = c(\"A\", \"\"), rel_heights = c(1.5, 1) )  p_ex2"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"script-from-panel-a-2","dir":"Articles","previous_headings":"Example 3 (Figure 4)","what":"Script from panel A","title":"Examples from the slendr paper","text":"","code":"map <- world(xrange = c(-13, 70), yrange = c(18, 65), crs = 3035) #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpHixcFC/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84  R1 <- region(   \"EHG range\", map,   polygon = list(c(26, 55), c(38, 53), c(48, 53), c(60, 53),                  c(60, 60), c(48, 63), c(38, 63), c(26, 60)) ) R2 <- region(   \"Europe\", map,   polygon = list(     c(-8, 35), c(-5, 36), c(10, 38), c(20, 35), c(25, 35),     c(33, 45), c(20, 58), c(-5, 60), c(-15, 50)   ) ) R3 <- region(   \"Anatolia\", map,   polygon = list(c(28, 35), c(40, 35), c(42, 40),                  c(30, 43), c(27, 40), c(25, 38)) ) R4 <- join(R2, R3) R5 <- region(   \"YAM range\", map,   polygon = list(c(26, 50), c(38, 49), c(48, 50),                  c(48, 56), c(38, 59), c(26, 56)) )  ooa_trajectory <- list(c(40, 30), c(50, 30), c(60, 40), c(45, 55)) map <- world(xrange = c(-13, 70), yrange = c(18, 65), crs = 3035) #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpHixcFC/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84  ooa <- population(   \"OOA\", time = 50000, N = 500, remove = 23000,    map = map, center = c(33, 30), radius = 400e3 ) %>%   move(trajectory = ooa_trajectory, start = 50000, end = 40000, snapshots = 30)  ehg <- population(   \"EHG\", time = 28000, N = 1000, parent = ooa, remove = 6000,   map = map, polygon = R1 )  eur <- population(   \"EUR\", time = 30000, N = 2000, parent = ooa,   map = map, polygon = R2 ) %>%   resize(N = 10000, time = 5000, end = 0, how = \"exponential\")  ana <- population(   \"ANA\", time = 25000, N = 4000, parent = ooa, remove = 3000,   map = map, polygon = R3 ) %>%   expand_range(by = 3e6, start = 10000, end = 7000, polygon = R4, snapshots = 15)  yam <- population(   \"YAM\", time = 7000, N = 600, parent = ehg, remove = 2500,   map = map, polygon = R5 ) %>%   move(trajectory = list(c(15, 50)), start = 5000, end = 3000, snapshots = 10)  gf <- list(   gene_flow(ana, to = yam, rate = 0.5, start = 6500, end = 5000),   gene_flow(ana, to = eur, rate = 0.6, start = 8000, end = 6000),   gene_flow(yam, to = eur, rate = 0.7, start = 3500, end = 3000) )  model <- compile_model(   populations = list(ooa, ehg, eur, ana, yam), gene_flow = gf,   generation_time = 30, resolution = 10e3,   competition = 150e3, mating = 120e3, dispersal = 90e3 )  samples <- schedule_sampling(   model, times = seq(0, 50000, by = 1000),   list(ehg, 20), list(ana, 20), list(yam, 20), list(eur, 20) )  plot_model(model, sizes = FALSE) plot_map(model)  ts <- slim(   model, burnin = 200000, samples = samples, random_seed = SEED,   sequence_length = 200000, recombination_rate = 1e-8 )"},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"plotting-code-2","dir":"Articles","previous_headings":"Example 3 (Figure 4)","what":"Plotting code","title":"Examples from the slendr paper","text":"","code":"p_map <- plot_map(model) +   theme(legend.position = \"bottom\") +   guides(alpha = \"none\")  p_ex3 <- plot_grid(   p_code,   plot_grid(     plot_model(model, sizes = FALSE),     p_map,     labels = c(\"B\", \"C\"), nrow = 2, rel_heights = c(1, 1)   ),   ncol = 2, labels = c(\"A\", \"\"), rel_widths = c(1, 1.2) )  p_ex3"},{"path":[]},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"script-from-panel-a-3","dir":"Articles","previous_headings":"Example 4 (Figure 5)","what":"Script from panel A","title":"Examples from the slendr paper","text":"","code":"ts_small <- ts_simplify(ts, simplify_to = c(\"EUR_578\", \"YAM_75\", \"ANA_163\", \"EHG_208\"))  # tskit uses zero-based indexing tree <- ts_phylo(ts_small, i = 15 / scaling) #> Starting checking the validity of tree... #> Found number of tips: n = 8  #> Found number of nodes: m = 7  #> Done. nodes <- ts_nodes(tree) edges <- ts_edges(tree)  ancestors <- ts_ancestors(ts, \"EUR_578\")"},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"plotting-code-3","dir":"Articles","previous_headings":"Example 4 (Figure 5)","what":"Plotting code","title":"Examples from the slendr paper","text":"explained top page, random number generation current SLiM slightly different version 4.0.1 available time writing paper. figure looks little different well (different random generation implies different movement individuals, different individuals sampled). record, ’s original figure used paper also originally part vignette:","code":"library(ggtree) #> ggtree v3.10.0 For help: https://yulab-smu.top/treedata-book/ #>  #> If you use the ggtree package suite in published research, please cite #> the appropriate paper(s): #>  #> Guangchuang Yu, David Smith, Huachen Zhu, Yi Guan, Tommy Tsan-Yuk Lam. #> ggtree: an R package for visualization and annotation of phylogenetic #> trees with their covariates and other associated data. Methods in #> Ecology and Evolution. 2017, 8(1):28-36. doi:10.1111/2041-210X.12628 #>  #> Guangchuang Yu.  Data Integration, Manipulation and Visualization of #> Phylogenetic Trees (1st edition). Chapman and Hall/CRC. 2022, #> doi:10.1201/9781003279242 #>  #> Guangchuang Yu. Using ggtree to visualize data on tree-like structures. #> Current Protocols in Bioinformatics. 2020, 69:e96. doi:10.1002/cpbi.96 #>  #>  #> Attaching package: 'ggtree' #> The following object is masked from 'package:tidyr': #>  #>     expand  # prepare annotation table for ggtree linking R phylo node ID (not tskit integer # ID!) of each node with its population name df <- as_tibble(nodes) %>% select(node = phylo_id, pop)  abs_comma <- function (x, ...) {   format(abs(x) / 1000, ..., scientific = FALSE, trim = TRUE) }  highlight_nodes <- as_tibble(nodes) %>% dplyr::filter(name == \"EUR_578\") %>% .$phylo_id  p_tree <- ggtree(tree, aes(color = pop, fill = pop)) %<+% df +   geom_tiplab(align = TRUE, geom = \"label\", offset = 2000,               color = \"white\", fontface = \"bold\", size = 2.7) +   geom_tiplab(align = TRUE, geom = NULL, linetype = \"dotted\", size = 0) +   geom_point2(aes(subset = (node %in% highlight_nodes)), color = \"black\", size = 2.7) +   geom_label2(aes(label = label, subset = !isTip),               color = \"black\", size = 2.7) +   theme_tree2() +   theme(legend.position = \"none\") +   xlab(\"time before present [thousand years ago]\") +   scale_x_continuous(limits = c(-80000, 31000), labels = abs_comma,                      breaks = -c(100, 80, 60, 40, 20, 0) * 1000) p_tree <- revts(p_tree)  # nodes$label <- ifelse(is.na(nodes$name), nodes$node_id, nodes$name) nodes$label <- sapply(1:nrow(nodes), function(i) {   if (is.na(nodes[i, ]$name))     nodes[i, ]$node_id   else {     ind <- nodes[i, ]$name     paste(nodes[!is.na(nodes$name) & nodes$name == ind, ]$node_id, collapse = \"&\")   } })  p_map <- ggplot() +   geom_sf(data = map) +   geom_sf(data = edges, aes(color = parent_pop), size = 0.5) +   geom_sf_label(data = nodes[!nodes$sampled, ],                 aes(label = node_id, fill = pop), size = 3) +   geom_sf_label(data = nodes[nodes$sampled, ],                 aes(label = label, fill = pop), size = 3,                 fontface = \"bold\", color = \"white\") +   coord_sf(xlim = c(3177066.1, 7188656.9),            ylim = c(757021.7, 5202983.3), expand = 0) +   guides(fill = guide_legend(\"\", override.aes = aes(label = \"\"))) +   guides(color = \"none\") +   scale_colour_discrete(drop = FALSE) +   scale_fill_discrete(drop = FALSE) +   theme_bw() +   theme(legend.position = \"bottom\",         axis.title.x = element_blank(),         axis.title.y = element_blank())  chrom_names <- stats::setNames(   c(\"EUR_578 (node 6)\", \"EUR_578 (node 7)\"),   unique(ancestors$node_id) )  p_ancestors <- ggplot() +   geom_sf(data = map) +   geom_sf(data = ancestors, size = 0.5, alpha = 0.25) +   geom_sf(data = sf::st_set_geometry(ancestors, \"parent_location\"),           aes(shape = parent_pop, color = parent_pop)) +   geom_sf(data = filter(ts_nodes(ts), name == \"EUR_578\"), size = 3) +   coord_sf(expand = 0) +   labs(x = \"longitude\", y = \"latitude\") +   theme_bw() +   facet_grid(. ~ node_id, labeller = labeller(node_id = chrom_names)) +   theme(legend.position = \"none\")  p_legend <- ggdraw() + draw_grob(grid::grobTree(get_legend(p_map)))  p_ex4 <- plot_grid(   p_code,   plot_grid(p_tree + theme(legend.position = \"none\"),             p_map + theme(legend.position = \"none\"),             labels = c(\"B\", \"C\"), rel_widths = c(1, 0.9)),   p_ancestors,   p_legend,   labels = c(\"A\", \"\", \"D\", \"\"),   nrow = 4, rel_heights = c(0.5, 1, 1, 0.1) )  p_ex4"},{"path":"https://www.slendr.net/articles/vignette-09-paper.html","id":"run-time-of-each-code-example-from-the-paper","dir":"Articles","previous_headings":"","what":"Run time of each code example from the paper","title":"Examples from the slendr paper","text":"following times measured 16’’ MacBook Pro Apple M1 Pro chip (2021 model), 32 GB RAM, running macOS Ventura Version 13.1. table , ex1, ex2, ex3, ex4 correspond runtimes code shown one four example figures slendr paper.","code":""},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"demographic-model","dir":"Articles","previous_headings":"","what":"Demographic model","title":"Extracting true ancestry tracts","text":"Let’s imagine following demographic model Neanderthal introgression ancestors non-Africans (represented “EUR” “PAP” populations, approximating European Papuan people living today), followed Denisovan introgression ancestors Papuans:","code":"anc_all <- population(\"ancestor_all\", time = 700e3, N = 10000, remove = 640e3) afr <- population(\"AFR\", parent = anc_all, time = 650e3, N = 10000) anc_arch <- population(\"ancestor_archaics\", parent = anc_all, time = 650e3, N = 10000, remove = 390e3) nea <- population(\"NEA\", parent = anc_arch, time = 400e3, N = 2000, remove = 30e3) den <- population(\"DEN\", parent = anc_arch, time = 400e3, N = 2000, remove = 30e3) nonafr <- population(\"nonAFR\", parent = afr, time = 100e3, N = 3000, remove = 39e3) eur <- population(\"EUR\", parent = nonafr, time = 45e3, N = 5000) pap <- population(\"PAP\", parent = nonafr, time = 45e3, N = 5000)  gf <- list(   gene_flow(from = nea, to = nonafr, rate = 0.03, start = 55000, end = 50000),   gene_flow(from = den, to = pap, rate = 0.07, start = 35000, end = 30000) )  model <- compile_model(   populations = list(anc_all, afr, anc_arch, nea, den, nonafr, eur, pap),   gene_flow = gf,   generation_time = 30,   serialize = FALSE )  plot_model(   model, sizes = FALSE,   order = c(\"AFR\", \"EUR\", \"nonAFR\", \"PAP\", \"ancestor_all\", \"DEN\", \"ancestor_archaics\", \"NEA\") )"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"tree-sequence-simulation","dir":"Articles","previous_headings":"","what":"Tree sequence simulation","title":"Extracting true ancestry tracts","text":"Let’s now simulate 50Mb tree sequence model, recording 50 diploid individuals EUR PAP populations:","code":"samples <- schedule_sampling(model, times = 0, list(eur, 50), list(pap, 50))  ts <- msprime(model, sequence_length = 100e6, recombination_rate = 1e-8, samples = samples, random_seed = 42)"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"extracting-ancestry-tracts","dir":"Articles","previous_headings":"","what":"Extracting ancestry tracts","title":"Extracting true ancestry tracts","text":"order extract tracts Neanderthal Denisovan ancestry, can use slendr’s new function ts_tracts() serves simplified R-friendly interface Python method tspop.get_pop_ancestry(). important piece information used function -called “census time”, records time recording “ancestral population” identity node ancestral subsegment sample set. Please see excellent vignette tspop information inner workings algorithm. case, let’s extract ancestry tracts corresponding ancestral nodes present 55 thousand years ago – time corresponds moment start archaic introgression: table ancestry tracts looks like. expect, see column indicating name individual, left right coordinates tract individual, well population name source ancestry tract:","code":"nea_tracts <- ts_tracts(ts, census = 55000) #>  #> PopAncestry summary #> Number of ancestral populations:     3 #> Number of sample chromosomes:        200 #> Number of ancestors:             44052 #> Total length of genomes:         20000000000.000000 #> Ancestral coverage:          20000000000.000000 den_tracts <- ts_tracts(ts, census = 35000) #>  #> PopAncestry summary #> Number of ancestral populations:     3 #> Number of sample chromosomes:        200 #> Number of ancestors:             55373 #> Total length of genomes:         20000000000.000000 #> Ancestral coverage:          20000000000.000000  tracts <- bind_rows(nea_tracts, den_tracts) tracts #> # A tibble: 17,002 × 8 #>    name  node_id pop   source_pop     left    right length source_pop_id #>    <chr>   <dbl> <fct> <fct>         <dbl>    <dbl>  <dbl>         <dbl> #>  1 EUR_1       0 EUR   NEA          781887   852668  70781             3 #>  2 EUR_1       0 EUR   NEA         1451389  1463837  12448             3 #>  3 EUR_1       0 EUR   NEA         1596995  1601441   4446             3 #>  4 EUR_1       0 EUR   NEA         1629857  1689656  59799             3 #>  5 EUR_1       0 EUR   NEA         3203711  3339565 135854             3 #>  6 EUR_1       0 EUR   NEA         3850629  3947923  97294             3 #>  7 EUR_1       0 EUR   NEA         6190700  6205627  14927             3 #>  8 EUR_1       0 EUR   NEA        10394211 10426667  32456             3 #>  9 EUR_1       0 EUR   NEA        13044685 13180258 135573             3 #> 10 EUR_1       0 EUR   NEA        14858956 14912068  53112             3 #> # ℹ 16,992 more rows"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"summaries-of-ancestral-proportions","dir":"Articles","previous_headings":"","what":"Summaries of ancestral proportions","title":"Extracting true ancestry tracts","text":"summarise ancestry proportions target EUR PAP populations, see EUR population carries ~3% Neanderthal ancestry also true PAP population. However, also see Papuans carry 7% Denisovan ancestry. consistent model, also expectation empirical data. Let’s visualize proportions individual level:","code":"summary <- tracts %>%   group_by(name, node_id, pop, source_pop) %>%   summarise(prop = sum(length) / 100e6) #> `summarise()` has grouped output by 'name', 'node_id', 'pop'. You can override #> using the `.groups` argument.  summary %>% group_by(pop, source_pop) %>% summarise(mean(prop)) %>% arrange(source_pop, pop) #> `summarise()` has grouped output by 'pop'. You can override using the `.groups` #> argument. #> # A tibble: 3 × 3 #> # Groups:   pop [2] #>   pop   source_pop `mean(prop)` #>   <fct> <fct>             <dbl> #> 1 EUR   NEA              0.0303 #> 2 PAP   NEA              0.0348 #> 3 PAP   DEN              0.0739 summary %>% ggplot(aes(source_pop, prop, color = source_pop, fill = source_pop)) +   geom_jitter() +   coord_cartesian(ylim = c(0, 0.2)) +   geom_hline(yintercept = c(0.03, 0.08), linetype = 2) +   ylab(\"ancestry proportion\") +   facet_wrap(~ pop) +   ggtitle(\"Ancestry proportions in each individual\",           \"(vertical lines represent 3% and 7% baseline expectations\")"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"chromosome-painting-of-ancestry-tracts","dir":"Articles","previous_headings":"","what":"“Chromosome painting” of ancestry tracts","title":"Extracting true ancestry tracts","text":"tracts object contains coordinates every single ancestry segment simulated individuals, can “paint” chromosome two archaic human ancestries:  lining NEA & DEN ancestry tracts EUR PAP populations, can see common origin Neanderthal ancestry non-African populations manifests significant overlap NEA tracts populations.","code":"tracts %>% mutate(chrom = paste(name, \" (node\", node_id, \")\")) %>% ggplot(aes(x = left, xend = right, y = chrom, yend = chrom, color = source_pop)) +   geom_segment(linewidth = 3) +   theme_minimal() +   labs(x = \"position [bp]\", y = \"haplotype\") +   ggtitle(\"True ancestry tracts along each chromosome\") +   theme(axis.text.y = element_blank(), panel.grid = element_blank()) +   facet_grid(pop ~ ., scales = \"free_y\")"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"average-tract-lengths","dir":"Articles","previous_headings":"","what":"Average tract lengths:","title":"Extracting true ancestry tracts","text":"Let’s compute simple summaries tract lengths simulated data, compare theoretical expectations. Theoretical expectations (Racimo Slatkin 2015, Box 1) Neanderthal tracts: Denisovan tracts: can see, simulations far theoretical expectations, giving us confidence simulations (ancestry tract extraction algorithm) working expected.","code":"tracts %>%   group_by(pop, source_pop) %>%   summarise(mean(length)) #> `summarise()` has grouped output by 'pop'. You can override using the `.groups` #> argument. #> # A tibble: 3 × 3 #> # Groups:   pop [2] #>   pop   source_pop `mean(length)` #>   <fct> <fct>               <dbl> #> 1 EUR   NEA                65765. #> 2 PAP   NEA                69204. #> 3 PAP   DEN               100317. m <- 0.03 t <- 52500 / 30 r <- 1e-8  mean_nea <- 1 / ((1 - m) * r * (t - 1)) mean_nea #> [1] 58943.84 m <- 0.07 t <- 37500 / 30 r <- 1e-8  mean_den <- 1 / ((1 - m) * r * (t - 1)) mean_den #> [1] 86090.38"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"distribution-of-ancestry-tract-lengths","dir":"Articles","previous_headings":"","what":"Distribution of ancestry tract lengths","title":"Extracting true ancestry tracts","text":"Finally, let’s plot distributions lengths ancestry tracts. case archaic human introgression well studied ’s perhaps exciting look figures. said, less well studied species, might interesting use kinds simulations inference introgression times proportions via Approximate Bayesian Computation another method:","code":"expectation_df <- data.frame(   pop = c(\"EUR\", \"PAP\", \"PAP\"),   source_pop = c(\"NEA\", \"NEA\", \"DEN\"),   length = c(mean_nea, mean_nea, mean_den) ) p_densities <- tracts %>% ggplot(aes(length, color = source_pop)) +   geom_density() +   geom_vline(data = expectation_df, aes(xintercept = length, color = source_pop),              linetype = 2) +   facet_wrap(~ pop) +   ggtitle(\"Distribution of tract lengths per different ancestries\")  cowplot::plot_grid(p_densities, p_densities + scale_x_log10(), nrow = 2)"},{"path":"https://www.slendr.net/articles/vignette-10-tracts.html","id":"pure-msprime-tree-sequence","dir":"Articles","previous_headings":"","what":"Pure msprime tree sequence","title":"Extracting true ancestry tracts","text":"Finally, sanity check, let’s use pure msprime simulation example official tspop documentation test ts_tracts() behaves expected even standard msprime tree-sequence object. First, let’s run simulation code exactly : Let’s save msprime tree sequence disk can load R (.e., approximating might want want use ts_tracts() without running slendr simulation first): Now let’s move R , load tree sequence slendr extract ancestry tracts using ts_tracts(): setting squashed = FALSE, get full, un-squashed ancestry segments, appropriate ancestral node ID: comparing two tables pandas data frames tspop documentation, can see obtained results.","code":"import msprime  pop_size = 500 sequence_length = 1e7 seed = 98765 rho = 3e-8  # Make the Demography object. demography = msprime.Demography() demography.add_population(name=\"RED\", initial_size=pop_size) #> Population(initial_size=500, growth_rate=0, name='RED', description='', extra_metadata={}, default_sampling_time=None, initially_active=None, id=0) demography.add_population(name=\"BLUE\", initial_size=pop_size) #> Population(initial_size=500, growth_rate=0, name='BLUE', description='', extra_metadata={}, default_sampling_time=None, initially_active=None, id=1) demography.add_population(name=\"ADMIX\", initial_size=pop_size) #> Population(initial_size=500, growth_rate=0, name='ADMIX', description='', extra_metadata={}, default_sampling_time=None, initially_active=None, id=2) demography.add_population(name=\"ANC\", initial_size=pop_size) #> Population(initial_size=500, growth_rate=0, name='ANC', description='', extra_metadata={}, default_sampling_time=None, initially_active=None, id=3) demography.add_admixture(     time=100, derived=\"ADMIX\", ancestral=[\"RED\", \"BLUE\"], proportions=[0.5, 0.5] ) #> Admixture(time=100, derived='ADMIX', ancestral=['RED', 'BLUE'], proportions=[0.5, 0.5]) demography.add_census(time=100.01) # Census is here! #> CensusEvent(time=100.01) demography.add_population_split(     time=1000, derived=[\"RED\", \"BLUE\"], ancestral=\"ANC\" ) #> PopulationSplit(time=1000, derived=['RED', 'BLUE'], ancestral='ANC')  # Simulate. ts = msprime.sim_ancestry(     samples={\"RED\": 0, \"BLUE\": 0, \"ADMIX\" : 2},     demography=demography,     random_seed=seed,     sequence_length=sequence_length,     recombination_rate=rho ) import tempfile path = tempfile.NamedTemporaryFile(suffix=\".trees\").name  ts.dump(path) sim_ts <- ts_load(reticulate::py$path)  squashed_tracts <- ts_tracts(sim_ts, census = 100.01, squashed = TRUE) #>  #> PopAncestry summary #> Number of ancestral populations:     2 #> Number of sample chromosomes:        4 #> Number of ancestors:             118 #> Total length of genomes:         40000000.000000 #> Ancestral coverage:          40000000.000000  head(squashed_tracts) #> # A tibble: 6 × 4 #>   sample    left   right population #>    <dbl>   <dbl>   <dbl>      <dbl> #> 1      0       0  419848          0 #> 2      0  419848  483009          1 #> 3      0  483009 1475765          0 #> 4      0 1475765 2427904          1 #> 5      0 2427904 3635390          0 #> 6      0 3635390 4606954          1 tail(squashed_tracts) #> # A tibble: 6 × 4 #>   sample    left    right population #>    <dbl>   <dbl>    <dbl>      <dbl> #> 1      3 7134130  7362300          1 #> 2      3 7362300  7369409          0 #> 3      3 7369409  7596783          1 #> 4      3 7596783  8289015          0 #> 5      3 8289015  8918727          1 #> 6      3 8918727 10000000          0 full_tracts <- ts_tracts(sim_ts, census = 100.01, squashed = FALSE) #>  #> PopAncestry summary #> Number of ancestral populations:     2 #> Number of sample chromosomes:        4 #> Number of ancestors:             118 #> Total length of genomes:         40000000.000000 #> Ancestral coverage:          40000000.000000  head(full_tracts) #> # A tibble: 6 × 5 #>   sample   left   right ancestor population #>    <int>  <dbl>   <dbl>    <int>      <int> #> 1      0      0   33027       74          0 #> 2      0  33027  155453       33          0 #> 3      0 155453  290542       46          0 #> 4      0 290542  419848       18          0 #> 5      0 419848  483009       83          1 #> 6      0 483009 1475765       28          0 tail(full_tracts) #> # A tibble: 6 × 5 #>   sample    left    right ancestor population #>    <int>   <dbl>    <dbl>    <int>      <int> #> 1      3 8477625  8672850       94          1 #> 2      3 8672850  8849756       95          1 #> 3      3 8849756  8918727      131          1 #> 4      3 8918727  9165035       44          0 #> 5      3 9165035  9176562       47          0 #> 6      3 9176562 10000000       58          0"},{"path":"https://www.slendr.net/authors.html","id":null,"dir":"","previous_headings":"","what":"Authors","title":"Authors and Citation","text":"Martin Petr. Author, maintainer.","code":""},{"path":"https://www.slendr.net/authors.html","id":"citation","dir":"","previous_headings":"","what":"Citation","title":"Authors and Citation","text":"Petr M (2024). slendr: Simulation Framework Spatiotemporal Population Genetics. R package version 0.9.1, https://github.com/bodkan/slendr.","code":"@Manual{,   title = {slendr: A Simulation Framework for Spatiotemporal Population Genetics},   author = {Martin Petr},   year = {2024},   note = {R package version 0.9.1},   url = {https://github.com/bodkan/slendr}, }"},{"path":[]},{"path":"https://www.slendr.net/index.html","id":"overview-","dir":"","previous_headings":"","what":"Overview","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"slendr R package toolbox defining population genetic models simulating genomic data entirely R. originally conceived framework simulating spatially-explicit genomic data real geographic landscapes grown much . page briefly summarizes slendr’s important features. much detailed description slendr architecture extensive set practical code examples can found paper PCI journal website.","code":""},{"path":"https://www.slendr.net/index.html","id":"citing-slendr","dir":"","previous_headings":"","what":"Citing slendr","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"slendr paper now published Peer Community Journal! use slendr work, please cite : Petr, Martin; Haller, Benjamin C.; Ralph, Peter L.; Racimo, Fernando. slendr: framework spatio-temporal population genomic simulations geographic landscapes. Peer Community Journal, Volume 3 (2023), article . e121. doi : 10.24072/pcjournal.354. Citations help justify development fixing bugs! Thank ! ❤️","code":""},{"path":"https://www.slendr.net/index.html","id":"main-features","dir":"","previous_headings":"","what":"Main features","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"brief summary slendr’s important features. R package allows : Program demographic models, including population splits, population size changes, gene-flow events using extremely simple declarative language entirely R (see vignette example model-definition interface). Even complex models can written little code require bare minimum R programming knowledge (thing user needs know call R function R data frame look like). Execute slendr models using efficient, tailor-made SLiM msprime simulation scripts bundled R package. simulation engines save outputs form efficient tree-sequence data structure. SLiM msprime programming needed! Load, process, analyse tree-sequence outputs via slendr’s built-R interface tree-sequence library tskit. can compute many population genetic statistics R immediately simulation finishes directly output tree sequences, without convert files formats (VCF, EIGENSTRAT) analysis different software. Encode complex models population movements landscape (see brief example model , extended explanation tutorial). knowledge cartographic geospatial analysis concepts needed. Simulate dynamic spatial demographic models using SLiM’s continuous-space simulation capabilities directly R (, SLiM programming required). outputs simulations saved tree sequences can analysed using standard R geospatial data analysis libraries. slendr performs conversion tree sequence tables appropriate spatial R data type automatically. Specify within-population individual dispersal dynamics R interface leveraging SLiM’s individual interaction parameters implemented SLiM back-end script. Schedule sampling events specify many individuals’ genomes, populations, times (optionally, locations) recorded simulation engine (SLiM msprime) output tree-sequence files. Utilizing flexibility R wealth libraries statistics, geospatial analysis graphics, combining power population genetic simulation frameworks SLiM msprime, slendr R package makes possible write entire simulation analytic pipelines without need leave R environment.","code":""},{"path":"https://www.slendr.net/index.html","id":"testing-the-r-package-in-an-online-rstudio-session","dir":"","previous_headings":"","what":"Testing the R package in an online RStudio session","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"can open RStudio session test examples vignettes directly web browser clicking button (installation needed!):  case RStudio instance appears starting slowly, please patient (Binder freely available service limited computational resources provided community). Binder crashes, try reloading web page, restart cloud session. get browser-based RStudio session, can navigate vignettes/ directory test examples !","code":""},{"path":"https://www.slendr.net/index.html","id":"installation","dir":"","previous_headings":"","what":"Installation","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"slendr now available CRAN means can install simply entering install.packages(\"slendr\") R console. like test latest features software (perhaps need bug fixes), can install devtools::install_github(\"bodkan/slendr\") (note requires R package devtools). ⚠️⚠️⚠️ Note: default slendr installation longer hard dependency geospatial R packages sf, stars, rnaturalearth. indend use slendr spatial genomic simulations data analysis, run install.packages(c(\"sf\", \"stars\", \"rnaturalearth\")) first. made change ease installation burden researchers might need slendr’s support traditional, non-spatial simulations (represent majority use-cases population genetics general). geospatial R packages imply non-negligible amount software dependencies , proven unnecessary hurdle many users might need (least first). ⚠️⚠️⚠️ software active development! like stay updated: Click “Watch” button project’s GitHub website. Follow social media posting progress updates (can find links homepage). time time, take look changelog post updates new features, breaking changes, etc.","code":""},{"path":"https://www.slendr.net/index.html","id":"example","dir":"","previous_headings":"","what":"Example","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"brief demonstration kind spatial model slendr originally designed simulate data . Please note although spatially-explicit population genetic model, slendr extensive support traditional, non-spatial simulations well. Furthermore, example shows specify simulate model R. doesn’t show analyse tree-sequence outputs compute population genetic statistics (important feature demonstrated tutorial).","code":""},{"path":"https://www.slendr.net/index.html","id":"id_1-setup-the-spatial-context","dir":"","previous_headings":"Example","what":"1. Setup the spatial context","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"Imagine wanted simulate spatio-temporal genomic data toy model history modern humans West Eurasia Africa migration. First, define spatial context simulation. represent “world” occupied populations model. can visualize defined world map using function plot_map provided package.  Although example use real Earth landscape, map can completely abstract (either blank user-defined landscape features continents, islands, corridors barriers).","code":"library(slendr)  # this sets up internal Python environment and needs to be ran only once! # (do not put this in your R scripts, run this command in the R console # after you (re-)install slendr) setup_env()  # activate the internal Python environment needed for simulation and # tree-sequence processing init_env()  map <- world(   xrange = c(-13, 70), # min-max longitude   yrange = c(18, 65),  # min-max latitude   crs = \"EPSG:3035\"    # coordinate reference system (CRS) for West Eurasia ) plot_map(map)"},{"path":"https://www.slendr.net/index.html","id":"id_2-define-broader-geographic-regions","dir":"","previous_headings":"Example","what":"2. Define broader geographic regions","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"order make definitions population ranges () easier, can define smaller regions map using function region. Note coordinates slendr specified geographic coordinate system (.e., degrees longitude latitude), internally represented projected CRS (case, EPSG 3035 specified ). makes easier us define spatial features simply reading coordinates regular map internal projected CRS makes simulations accurate (distances shapes distorted can use CRS tailored region world working ). projected CRS takes care projection part world ’re interested three-dimensional Earth surface two-dimensional map. , can use generic plot_map function visualize objects, making sure specified correctly:","code":"africa <- region(   \"Africa\", map,   polygon = list(c(-18, 20), c(38, 20), c(30, 33),                  c(20, 33), c(10, 38), c(-6, 35)) ) europe <- region(   \"Europe\", map,   polygon = list(     c(-8, 35), c(-5, 36), c(10, 38), c(20, 35), c(25, 35),     c(33, 45), c(20, 58), c(-5, 60), c(-15, 50)   ) ) anatolia <- region(   \"Anatolia\", map,   polygon = list(c(28, 35), c(40, 35), c(42, 40),                  c(30, 43), c(27, 40), c(25, 38)) ) plot_map(africa, europe, anatolia)"},{"path":"https://www.slendr.net/index.html","id":"id_3-define-demographic-history-and-population-boundaries","dir":"","previous_headings":"Example","what":"3. Define demographic history and population boundaries","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"important function slendr package population(), used define names, split times, sizes spatial ranges populations. , specify times years present, distances kilometers. makes sense models, times can also given forward direction. also note functions move() expand_range() designed take slendr population object schedule spatial dynamics appropriate times model simulation (happen later step). Note order make example executable reasonable time extremely old laptop, decreased sizes populations unrealistic levels. speed SLiM simulation later step. can use function plot_map get “compressed” overview spatio-temporal range dynamics encoded model far (prior simulation ).","code":"afr <- population( # African ancestral population   \"AFR\", time = 52000, N = 3000, map = map, polygon = africa )  ooa <- population( # population of the first migrants out of Africa   \"OOA\", parent = afr, time = 51000, N = 500, remove = 25000,   center = c(33, 30), radius = 400e3 ) %>%   move(     trajectory = list(c(40, 30), c(50, 30), c(60, 40)),     start = 50000, end = 40000, snapshots = 20   )  ehg <- population( # Eastern hunter-gatherers   \"EHG\", parent = ooa, time = 28000, N = 1000, remove = 6000,   polygon = list(     c(26, 55), c(38, 53), c(48, 53), c(60, 53),     c(60, 60), c(48, 63), c(38, 63), c(26, 60)) )  eur <- population( # European population   name = \"EUR\", parent = ehg, time = 25000, N = 2000,   polygon = europe )  ana <- population( # Anatolian farmers   name = \"ANA\", time = 28000, N = 3000, parent = ooa, remove = 4000,   center = c(34, 38), radius = 500e3, polygon = anatolia ) %>%   expand_range( # expand the range by 2.500 km     by = 2500e3, start = 10000, end = 7000,     polygon = join(europe, anatolia), snapshots = 20   )  yam <- population( # Yamnaya steppe population   name = \"YAM\", time = 7000, N = 500, parent = ehg, remove = 2500,   polygon = list(c(26, 50), c(38, 49), c(48, 50),                  c(48, 56), c(38, 59), c(26, 56)) ) %>%   move(trajectory = list(c(15, 50)), start = 5000, end = 3000, snapshots = 10) plot_map(afr, ooa, ehg, eur, ana, yam)"},{"path":"https://www.slendr.net/index.html","id":"id_4-define-gene-flow-events","dir":"","previous_headings":"Example","what":"4. Define gene-flow events","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"default, populations slendr mix even overlapping. order schedule gene-flow event two populations, can use function gene_flow. want specify multiple events , can collect events simple R list:","code":"gf <- list(   gene_flow(from = ana, to = yam, rate = 0.5, start = 6500, end = 6400, overlap = FALSE),   gene_flow(from = ana, to = eur, rate = 0.5, start = 8000, end = 6000),   gene_flow(from = yam, to = eur, rate = 0.75, start = 4000, end = 3000) )"},{"path":"https://www.slendr.net/index.html","id":"id_5-compile-the-model-to-a-set-of-configuration-files","dir":"","previous_headings":"Example","what":"5. Compile the model to a set of configuration files","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"run simulation, compile individual model components (population objects gene-flow events) single R object, specifying additional model parameters. Additionally, performs internal consistency checks, making sure model parameters (split times, gene flow times, etc.) make sense (potentially quite computationally costly) simulation even run. Compiled model kept R object can passed different functions. example, simulate data slim() engine (also run simulation coalescent engine via msprime() function).","code":"model <- compile_model(   populations = list(afr, ooa, ehg, eur, ana, yam), # populations defined above   gene_flow = gf,   generation_time = 30,   resolution = 10e3,                   # resolution in meters per pixel   competition = 130e3, mating = 100e3, # spatial interaction parameters   dispersal = 70e3,                    # how far can offspring end up from their parents )"},{"path":"https://www.slendr.net/index.html","id":"id_6-visualize-the-model","dir":"","previous_headings":"Example","what":"6. Visualize the model","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"package provides R shiny-based browser app explore_model() checking model dynamics interactively visually. complex models, much better static spatial plots one showed step 2 : function two modes: Plotting (“playing”) spatial map dynamics:  Displaying demographic history graph (splits gene-flow events) embedded specified model:","code":"explore_model(model)"},{"path":"https://www.slendr.net/index.html","id":"id_7-run-the-model-in-slim","dir":"","previous_headings":"Example","what":"7. Run the model in SLiM","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"Finally, can execute compiled model SLiM. run simulation batch mode, also run SLiMgui setting method = \"gui\". allow us inspect spatial simulation happens real time. slim function generates complete SLiM script tailored run spatial model defined . saves , user, tremendous amount time, don’t write new SLiM code every time design new demographic model. output simulation run slendr model always tree sequence, loaded object ts_slim. specified , slendr’s SLiM backend simulate 10 Mb sequence individual, produce tree sequence output simulation run can analysed many built-population genetic functions. default, individuals living end simulation recorded samples tree sequence. specific set samples (ancient modern) needed, can defined accordingly using dedicated function. Note although defined spatial model, just easily simulated standard, non-spatial data running model slendr’s msprime() back end without need make changes: quick overview SLiM simulation run summarised GIF animation. , please note simulation extremely simplified. simulated small number individuals population, also didn’t specify dispersal dynamics populations look clumped.  point, either compute population genetic statistics interest perhaps analyse spatial features genealogies simulated model.","code":"ts_slim <- slim(model, sequence_length = 10e6, recombination_rate = 1e-8,                 method = \"batch\", random_seed = 314159) ts_msprime <- msprime(model, sequence_length = 10e6, recombination_rate = 1e-8) animate_model(model = model, file = locations_file, steps = 50, width = 500, height = 300)"},{"path":"https://www.slendr.net/index.html","id":"further-information","dir":"","previous_headings":"","what":"Further information","title":"A Simulation Framework for Spatiotemporal Population Genetics","text":"example provides brief incomplete overview full functionality slendr package. much slendr demonstrated . instance: can tweak parameters influencing dispersal dynamics (“clumpy” populations , far can offspring migrate parents, etc.) define change time. instance, can see animation , African population forms single “blob” really isn’t spread across entire population range. Tweaking dispersal parameters show vignette helps avoid . can use slendr program non-spatial models, means conceivable traditional, random-mating demographic model can simulated lines R code. can learn vignette (detail vignette). SLiM simulations can often quite slow compared coalescent counterparts, also provide functionality allowing simulate slendr models (without change!) using built-msprime back end script. See vignette tutorial works. can build complex spatial models still abstract (assuming real geographic location), including traditional simulations demes lattice structure. complete example shown vignette. slendr & SLiM save data tree-sequence file format, thanks R package reticulate interfacing Python code, full power tskit pyslim manipulating tree-sequence data right fingertips, within convenient environment R. extended example can found vignette. spatially explicit population models, slendr package automatically converts simulated output data format makes possible analyse many available R packages geospatial data analysis. brief description functionality can found vignette. can find complete reproducible code behind examples paper dedicated R vignette .","code":""},{"path":"https://www.slendr.net/reference/animate_model.html","id":null,"dir":"Reference","previous_headings":"","what":"Animate the simulated population dynamics — animate_model","title":"Animate the simulated population dynamics — animate_model","text":"Animate simulated population dynamics","code":""},{"path":"https://www.slendr.net/reference/animate_model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Animate the simulated population dynamics — animate_model","text":"","code":"animate_model(model, file, steps, gif = NULL, width = 800, height = 560)"},{"path":"https://www.slendr.net/reference/animate_model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Animate the simulated population dynamics — animate_model","text":"model Compiled slendr_model model object file Path table saved individual locations steps many frames animation ? gif Path output GIF file (animation object returned default) width, height Dimensions animation pixels","code":""},{"path":"https://www.slendr.net/reference/animate_model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Animate the simulated population dynamics — animate_model","text":"gif = NULL, return gganimate animation object. Otherwise GIF file saved value returned.","code":""},{"path":"https://www.slendr.net/reference/area.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the area covered by the given slendr object — area","title":"Calculate the area covered by the given slendr object — area","text":"Calculate area covered given slendr object","code":""},{"path":"https://www.slendr.net/reference/area.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the area covered by the given slendr object — area","text":"","code":"area(x)"},{"path":"https://www.slendr.net/reference/area.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the area covered by the given slendr object — area","text":"x Object class slendr","code":""},{"path":"https://www.slendr.net/reference/area.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the area covered by the given slendr object — area","text":"Area covered input object. slendr_pop given, table population range area time point returned. slendr_region slendr_world object specified, total area covered object's spatial boundary returned.","code":""},{"path":"https://www.slendr.net/reference/area.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the area covered by the given slendr object — area","text":"","code":"region_a <- region(\"A\", center = c(20, 50), radius = 20) region_b <- region(\"B\", polygon = list(c(50, 40), c(70, 40), c(70, 60), c(50, 60))) plot_map(region_a, region_b)   # note that area won't be *exactly* equal to pi*r^2: #   https://stackoverflow.com/a/65280376 area(region_a) #> [1] 1256.063  area(region_b) #> [1] 400"},{"path":"https://www.slendr.net/reference/as.phylo.slendr_phylo.html","id":null,"dir":"Reference","previous_headings":"","what":"Convert an annotated slendr_phylo object to a phylo object — as.phylo.slendr_phylo","title":"Convert an annotated slendr_phylo object to a phylo object — as.phylo.slendr_phylo","text":"function servers workaround around ggtree error: Error UseMethod(\".phylo\") :     applicable method '.phylo' applied object class     \"c('phylo', 'slendr_phylo')\"","code":""},{"path":"https://www.slendr.net/reference/as.phylo.slendr_phylo.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Convert an annotated slendr_phylo object to a phylo object — as.phylo.slendr_phylo","text":"","code":"# S3 method for slendr_phylo as.phylo(x, ...)"},{"path":"https://www.slendr.net/reference/as.phylo.slendr_phylo.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Convert an annotated slendr_phylo object to a phylo object — as.phylo.slendr_phylo","text":"x Tree object class slendr_phylo ... Additional (unused) arguments .phylo S3 method","code":""},{"path":"https://www.slendr.net/reference/as.phylo.slendr_phylo.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Convert an annotated slendr_phylo object to a phylo object — as.phylo.slendr_phylo","text":"Standard phylogenetic tree object implemented R package ape","code":""},{"path":"https://www.slendr.net/reference/check_dependencies.html","id":null,"dir":"Reference","previous_headings":"","what":"Check that the required dependencies are available for slendr to work — check_dependencies","title":"Check that the required dependencies are available for slendr to work — check_dependencies","text":"Check required dependencies available slendr work","code":""},{"path":"https://www.slendr.net/reference/check_dependencies.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check that the required dependencies are available for slendr to work — check_dependencies","text":"","code":"check_dependencies(python = FALSE, slim = FALSE, quit = FALSE)"},{"path":"https://www.slendr.net/reference/check_dependencies.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check that the required dependencies are available for slendr to work — check_dependencies","text":"python slendr Python environment required? slim SLiM required? quit R interpreter quit required slendr dependencies missing? option (turned default, set FALSE) used mainly avoiding running slendr man page examples machines lack dependencies. set TRUE, logical value returned.","code":""},{"path":"https://www.slendr.net/reference/check_dependencies.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check that the required dependencies are available for slendr to work — check_dependencies","text":"quit = TRUE, values returned, quit = FALSE, scalar logical value returned indicating whether dependencies present.","code":""},{"path":"https://www.slendr.net/reference/check_env.html","id":null,"dir":"Reference","previous_headings":"","what":"Check that the active Python environment is setup for slendr — check_env","title":"Check that the active Python environment is setup for slendr — check_env","text":"function inspects Python environment activated reticulate package prints versions slendr Python dependencies console.","code":""},{"path":"https://www.slendr.net/reference/check_env.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check that the active Python environment is setup for slendr — check_env","text":"","code":"check_env(verbose = TRUE)"},{"path":"https://www.slendr.net/reference/check_env.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check that the active Python environment is setup for slendr — check_env","text":"verbose log message printed? FALSE, logical value returned (invisibly).","code":""},{"path":"https://www.slendr.net/reference/check_env.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check that the active Python environment is setup for slendr — check_env","text":"Either TRUE (slendr Python environment present) FALSE (slendr Python environment present).","code":""},{"path":"https://www.slendr.net/reference/check_env.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Check that the active Python environment is setup for slendr — check_env","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated. check_env() #> Summary of the currently active Python environment: #>  #> Python binary: /Users/mp/Library/r-miniconda-arm64/envs/Python-3.12_msprime-1.3.1_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2/bin/python  #> Python version: 3.12.2 | packaged by conda-forge | (main, Feb 16 2024, 20:54:21) [Clang 16.0.6 ]  #>  #> slendr requirements: #>  - tskit: version 0.5.6 ✓  #>  - msprime: version 1.3.1 ✓  #>  - pyslim: version 1.0.4 ✓  #>  - tspop: present ✓"},{"path":"https://www.slendr.net/reference/clear_env.html","id":null,"dir":"Reference","previous_headings":"","what":"Remove the automatically created slendr Python environment — clear_env","title":"Remove the automatically created slendr Python environment — clear_env","text":"Remove automatically created slendr Python environment","code":""},{"path":"https://www.slendr.net/reference/clear_env.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Remove the automatically created slendr Python environment — clear_env","text":"","code":"clear_env(force = FALSE, all = FALSE)"},{"path":"https://www.slendr.net/reference/clear_env.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Remove the automatically created slendr Python environment — clear_env","text":"force Ask deleting environment? (present past) slendr Python environments removed (default FALSE) just current environment?","code":""},{"path":"https://www.slendr.net/reference/clear_env.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Remove the automatically created slendr Python environment — clear_env","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/compile_model.html","id":null,"dir":"Reference","previous_headings":"","what":"Compile a slendr demographic model — compile_model","title":"Compile a slendr demographic model — compile_model","text":"First, compiles vectorized population spatial maps series binary raster PNG files, format SLiM understands uses define population boundaries. extracts demographic model defined user (.e. population divergences gene flow events) series tables later used built-SLiM script program timing simulation events.","code":""},{"path":"https://www.slendr.net/reference/compile_model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compile a slendr demographic model — compile_model","text":"","code":"compile_model(   populations,   generation_time,   path = NULL,   resolution = NULL,   competition = NULL,   mating = NULL,   dispersal = NULL,   gene_flow = list(),   overwrite = FALSE,   force = FALSE,   simulation_length = NULL,   direction = NULL,   serialize = TRUE,   slim_script = NULL,   description = \"\",   sim_length = NULL )"},{"path":"https://www.slendr.net/reference/compile_model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compile a slendr demographic model — compile_model","text":"populations Object(s) slendr_pop class (multiple objects need specified list) generation_time Generation time (model time units) path Output directory model configuration files loaded backend SLiM script. NULL, model configuration files saved temporary directory. resolution many distance units per pixel? competition, mating Maximum spatial competition mating choice distance dispersal Standard deviation normal distribution parent-offspring distance gene_flow Gene flow events generated gene_flow function (either list data.frame objects format defined gene_flow function, single data.frame) overwrite Completely delete specified directory, case already exists, create new one? force Force deletion model directory already present? Useful non-interactive uses. interactive mode, user asked confirm deletion manually. simulation_length Total length simulation (required forward time models, optional models specified backward time units default run \"present time\") direction Intended direction time. normal circumstances parameter inferred model need set manually. serialize model files serialized disk? , R model object returned files created. speeds simulation msprime prevents using SLiM back end. slim_script Path SLiM script used executing model (default, bundled backend script used). NULL, SLiM script bundled slendr used. description Optional short description model sim_length Deprecated. Use simulation_length instead.","code":""},{"path":"https://www.slendr.net/reference/compile_model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Compile a slendr demographic model — compile_model","text":"Compiled slendr_model model object encapsulates information specified model (populations involved, much gene flow occur, spatial resolution map, spatial dispersal mating parameters used SLiM simulation, applicable)","code":""},{"path":"https://www.slendr.net/reference/compile_model.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Compile a slendr demographic model — compile_model","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/distance.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the distance between a pair of spatial boundaries — distance","title":"Calculate the distance between a pair of spatial boundaries — distance","text":"Calculate distance pair spatial boundaries","code":""},{"path":"https://www.slendr.net/reference/distance.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the distance between a pair of spatial boundaries — distance","text":"","code":"distance(x, y, measure, time = NULL)"},{"path":"https://www.slendr.net/reference/distance.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the distance between a pair of spatial boundaries — distance","text":"x, y Objects class slendr measure measure distance? can either 'border' (distance borders x y) 'center' (distance centroids). time Time closest spatial maps x y represent slendr_pop population boundaries (ignored general slendr_region objects)","code":""},{"path":"https://www.slendr.net/reference/distance.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the distance between a pair of spatial boundaries — distance","text":"coordinate reference system specified, distance projected units (.e. meters) returned. Otherwise function returns normal Euclidean distance.","code":""},{"path":"https://www.slendr.net/reference/distance.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the distance between a pair of spatial boundaries — distance","text":"","code":"# create two regions on a blank abstract landscape region_a <- region(\"A\", center = c(20, 50), radius = 20) region_b <- region(\"B\", center = c(80, 50), radius = 20) plot_map(region_a, region_b)   # compute the distance between the centers of both population ranges distance(region_a, region_b, measure = \"center\") #> [1] 60  # compute the distance between the borders of both population ranges distance(region_a, region_b, measure = \"border\") #> [1] 20"},{"path":"https://www.slendr.net/reference/expand_range.html","id":null,"dir":"Reference","previous_headings":"","what":"Expand the population range — expand_range","title":"Expand the population range — expand_range","text":"Expands spatial population range specified distance given time-window","code":""},{"path":"https://www.slendr.net/reference/expand_range.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Expand the population range — expand_range","text":"","code":"expand_range(   pop,   by,   end,   start,   overlap = 0.8,   snapshots = NULL,   polygon = NULL,   lock = FALSE,   verbose = TRUE )"},{"path":"https://www.slendr.net/reference/expand_range.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Expand the population range — expand_range","text":"pop Object class slendr_pop many units distance expand ? start, end expansion start/end? overlap Minimum overlap subsequent spatial boundaries snapshots number intermediate snapshots (overrides overlap parameter) polygon Geographic region restrict expansion lock Maintain density individuals. FALSE (default), number individuals population change. TRUE, number individuals simulated changed (increased decreased) appropriately, match new population range area. verbose Report progress generating intermediate spatial boundaries?","code":""},{"path":"https://www.slendr.net/reference/expand_range.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Expand the population range — expand_range","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/expand_range.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Expand the population range — expand_range","text":"Note slendr models accomodate SLiM msprime back ends, population sizes times events rounded nearest integer value.","code":""},{"path":"https://www.slendr.net/reference/expand_range.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Expand the population range — expand_range","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/explore_model.html","id":null,"dir":"Reference","previous_headings":"","what":"Open an interactive browser of the spatial model — explore_model","title":"Open an interactive browser of the spatial model — explore_model","text":"Open interactive browser spatial model","code":""},{"path":"https://www.slendr.net/reference/explore_model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Open an interactive browser of the spatial model — explore_model","text":"","code":"explore_model(model)"},{"path":"https://www.slendr.net/reference/explore_model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Open an interactive browser of the spatial model — explore_model","text":"model Compiled slendr_model model object","code":""},{"path":"https://www.slendr.net/reference/explore_model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Open an interactive browser of the spatial model — explore_model","text":"return value, called order start interactive browser-based interface explore dynamics slendr model","code":""},{"path":"https://www.slendr.net/reference/gene_flow.html","id":null,"dir":"Reference","previous_headings":"","what":"Define a gene-flow event between two populations — gene_flow","title":"Define a gene-flow event between two populations — gene_flow","text":"Define gene-flow event two populations","code":""},{"path":"https://www.slendr.net/reference/gene_flow.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Define a gene-flow event between two populations — gene_flow","text":"","code":"gene_flow(from, to, rate, start, end, overlap = TRUE)"},{"path":"https://www.slendr.net/reference/gene_flow.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Define a gene-flow event between two populations — gene_flow","text":", Objects class slendr_pop rate Scalar value range (0, 1] specifying proportion migration given time period start, end Start end gene-flow event overlap Require spatial overlap admixing populations?  (default TRUE)","code":""},{"path":"https://www.slendr.net/reference/gene_flow.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Define a gene-flow event between two populations — gene_flow","text":"Object class data.frame containing parameters specified gene-flow event.","code":""},{"path":"https://www.slendr.net/reference/gene_flow.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Define a gene-flow event between two populations — gene_flow","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/get_env.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the name of the current slendr Python environment — get_env","title":"Get the name of the current slendr Python environment — get_env","text":"Get name current slendr Python environment","code":""},{"path":"https://www.slendr.net/reference/get_env.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the name of the current slendr Python environment — get_env","text":"","code":"get_env()"},{"path":"https://www.slendr.net/reference/get_env.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the name of the current slendr Python environment — get_env","text":"Name slendr Python environment","code":""},{"path":"https://www.slendr.net/reference/init_env.html","id":null,"dir":"Reference","previous_headings":"","what":"Activate slendr's own dedicated Python environment — init_env","title":"Activate slendr's own dedicated Python environment — init_env","text":"function attempts activate dedicated slendr Miniconda Python environment previously set via setup_env.","code":""},{"path":"https://www.slendr.net/reference/init_env.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Activate slendr's own dedicated Python environment — init_env","text":"","code":"init_env(quiet = FALSE)"},{"path":"https://www.slendr.net/reference/init_env.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Activate slendr's own dedicated Python environment — init_env","text":"quiet informative messages printed console? Default FALSE.","code":""},{"path":"https://www.slendr.net/reference/init_env.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Activate slendr's own dedicated Python environment — init_env","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/join.html","id":null,"dir":"Reference","previous_headings":"","what":"Merge two spatial slendr objects into one — join","title":"Merge two spatial slendr objects into one — join","text":"Merge two spatial slendr objects one","code":""},{"path":"https://www.slendr.net/reference/join.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Merge two spatial slendr objects into one — join","text":"","code":"join(x, y, name = NULL)"},{"path":"https://www.slendr.net/reference/join.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Merge two spatial slendr objects into one — join","text":"x Object class slendr y Object class slendr name Optional name resulting geographic region. missing, name constructed function arguments.","code":""},{"path":"https://www.slendr.net/reference/join.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Merge two spatial slendr objects into one — join","text":"Object class slendr_region encodes standard spatial object class sf several additional attributes (importantly corresponding slendr_map object, applicable).","code":""},{"path":"https://www.slendr.net/reference/join.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Merge two spatial slendr objects into one — join","text":"","code":"# create a blank abstract world 1000x1000 distance units in size blank_map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # it is possible to construct custom landscapes (islands, corridors, etc.) island1 <- region(\"island1\", polygon = list(c(10, 30), c(50, 30), c(40, 50), c(0, 40))) island2 <- region(\"island2\", polygon = list(c(60, 60), c(80, 40), c(100, 60), c(80, 80))) island3 <- region(\"island3\", center = c(20, 80), radius = 10) archipelago <- island1 %>% join(island2) %>% join(island3)  custom_map <- world(xrange = c(1, 100), c(1, 100), landscape = archipelago)  # real Earth landscapes can be defined using freely-available Natural Earth # project data and with the possibility to specify an appropriate Coordinate # Reference System, such as this example of a map of Europe  real_map <- world(xrange = c(-15, 40), yrange = c(30, 60), crs = \"EPSG:3035\") #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpqG5m0G/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84"},{"path":"https://www.slendr.net/reference/move.html","id":null,"dir":"Reference","previous_headings":"","what":"Move the population to a new location in a given amount of time — move","title":"Move the population to a new location in a given amount of time — move","text":"function defines displacement population along given trajectory given time frame","code":""},{"path":"https://www.slendr.net/reference/move.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Move the population to a new location in a given amount of time — move","text":"","code":"move(   pop,   trajectory,   end,   start,   overlap = 0.8,   snapshots = NULL,   verbose = TRUE )"},{"path":"https://www.slendr.net/reference/move.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Move the population to a new location in a given amount of time — move","text":"pop Object class slendr_pop trajectory List two-dimensional vectors (longitude, latitude) specifying migration trajectory start, end Start/end points population migration overlap Minimum overlap subsequent spatial boundaries snapshots number intermediate snapshots (overrides overlap parameter) verbose Show progress searching number sufficient snapshots?","code":""},{"path":"https://www.slendr.net/reference/move.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Move the population to a new location in a given amount of time — move","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/move.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Move the population to a new location in a given amount of time — move","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/msprime.html","id":null,"dir":"Reference","previous_headings":"","what":"Run a slendr model in msprime — msprime","title":"Run a slendr model in msprime — msprime","text":"function execute built-msprime script run compiled slendr demographic model.","code":""},{"path":"https://www.slendr.net/reference/msprime.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Run a slendr model in msprime — msprime","text":"","code":"msprime(   model,   sequence_length,   recombination_rate,   samples = NULL,   output = NULL,   random_seed = NULL,   load = TRUE,   verbose = FALSE,   debug = FALSE,   run = TRUE )"},{"path":"https://www.slendr.net/reference/msprime.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Run a slendr model in msprime — msprime","text":"model Model object created compile function sequence_length Total length simulated sequence (base-pairs) recombination_rate Recombination rate simulated sequence (recombinations per basepair per generation) samples data frame times given number individuals remembered tree-sequence (see schedule_sampling function can generate sampling schedule correct format). missing, individuals present end simulation recorded tree-sequence output file. output Path output tree-sequence file. NULL (default), tree sequence saved temporary file. random_seed Random seed (NULL, seed generated 0 maximum integer number available) load final tree sequence immediately loaded returned? Default TRUE. alternative (FALSE) useful tree-sequence file written custom location loaded later point. verbose Write output log console (default FALSE)? debug Write msprime's debug log console (default FALSE)? run msprime engine run? FALSE, command line msprime command printed (returned invisibly character vector) executed.","code":""},{"path":"https://www.slendr.net/reference/msprime.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Run a slendr model in msprime — msprime","text":"tree-sequence object loaded via Python-R reticulate interface function ts_load (internally represented Python object tskit.trees.TreeSequence). Optionally, depending value arguments load = run =, nothing character vector, respectively.","code":""},{"path":"https://www.slendr.net/reference/msprime.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Run a slendr model in msprime — msprime","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # afr and eur objects would normally be created before slendr model compilation, # but here we take them out of the model object already compiled for this # example (in a standard slendr simulation pipeline, this wouldn't be necessary) afr <- model$populations[[\"AFR\"]] eur <- model$populations[[\"EUR\"]] chimp <- model$populations[[\"CH\"]]  # schedule the sampling of a couple of ancient and present-day individuals # given model at 20 ky, 10 ky, 5ky ago and at present-day (time 0) modern_samples <- schedule_sampling(model, times = 0, list(afr, 10), list(eur, 100), list(chimp, 1)) ancient_samples <- schedule_sampling(model, times = c(40000, 30000, 20000, 10000), list(eur, 1))  # sampling schedules are just data frames and can be merged easily samples <- rbind(modern_samples, ancient_samples)  # run a simulation using the msprime back end from a compiled slendr model object ts <- msprime(model, sequence_length = 1e5, recombination_rate = 0, samples = samples)  # automatic loading of a simulated output can be prevented by `load = FALSE`, which can be # useful when a custom path to a tree-sequence output is given for later downstream analyses output_file <- tempfile(fileext = \".trees\") msprime(model, sequence_length = 1e5, recombination_rate = 0, samples = samples,         output = output_file, load = FALSE, random_seed = 42) # ... at a later stage: ts <- ts_load(output_file, model)  summary(ts) #> ╔═══════════════════════════╗ #> ║TreeSequence               ║ #> ╠═══════════════╤═══════════╣ #> ║Trees          │          1║ #> ╟───────────────┼───────────╢ #> ║Sequence Length│     100000║ #> ╟───────────────┼───────────╢ #> ║Time Units     │generations║ #> ╟───────────────┼───────────╢ #> ║Sample Nodes   │        230║ #> ╟───────────────┼───────────╢ #> ║Total Size     │   39.4 KiB║ #> ╚═══════════════╧═══════════╝ #> ╔═══════════╤════╤═════════╤════════════╗ #> ║Table      │Rows│Size     │Has Metadata║ #> ╠═══════════╪════╪═════════╪════════════╣ #> ║Edges      │ 471│ 14.7 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Individuals│ 115│  3.2 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Migrations │   0│  8 Bytes│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Mutations  │   0│ 16 Bytes│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Nodes      │ 472│ 12.9 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Populations│   4│338 Bytes│         Yes║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Provenances│   1│  2.9 KiB│          No║ #> ╟───────────┼────┼─────────┼────────────╢ #> ║Sites      │   0│ 16 Bytes│          No║ #> ╚═══════════╧════╧═════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/overlap.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate the overlap of two slendr objects — overlap","title":"Generate the overlap of two slendr objects — overlap","text":"Generate overlap two slendr objects","code":""},{"path":"https://www.slendr.net/reference/overlap.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate the overlap of two slendr objects — overlap","text":"","code":"overlap(x, y, name = NULL)"},{"path":"https://www.slendr.net/reference/overlap.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate the overlap of two slendr objects — overlap","text":"x Object class slendr y Object class slendr name Optional name resulting geographic region. missing, name constructed function arguments.","code":""},{"path":"https://www.slendr.net/reference/overlap.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generate the overlap of two slendr objects — overlap","text":"Object class slendr_region encodes standard spatial object class sf several additional attributes (importantly corresponding slendr_map object, applicable).","code":""},{"path":"https://www.slendr.net/reference/pipe.html","id":null,"dir":"Reference","previous_headings":"","what":"Pipe operator — %>%","title":"Pipe operator — %>%","text":"Pipe operator","code":""},{"path":"https://www.slendr.net/reference/pipe.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Pipe operator — %>%","text":"","code":"lhs %>% rhs"},{"path":"https://www.slendr.net/reference/pipe.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Pipe operator — %>%","text":"See magrittr::%>% details.","code":""},{"path":"https://www.slendr.net/reference/plot_map.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot slendr geographic features on a map — plot_map","title":"Plot slendr geographic features on a map — plot_map","text":"Plots objects three slendr spatial classes (slendr_map, slendr_region, slendr_pop).","code":""},{"path":"https://www.slendr.net/reference/plot_map.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot slendr geographic features on a map — plot_map","text":"","code":"plot_map(   ...,   time = NULL,   gene_flow = FALSE,   graticules = \"original\",   intersect = TRUE,   show_map = TRUE,   title = NULL,   interpolated_maps = NULL )"},{"path":"https://www.slendr.net/reference/plot_map.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot slendr geographic features on a map — plot_map","text":"... Objects classes slendr_map, slendr_region, slendr_pop time Plot concrete time point gene_flow Indicate geneflow events arrow graticules Plot graticules original Coordinate Reference System (longitude-latitude), internal CRS (meters)? intersect Intersect population boundaries landscape geographic boundaries (default TRUE)? show_map Show underlying world map title Title plot interpolated_maps Interpolated spatial boundaries populations time points (used plotting using explore shiny app)","code":""},{"path":"https://www.slendr.net/reference/plot_map.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot slendr geographic features on a map — plot_map","text":"ggplot2 object visualized slendr map","code":""},{"path":"https://www.slendr.net/reference/plot_model.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot demographic history encoded in a slendr model — plot_model","title":"Plot demographic history encoded in a slendr model — plot_model","text":"Plot demographic history encoded slendr model","code":""},{"path":"https://www.slendr.net/reference/plot_model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot demographic history encoded in a slendr model — plot_model","text":"","code":"plot_model(   model,   sizes = TRUE,   proportions = FALSE,   gene_flow = TRUE,   log = FALSE,   order = NULL,   file = NULL,   samples = NULL,   ... )"},{"path":"https://www.slendr.net/reference/plot_model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot demographic history encoded in a slendr model — plot_model","text":"model Compiled slendr_model model object sizes population size changes visualized? proportions gene flow proportions visualized (FALSE default prevent cluttering overplotting) gene_flow gene-flow arrows visualized (default TRUE). log y-axis plotted log scale? Useful models long time-scales. order Order populations along x-axis, given character vector population names. NULL (default), default plotting algorithm used, ordering populations ancestral recent using -order tree traversal. file Output file figure saved via ggsave samples Sampling schedule visualized model ... Optional argument passed ggsave","code":""},{"path":"https://www.slendr.net/reference/plot_model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot demographic history encoded in a slendr model — plot_model","text":"ggplot2 object visualized slendr model","code":""},{"path":"https://www.slendr.net/reference/plot_model.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot demographic history encoded in a slendr model — plot_model","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence path <- system.file(\"extdata/models/introgression\", package = \"slendr\") model <- read_model(path)  plot_model(model, sizes = FALSE, log = TRUE)"},{"path":"https://www.slendr.net/reference/population.html","id":null,"dir":"Reference","previous_headings":"","what":"Define a population — population","title":"Define a population — population","text":"Defines parameters population (non-spatial spatial).","code":""},{"path":"https://www.slendr.net/reference/population.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Define a population — population","text":"","code":"population(   name,   time,   N,   parent = NULL,   map = FALSE,   center = NULL,   radius = NULL,   polygon = NULL,   remove = NULL,   intersect = TRUE,   competition = NA,   mating = NA,   dispersal = NA,   dispersal_fun = NULL,   aquatic = FALSE )"},{"path":"https://www.slendr.net/reference/population.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Define a population — population","text":"name Name population time Time population's first appearance N Number individuals time first appearance parent Parent population object NULL (indicates population ancestor, first population \"lineage\") map Object type slendr_map defines world context (created using world function). value FALSE provided, non-spatial model run. center Two-dimensional vector specifying center circular range radius Radius circular range polygon List vector pairs, defining corners polygon range geographic region class slendr_region polygon coordinates extracted (see region() function) remove Time population removed intersect Intersect population's boundaries landscape features? competition, mating Maximum spatial competition mating choice distance dispersal Standard deviation normal distribution distance offspring disperses parent dispersal_fun Distribution function governing dispersal offspring. One \"normal\", \"uniform\", \"cauchy\", \"exponential\", \"brownian\" (vertical horizontal displacements drawn normal distribution independently). aquatic species aquatic (FALSE default, .e. terrestrial species)?","code":""},{"path":"https://www.slendr.net/reference/population.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Define a population — population","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/population.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Define a population — population","text":"four ways specify spatial boundary: ) circular range specified using center coordinate radius, ii) polygon specified list two-dimensional vector coordinates, iii) polygon ii), defined (named) using region function, iv) just world map specified (circular polygon range parameters set default NULL value), population allowed occupy entire landscape. Note slendr models accomodate SLiM msprime back ends, population sizes split times rounded nearest integer value.","code":""},{"path":"https://www.slendr.net/reference/population.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Define a population — population","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/print.slendr_pop.html","id":null,"dir":"Reference","previous_headings":"","what":"Print a short summary of a slendr object — print.slendr_pop","title":"Print a short summary of a slendr object — print.slendr_pop","text":"spatial objects slendr package internally represented Simple Features (sf) objects. fact hidden circumstances , goal slendr package provide functionality much higher level (population boundaries, geographic regions, instead individual polygons \"low-level\" geometric objects), without users worry low-level details involved handling spatial geometries. However, full sf object representation can always printed calling x[].","code":""},{"path":"https://www.slendr.net/reference/print.slendr_pop.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print a short summary of a slendr object — print.slendr_pop","text":"","code":"# S3 method for slendr_pop print(x, ...)  # S3 method for slendr_region print(x, ...)  # S3 method for slendr_map print(x, ...)  # S3 method for slendr_model print(x, ...)"},{"path":"https://www.slendr.net/reference/print.slendr_pop.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print a short summary of a slendr object — print.slendr_pop","text":"x Object class slendr (either slendr_pop, slendr_map, slendr_region, slendr_table) ... Additional arguments passed print","code":""},{"path":"https://www.slendr.net/reference/print.slendr_pop.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print a short summary of a slendr object — print.slendr_pop","text":"return value, used printing","code":""},{"path":"https://www.slendr.net/reference/print.slendr_ts.html","id":null,"dir":"Reference","previous_headings":"","what":"Print tskit's summary table of the Python tree-sequence object — print.slendr_ts","title":"Print tskit's summary table of the Python tree-sequence object — print.slendr_ts","text":"Print tskit's summary table Python tree-sequence object","code":""},{"path":"https://www.slendr.net/reference/print.slendr_ts.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Print tskit's summary table of the Python tree-sequence object — print.slendr_ts","text":"","code":"# S3 method for slendr_ts print(x, ...)"},{"path":"https://www.slendr.net/reference/print.slendr_ts.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Print tskit's summary table of the Python tree-sequence object — print.slendr_ts","text":"x Tree object class slendr_phylo ... Additional arguments normally passed print (used case)","code":""},{"path":"https://www.slendr.net/reference/print.slendr_ts.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Print tskit's summary table of the Python tree-sequence object — print.slendr_ts","text":"return value, simply prints tskit summary table terminal","code":""},{"path":"https://www.slendr.net/reference/read_model.html","id":null,"dir":"Reference","previous_headings":"","what":"Read a previously serialized model configuration — read_model","title":"Read a previously serialized model configuration — read_model","text":"Reads configuration tables model data location previously compiled compile function.","code":""},{"path":"https://www.slendr.net/reference/read_model.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Read a previously serialized model configuration — read_model","text":"","code":"read_model(path)"},{"path":"https://www.slendr.net/reference/read_model.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Read a previously serialized model configuration — read_model","text":"path Directory required configuration files","code":""},{"path":"https://www.slendr.net/reference/read_model.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Read a previously serialized model configuration — read_model","text":"Compiled slendr_model model object encapsulates information specified model (populations involved, much gene flow occur, spatial resolution map, spatial dispersal mating parameters used SLiM simulation, applicable)","code":""},{"path":"https://www.slendr.net/reference/read_model.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Read a previously serialized model configuration — read_model","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence path <- system.file(\"extdata/models/introgression\", package = \"slendr\") model <- read_model(path)  plot_model(model, sizes = FALSE, log = TRUE)"},{"path":"https://www.slendr.net/reference/region.html","id":null,"dir":"Reference","previous_headings":"","what":"Define a geographic region — region","title":"Define a geographic region — region","text":"Creates geographic region (polygon) given map gives name. can used define objects can reused multiple places slendr script (region arguments population) without repeatedly define polygon coordinates.","code":""},{"path":"https://www.slendr.net/reference/region.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Define a geographic region — region","text":"","code":"region(name = NULL, map = NULL, center = NULL, radius = NULL, polygon = NULL)"},{"path":"https://www.slendr.net/reference/region.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Define a geographic region — region","text":"name Name geographic region map Object type sf defines map center Two-dimensional vector specifying center circular range radius Radius circular range polygon List vector pairs, defining corners polygon range geographic region class slendr_region polygon coordinates extracted (see region() function)","code":""},{"path":"https://www.slendr.net/reference/region.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Define a geographic region — region","text":"Object class slendr_region encodes standard spatial object class sf several additional attributes (importantly corresponding slendr_map object, applicable).","code":""},{"path":"https://www.slendr.net/reference/region.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Define a geographic region — region","text":"","code":"# create a blank abstract world 1000x1000 distance units in size blank_map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # it is possible to construct custom landscapes (islands, corridors, etc.) island1 <- region(\"island1\", polygon = list(c(10, 30), c(50, 30), c(40, 50), c(0, 40))) island2 <- region(\"island2\", polygon = list(c(60, 60), c(80, 40), c(100, 60), c(80, 80))) island3 <- region(\"island3\", center = c(20, 80), radius = 10) archipelago <- island1 %>% join(island2) %>% join(island3)  custom_map <- world(xrange = c(1, 100), c(1, 100), landscape = archipelago)  # real Earth landscapes can be defined using freely-available Natural Earth # project data and with the possibility to specify an appropriate Coordinate # Reference System, such as this example of a map of Europe  real_map <- world(xrange = c(-15, 40), yrange = c(30, 60), crs = \"EPSG:3035\") #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpqG5m0G/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84"},{"path":"https://www.slendr.net/reference/reproject.html","id":null,"dir":"Reference","previous_headings":"","what":"Reproject coordinates between coordinate systems — reproject","title":"Reproject coordinates between coordinate systems — reproject","text":"Converts coordinates compiled raster map (.e. pixel units) different Geographic Coordinate Systems (CRS).","code":""},{"path":"https://www.slendr.net/reference/reproject.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Reproject coordinates between coordinate systems — reproject","text":"","code":"reproject(   from,   to,   x = NULL,   y = NULL,   coords = NULL,   model = NULL,   add = FALSE,   input_prefix = \"\",   output_prefix = \"new\" )"},{"path":"https://www.slendr.net/reference/reproject.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Reproject coordinates between coordinate systems — reproject","text":", Either CRS code accepted GDAL, valid integer EPSG value, object class crs, value \"raster\" (converting /pixel coordinates), \"world\" (converting /whatever CRS set underlying map) x, y Coordinates two dimensions (missing, coordinates expected data.frame specified coords parameter columns \"x\" \"y\") coords data.frame-like object coordinates columns \"x\" \"y\" model Object class slendr_model add Add column coordinates input data.frame coords (coordinates otherwise returned separate object)? input_prefix, output_prefix Input output prefixes data frame columns spatial coordinates","code":""},{"path":"https://www.slendr.net/reference/reproject.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Reproject coordinates between coordinate systems — reproject","text":"Data.frame converted two-dimensional coordinates given input","code":""},{"path":"https://www.slendr.net/reference/reproject.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Reproject coordinates between coordinate systems — reproject","text":"","code":"lon_lat_df <- data.frame(x = c(30, 0, 15), y = c(60, 40, 10))  reproject(   from = \"epsg:4326\",   to = \"epsg:3035\",   coords = lon_lat_df,   add = TRUE # add converted [lon,lat] coordinates as a new column ) #> # A tibble: 3 × 4 #>       x     y     newx      newy #>   <dbl> <dbl>    <dbl>     <dbl> #> 1    30    60 5422493.  4256803. #> 2     0    40 3465349.  1934879. #> 3    15    10 4907297. -1328914."},{"path":"https://www.slendr.net/reference/resize.html","id":null,"dir":"Reference","previous_headings":"","what":"Change the population size — resize","title":"Change the population size — resize","text":"Resizes population starting current value N individuals specified value","code":""},{"path":"https://www.slendr.net/reference/resize.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Change the population size — resize","text":"","code":"resize(pop, N, how, time, end = NULL)"},{"path":"https://www.slendr.net/reference/resize.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Change the population size — resize","text":"pop Object class slendr_pop N Population size change change population size (options \"step\" \"exponential\") time Time population size change end End population size change period (used exponential change events)","code":""},{"path":"https://www.slendr.net/reference/resize.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Change the population size — resize","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/resize.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Change the population size — resize","text":"case exponential size change, final N larger current size, population exponentially growing specified time period reaches N individuals. N smaller, population shrink exponentially. Note slendr models accomodate SLiM msprime back ends, population sizes split times rounded nearest integer value.","code":""},{"path":"https://www.slendr.net/reference/resize.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Change the population size — resize","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":null,"dir":"Reference","previous_headings":"","what":"Define sampling events for a given set of populations — schedule_sampling","title":"Define sampling events for a given set of populations — schedule_sampling","text":"Schedule sampling events specified times , optionally, given set locations landscape","code":""},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Define sampling events for a given set of populations — schedule_sampling","text":"","code":"schedule_sampling(model, times, ..., locations = NULL, strict = FALSE)"},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Define sampling events for a given set of populations — schedule_sampling","text":"model Object class slendr_model times Integer vector times (model time units) schedule remembering individuals tree-sequence ... Lists two elements (slendr_pop population object-<number individuals sample), representing populations many individuals remembered times given times locations List vector pairs, defining two-dimensional coordinates locations closest number individuals given populations sampled. NULL (default), individuals sampled randomly throughout spatial boundary. strict occurence population present given time result error? Default FALSE, meaning invalid sampling times populations quietly ignored.","code":""},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Define sampling events for a given set of populations — schedule_sampling","text":"Data frame three columns: time sampling, population sample , many individuals sample","code":""},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Define sampling events for a given set of populations — schedule_sampling","text":"times locations given, sampling scheduled specified location given time-point. Note time-, interest simplicity, sanity checks performed locations given except restriction sampling points must fall within bounding box around simulated world map. , slendr simply instruct SLiM backend script sample individuals close sampling points given possible, regardless whether points lie within population spatial boundary particular moment time.","code":""},{"path":"https://www.slendr.net/reference/schedule_sampling.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Define sampling events for a given set of populations — schedule_sampling","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence path <- system.file(\"extdata/models/introgression\", package = \"slendr\") model <- read_model(path)  # afr and eur objects would normally be created before slendr model compilation, # but here we take them out of the model object already compiled for this # example (in a standard slendr simulation pipeline, this wouldn't be necessary) afr <- model$populations[[\"AFR\"]] eur <- model$populations[[\"EUR\"]]  # schedule the recording of 10 African and 100 European individuals from a # given model at 20 ky, 10 ky, 5ky ago and at present-day (time 0) schedule <- schedule_sampling(   model, times = c(20000, 10000, 5000, 0),   list(afr, 10), list(eur, 100) )  # the result of `schedule_sampling` is a simple data frame (note that the locations # of sampling locations have `NA` values because the model is non-spatial) schedule #> # A tibble: 8 × 7 #>    time pop       n y_orig x_orig y     x     #>   <int> <chr> <int> <lgl>  <lgl>  <lgl> <lgl> #> 1     0 AFR      10 NA     NA     NA    NA    #> 2     0 EUR     100 NA     NA     NA    NA    #> 3  5000 AFR      10 NA     NA     NA    NA    #> 4  5000 EUR     100 NA     NA     NA    NA    #> 5 10000 AFR      10 NA     NA     NA    NA    #> 6 10000 EUR     100 NA     NA     NA    NA    #> 7 20000 AFR      10 NA     NA     NA    NA    #> 8 20000 EUR     100 NA     NA     NA    NA"},{"path":"https://www.slendr.net/reference/set_dispersal.html","id":null,"dir":"Reference","previous_headings":"","what":"Change dispersal parameters — set_dispersal","title":"Change dispersal parameters — set_dispersal","text":"Changes either competition interactive distance, mating choice distance, dispersal offspring parent","code":""},{"path":"https://www.slendr.net/reference/set_dispersal.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Change dispersal parameters — set_dispersal","text":"","code":"set_dispersal(   pop,   time,   competition = NA,   mating = NA,   dispersal = NA,   dispersal_fun = NULL )"},{"path":"https://www.slendr.net/reference/set_dispersal.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Change dispersal parameters — set_dispersal","text":"pop Object class slendr_pop time Time population size change competition, mating Maximum spatial competition mating choice distance dispersal Standard deviation normal distribution distance offspring disperses parent dispersal_fun Distribution function governing dispersal offspring. One \"normal\", \"uniform\", \"cauchy\", \"exponential\", \"brownian\" (vertical horizontal displacements drawn normal distribution independently).","code":""},{"path":"https://www.slendr.net/reference/set_dispersal.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Change dispersal parameters — set_dispersal","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/set_dispersal.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Change dispersal parameters — set_dispersal","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/set_range.html","id":null,"dir":"Reference","previous_headings":"","what":"Update the population range — set_range","title":"Update the population range — set_range","text":"function allows manual control spatial map changes addition expand move functions","code":""},{"path":"https://www.slendr.net/reference/set_range.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Update the population range — set_range","text":"","code":"set_range(   pop,   time,   center = NULL,   radius = NULL,   polygon = NULL,   lock = FALSE )"},{"path":"https://www.slendr.net/reference/set_range.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Update the population range — set_range","text":"pop Object class slendr_pop time Time change center Two-dimensional vector specifying center circular range radius Radius circular range polygon List vector pairs, defining corners polygon range (see also region argument) geographic region class slendr_region polygon coordinates extracted lock Maintain density individuals. FALSE (default), number individuals population change. TRUE, number individuals simulated changed (increased decreased) appropriately, match new population range area.","code":""},{"path":"https://www.slendr.net/reference/set_range.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Update the population range — set_range","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/set_range.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Update the population range — set_range","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/setup_env.html","id":null,"dir":"Reference","previous_headings":"","what":"Setup a dedicated Python virtual environment for slendr — setup_env","title":"Setup a dedicated Python virtual environment for slendr — setup_env","text":"function automatically download Python miniconda distribution dedicated R-Python interface. also create slendr-specific Python environment required Python dependencies.","code":""},{"path":"https://www.slendr.net/reference/setup_env.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Setup a dedicated Python virtual environment for slendr — setup_env","text":"","code":"setup_env(quiet = FALSE, agree = FALSE, pip = NULL)"},{"path":"https://www.slendr.net/reference/setup_env.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Setup a dedicated Python virtual environment for slendr — setup_env","text":"quiet informative messages printed console? Default FALSE. agree Automatically agree questions? pip pip used instead conda installing slendr's Python dependencies? Note still use conda distribution install Python , change repository slendr install Python dependencies. Unless explicitly set TRUE, Python dependencies installed conda repositories default, expect case osx-arm64 Mac architecture, conda dependencies broken.","code":""},{"path":"https://www.slendr.net/reference/setup_env.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Setup a dedicated Python virtual environment for slendr — setup_env","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/shrink_range.html","id":null,"dir":"Reference","previous_headings":"","what":"Shrink the population range — shrink_range","title":"Shrink the population range — shrink_range","text":"Shrinks spatial population range specified distance given time-window","code":""},{"path":"https://www.slendr.net/reference/shrink_range.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Shrink the population range — shrink_range","text":"","code":"shrink_range(   pop,   by,   end,   start,   overlap = 0.8,   snapshots = NULL,   lock = FALSE,   verbose = TRUE )"},{"path":"https://www.slendr.net/reference/shrink_range.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Shrink the population range — shrink_range","text":"pop Object class slendr_pop many units distance shrink ? start, end boundary shrinking start/end? overlap Minimum overlap subsequent spatial boundaries snapshots number intermediate snapshots (overrides overlap parameter) lock Maintain density individuals. FALSE (default), number individuals population change. TRUE, number individuals simulated changed (increased decreased) appropriately, match new population range area. verbose Report progress generating intermediate spatial boundaries?","code":""},{"path":"https://www.slendr.net/reference/shrink_range.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Shrink the population range — shrink_range","text":"Object class slendr_pop, contains population parameters name, time appearance simulation, parent population (), spatial parameters map spatial boundary.","code":""},{"path":"https://www.slendr.net/reference/shrink_range.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Shrink the population range — shrink_range","text":"Note slendr models accomodate SLiM msprime back ends, population sizes split times rounded nearest integer value.","code":""},{"path":"https://www.slendr.net/reference/shrink_range.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Shrink the population range — shrink_range","text":"","code":"# spatial definitions -----------------------------------------------------  # create a blank abstract world 1000x1000 distance units in size map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # create a circular population with the center of a population boundary at # [200, 800] and a radius of 100 distance units, 1000 individuals at time 1 # occupying a map just specified pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100)  # printing a population object to a console shows a brief summary pop1 #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 1  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000)  # create another population occupying a polygon range, splitting from pop1 # at a given time point (note that specifying a map is not necessary because # it is \"inherited\" from the parent) pop2 <- population(\"pop2\", N = 100, time = 50, parent = pop1,                         polygon = list(c(100, 100), c(320, 30), c(500, 200),                                   c(500, 400), c(300, 450), c(100, 400)))  pop3 <- population(\"pop3\", N = 200, time = 80, parent = pop2,                    center = c(800, 800), radius = 200)  # move \"pop1\" to another location along a specified trajectory and saved the # resulting object to the same variable (the number of intermediate spatial # snapshots can be also determined automatically by leaving out the # `snapshots = ` argument) pop1_moved <- move(pop1, start = 100, end = 200, snapshots = 6,                    trajectory = list(c(600, 820), c(800, 400), c(800, 150))) pop1_moved #> slendr 'population' object  #> --------------------------  #> name: pop1  #> habitat: terrestrial #>  #> number of spatial maps: 10  #> map: abstract spatial landscape with custom features #> stays until the end of the simulation #>  #> population history overview: #>   - time 1: created as an ancestral population (N = 1000) #>   - time 100-200: movement across a landscape  # many slendr functions are pipe-friendly, making it possible to construct # pipelines which construct entire history of a population pop1 <- population(\"pop1\", N = 1000, time = 1,                    map = map, center = c(200, 800), radius = 100) %>%   move(start = 100, end = 200, snapshots = 6,        trajectory = list(c(400, 800), c(600, 700), c(800, 400), c(800, 150))) %>%   set_range(time = 300, polygon = list(     c(400, 0), c(1000, 0), c(1000, 600), c(900, 400), c(800, 250),     c(600, 100), c(500, 50))   )  # population ranges can expand by a given distance in all directions pop2 <- expand_range(pop2, by = 200, start = 50, end = 150, snapshots = 3)  # we can check the positions of all populations interactively by plotting their # ranges together on a single map plot_map(pop1, pop2, pop3)   # gene flow events --------------------------------------------------------  # individual gene flow events can be saved to a list gf <- list(   gene_flow(from = pop1, to = pop3, start = 150, end = 200, rate = 0.15),   gene_flow(from = pop1, to = pop2, start = 300, end = 330, rate = 0.25) )  # compilation -------------------------------------------------------------  # compile model components in a serialized form to dist, returning a single # slendr model object (in practice, the resolution should be smaller) model <- compile_model(   populations = list(pop1, pop2, pop3), generation_time = 1,   resolution = 100, simulation_length = 500,   competition = 5, mating = 5, dispersal = 1 )"},{"path":"https://www.slendr.net/reference/slendr.html","id":null,"dir":"Reference","previous_headings":"","what":"A Simulation Framework for Spatiotemporal Population Genetics — slendr","title":"A Simulation Framework for Spatiotemporal Population Genetics — slendr","text":"framework simulating spatially explicit genomic data leverages real cartographic information programmatic visual encoding spatiotemporal population dynamics real geographic landscapes. Population genetic models automatically executed 'SLiM' software behind scenes, using custom built-simulation 'SLiM' script. Additionally, fully abstract spatial models tied specific geographic location supported, users can also simulate data standard, non-spatial, random-mating models. can simulated either 'SLiM' built-back-end script, using efficient coalescent population genetics simulator 'msprime' custom-built 'Python' script bundled R package. Simulated genomic data saved tree-sequence format can loaded, manipulated, summarised using tree-sequence functionality via R interface 'Python' module 'tskit'. Complete model configuration, simulation analysis pipelines can therefore constructed without need leave R environment, eliminating friction disparate tools population genetic simulations data analysis.","code":""},{"path":"https://www.slendr.net/reference/slendr.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"A Simulation Framework for Spatiotemporal Population Genetics — slendr","text":"can find installation instructions, reference manual, tutorials https://www.slendr.net.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/slendr.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"A Simulation Framework for Spatiotemporal Population Genetics — slendr","text":"Maintainer: Martin Petr contact@bodkan.net (ORCID)","code":""},{"path":"https://www.slendr.net/reference/slim.html","id":null,"dir":"Reference","previous_headings":"","what":"Run a slendr model in SLiM — slim","title":"Run a slendr model in SLiM — slim","text":"function execute SLiM script generated compile function compilation slendr demographic model.","code":""},{"path":"https://www.slendr.net/reference/slim.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Run a slendr model in SLiM — slim","text":"","code":"slim(   model,   sequence_length,   recombination_rate,   samples = NULL,   output = NULL,   burnin = 0,   max_attempts = 1,   spatial = !is.null(model$world),   coalescent_only = TRUE,   method = c(\"batch\", \"gui\"),   random_seed = NULL,   run = TRUE,   verbose = FALSE,   load = TRUE,   locations = NULL,   slim_path = NULL )"},{"path":"https://www.slendr.net/reference/slim.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Run a slendr model in SLiM — slim","text":"model Model object created compile function sequence_length Total length simulated sequence (base-pairs) recombination_rate Recombination rate simulated sequence (recombinations per basepair per generation) samples data frame times given number individuals remembered tree-sequence (see schedule_sampling function can generate sampling schedule correct format). missing, individuals present end simulation recorded tree-sequence output file. output Path output tree-sequence file. NULL (default), tree sequence saved temporary file. burnin Length burnin (model's time units, .e. years) max_attempts many attempts made place offspring near one parents? Serves prevent infinite loops SLiM backend. Default value 1. spatial model executed spatial mode? default, world map specified model definition, simulation proceed spatial mode. coalescent_only initializeTreeSeq(retainCoalescentOnly =   <...>) set TRUE (default) FALSE? See \"retainCoalescentOnly\" SLiM manual detail. method run script? (\"gui\" - open SLiMgui, \"batch\" - run command line) random_seed Random seed (NULL, seed generated 0 maximum integer number available) run SLiM engine run? FALSE, command line SLiM command printed (returned invisibly character vector) executed. verbose Write SLiM output log console (default FALSE)? load final tree sequence immediately loaded returned? Default TRUE. alternative (FALSE) useful tree-sequence file written custom location loaded later point. locations NULL, locations saved. Otherwise, path file locations individual throughout simulation saved (likely use animate_model). slim_path Path appropriate SLiM binary (useful slim binary $PATH). Note argument must specified function run Windows.","code":""},{"path":"https://www.slendr.net/reference/slim.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Run a slendr model in SLiM — slim","text":"tree-sequence object loaded via Python-R reticulate interface function ts_load (internally represented Python object tskit.trees.TreeSequence). Optionally, depending value arguments load = run =, nothing character vector, respectively.","code":""},{"path":"https://www.slendr.net/reference/slim.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Run a slendr model in SLiM — slim","text":"","code":"check_dependencies(python = TRUE, slim = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # afr and eur objects would normally be created before slendr model compilation, # but here we take them out of the model object already compiled for this # example (in a standard slendr simulation pipeline, this wouldn't be necessary) afr <- model$populations[[\"AFR\"]] eur <- model$populations[[\"EUR\"]] chimp <- model$populations[[\"CH\"]]  # schedule the sampling of a couple of ancient and present-day individuals # given model at 20 ky, 10 ky, 5ky ago and at present-day (time 0) modern_samples <- schedule_sampling(model, times = 0, list(afr, 5), list(eur, 5), list(chimp, 1)) ancient_samples <- schedule_sampling(model, times = c(30000, 20000, 10000), list(eur, 1))  # sampling schedules are just data frames and can be merged easily samples <- rbind(modern_samples, ancient_samples)  # run a simulation using the SLiM back end from a compiled slendr model object and return # a tree-sequence output ts <- slim(model, sequence_length = 1e5, recombination_rate = 0, samples = samples)  # automatic loading of a simulated output can be prevented by `load = FALSE`, which can be # useful when a custom path to a tree-sequence output is given for later downstream analyses output_file <- tempfile(fileext = \".trees\") slim(model, sequence_length = 1e5, recombination_rate = 0, samples = samples,      output = output_file, load = FALSE) # ... at a later stage: ts <- ts_load(output_file, model)  ts #> ╔═══════════════════════╗ #> ║TreeSequence           ║ #> ╠═══════════════╤═══════╣ #> ║Trees          │      1║ #> ╟───────────────┼───────╢ #> ║Sequence Length│ 100000║ #> ╟───────────────┼───────╢ #> ║Time Units     │  ticks║ #> ╟───────────────┼───────╢ #> ║Sample Nodes   │  10046║ #> ╟───────────────┼───────╢ #> ║Total Size     │2.6 MiB║ #> ╚═══════════════╧═══════╝ #> ╔═══════════╤═════╤═════════╤════════════╗ #> ║Table      │Rows │Size     │Has Metadata║ #> ╠═══════════╪═════╪═════════╪════════════╣ #> ║Edges      │18361│573.8 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Individuals│12858│  1.2 MiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Migrations │    0│  8 Bytes│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Mutations  │    0│  1.2 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Nodes      │18362│682.1 KiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Populations│    4│  2.5 KiB│         Yes║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Provenances│    1│ 34.7 KiB│          No║ #> ╟───────────┼─────┼─────────┼────────────╢ #> ║Sites      │    0│ 16 Bytes│          No║ #> ╚═══════════╧═════╧═════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/subtract.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate the difference between two slendr objects — subtract","title":"Generate the difference between two slendr objects — subtract","text":"Generate difference two slendr objects","code":""},{"path":"https://www.slendr.net/reference/subtract.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate the difference between two slendr objects — subtract","text":"","code":"subtract(x, y, name = NULL)"},{"path":"https://www.slendr.net/reference/subtract.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate the difference between two slendr objects — subtract","text":"x Object class slendr y Object class slendr name Optional name resulting geographic region. missing, name constructed function arguments.","code":""},{"path":"https://www.slendr.net/reference/subtract.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generate the difference between two slendr objects — subtract","text":"Object class slendr_region encodes standard spatial object class sf several additional attributes (importantly corresponding slendr_map object, applicable).","code":""},{"path":"https://www.slendr.net/reference/summary.slendr_nodes.html","id":null,"dir":"Reference","previous_headings":"","what":"Summarise the contents of a ts_nodes result — summary.slendr_nodes","title":"Summarise the contents of a ts_nodes result — summary.slendr_nodes","text":"Summarise contents ts_nodes result","code":""},{"path":"https://www.slendr.net/reference/summary.slendr_nodes.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Summarise the contents of a ts_nodes result — summary.slendr_nodes","text":"","code":"# S3 method for slendr_nodes summary(object, ...)"},{"path":"https://www.slendr.net/reference/summary.slendr_nodes.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Summarise the contents of a ts_nodes result — summary.slendr_nodes","text":"object Data frame produced function ts_nodes ... Additional formal arguments summary method (unused )","code":""},{"path":"https://www.slendr.net/reference/summary.slendr_nodes.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Summarise the contents of a ts_nodes result — summary.slendr_nodes","text":"Used output terminal","code":""},{"path":"https://www.slendr.net/reference/ts_afs.html","id":null,"dir":"Reference","previous_headings":"","what":"Compute the allele frequency spectrum (AFS) — ts_afs","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"function computes AFS respect given set individuals nodes.","code":""},{"path":"https://www.slendr.net/reference/ts_afs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"","code":"ts_afs(   ts,   sample_sets = NULL,   mode = c(\"site\", \"branch\", \"node\"),   windows = NULL,   span_normalise = FALSE,   polarised = TRUE )"},{"path":"https://www.slendr.net/reference/ts_afs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set). NULL, allele frequency spectrum individuals tree sequence computed. mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) added automatically) span_normalise Argument passed tskit's allele_frequency_spectrum method polarised TRUE (default) allele frequency spectrum folded (.e. counts assume knowledge allele ancestral, derived, known simulation)","code":""},{"path":"https://www.slendr.net/reference/ts_afs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"Allele frequency spectrum values given sample set. Note contents first last elements AFS might surprise . Read links description detail tskit handles things.","code":""},{"path":"https://www.slendr.net/reference/ts_afs.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"information format result dimensions, particular interpretation first last element AFS (complete = TRUE), please see tskit manual https://tskit.dev/tskit/docs/stable/python-api.html example section dedicated AFS https://tskit.dev/tutorials/analysing_tree_sequences.html#allele-frequency-spectra.","code":""},{"path":"https://www.slendr.net/reference/ts_afs.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Compute the allele frequency spectrum (AFS) — ts_afs","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  samples <- ts_samples(ts) %>% .[.$pop %in% c(\"AFR\", \"EUR\"), ]  # compute AFS for the given set of individuals ts_afs(ts, sample_sets = list(samples$name)) #>  [1] 1080   73   21    8    1    4    0    0    0    0   11    0    0    0    0 #> [16]    1    0    3    5   10  935"},{"path":"https://www.slendr.net/reference/ts_ancestors.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","title":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","text":"Extract (spatio-)temporal ancestral history given nodes/individuals","code":""},{"path":"https://www.slendr.net/reference/ts_ancestors.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","text":"","code":"ts_ancestors(ts, x, verbose = FALSE, complete = TRUE)"},{"path":"https://www.slendr.net/reference/ts_ancestors.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","text":"ts Tree sequence object class slendr_ts x Either individual name integer node ID verbose Report progress ancestry path generation? complete every individual tree sequence need complete metadata recorded? TRUE, individuals/nodes complete metadata included reconstruction ancestral relationships. instance, nodes added coalescent recapitation phase included spatial information associated .","code":""},{"path":"https://www.slendr.net/reference/ts_ancestors.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","text":"table ancestral nodes given tree-sequence node way root tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_ancestors.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract (spatio-)temporal ancestral history for given nodes/individuals — ts_ancestors","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # find the complete ancestry information for a given individual ts_ancestors(ts, \"EUR_1\", verbose = TRUE) #> Collecting ancestors of EUR_1 [1/1]... #>  #> Generating data about spatial relationships of nodes... #> # A tibble: 222 × 12 #>    name  pop   node_id level child_id parent_id child_time parent_time child_pop #>    <chr> <fct>   <int> <fct>    <int>     <int>      <dbl>       <dbl> <fct>     #>  1 EUR_1 EUR        16 1           16        33          0         500 EUR       #>  2 EUR_1 EUR        16 2           33        43        500       21830 EUR       #>  3 EUR_1 EUR        16 2           33        47        500       28730 EUR       #>  4 EUR_1 EUR        16 2           33        49        500       30380 EUR       #>  5 EUR_1 EUR        16 2           33        51        500       35360 EUR       #>  6 EUR_1 EUR        16 2           33        54        500       39380 EUR       #>  7 EUR_1 EUR        16 2           33        57        500       41510 EUR       #>  8 EUR_1 EUR        16 2           33        58        500       44810 EUR       #>  9 EUR_1 EUR        16 2           33        59        500       46040 EUR       #> 10 EUR_1 EUR        16 2           33        62        500       51680 EUR       #> # ℹ 212 more rows #> # ℹ 3 more variables: parent_pop <fct>, left_pos <dbl>, right_pos <dbl>"},{"path":"https://www.slendr.net/reference/ts_coalesced.html","id":null,"dir":"Reference","previous_headings":"","what":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","title":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","text":"Check trees tree sequence fully coalesced","code":""},{"path":"https://www.slendr.net/reference/ts_coalesced.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","text":"","code":"ts_coalesced(ts, return_failed = FALSE)"},{"path":"https://www.slendr.net/reference/ts_coalesced.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","text":"ts Tree sequence object class slendr_ts return_failed Report back trees failed coalescence check?","code":""},{"path":"https://www.slendr.net/reference/ts_coalesced.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","text":"TRUE FALSE value return_failed = FALSE, otherwise vector (tskit Python 0-based) indices trees failed coalescence test","code":""},{"path":"https://www.slendr.net/reference/ts_coalesced.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Check that all trees in the tree sequence are fully coalesced — ts_coalesced","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  ts_coalesced(ts) # is the tree sequence fully coalesced? (TRUE or FALSE) #> [1] TRUE  # returns a vector of tree sequence segments which are not coalesced not_coalesced <- ts_coalesced(ts, return_failed = TRUE)"},{"path":"https://www.slendr.net/reference/ts_descendants.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract all descendants of a given tree-sequence node — ts_descendants","title":"Extract all descendants of a given tree-sequence node — ts_descendants","text":"Extract descendants given tree-sequence node","code":""},{"path":"https://www.slendr.net/reference/ts_descendants.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract all descendants of a given tree-sequence node — ts_descendants","text":"","code":"ts_descendants(ts, x, verbose = FALSE, complete = TRUE)"},{"path":"https://www.slendr.net/reference/ts_descendants.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract all descendants of a given tree-sequence node — ts_descendants","text":"ts Tree sequence object class slendr_ts x integer node ID ancestral node verbose Report progress ancestry path generation? complete every individual tree sequence need complete metadata recorded? TRUE, individuals/nodes complete metadata included reconstruction ancestral relationships. instance, nodes added coalescent recapitation phase included spatial information associated .","code":""},{"path":"https://www.slendr.net/reference/ts_descendants.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract all descendants of a given tree-sequence node — ts_descendants","text":"table descendant nodes given tree-sequence node way leaves tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_descendants.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract all descendants of a given tree-sequence node — ts_descendants","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # find the complete descendancy information for a given individual ts_descendants(ts, x = 62, verbose = TRUE) #>  #> Generating data about spatial relationships of nodes... #> # A tibble: 4 × 12 #>   name  pop   node_id level child_id parent_id child_time parent_time child_pop #>   <chr> <fct>   <dbl> <fct>    <int>     <int>      <dbl>       <dbl> <fct>     #> 1 EUR_4 EUR        62 1           22        62          0       51680 EUR       #> 2 NA    EUR        62 1           33        62        500       51680 EUR       #> 3 EUR_1 EUR        62 2           16        33          0         500 EUR       #> 4 EUR_2 EUR        62 2           18        33          0         500 EUR       #> # ℹ 3 more variables: parent_pop <fct>, left_pos <dbl>, right_pos <dbl>"},{"path":"https://www.slendr.net/reference/ts_divergence.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate pairwise divergence between sets of individuals — ts_divergence","title":"Calculate pairwise divergence between sets of individuals — ts_divergence","text":"Calculate pairwise divergence sets individuals","code":""},{"path":"https://www.slendr.net/reference/ts_divergence.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate pairwise divergence between sets of individuals — ts_divergence","text":"","code":"ts_divergence(   ts,   sample_sets,   mode = c(\"site\", \"branch\", \"node\"),   windows = NULL,   span_normalise = TRUE )"},{"path":"https://www.slendr.net/reference/ts_divergence.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate pairwise divergence between sets of individuals — ts_divergence","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set) mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) specified added automatically. span_normalise Divide result span window? Default TRUE, see tskit documentation detail.","code":""},{"path":"https://www.slendr.net/reference/ts_divergence.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate pairwise divergence between sets of individuals — ts_divergence","text":"pairwise calculation, either single divergence value vector divergence values (one window)","code":""},{"path":"https://www.slendr.net/reference/ts_divergence.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate pairwise divergence between sets of individuals — ts_divergence","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # collect sampled individuals from all populations in a list sample_sets <- ts_samples(ts) %>%   split(., .$pop) %>%   lapply(function(pop) pop$name)  # compute the divergence between individuals from each sample set (list of # individual names generated in the previous step) ts_divergence(ts, sample_sets) %>% .[order(.$divergence), ] #> # A tibble: 6 × 3 #>   x     y     divergence #>   <chr> <chr>      <dbl> #> 1 AFR   EUR    0.0000602 #> 2 EUR   NEA    0.000289  #> 3 AFR   NEA    0.000305  #> 4 CH    NEA    0.00397   #> 5 AFR   CH     0.00398   #> 6 CH    EUR    0.00398"},{"path":"https://www.slendr.net/reference/ts_diversity.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate diversity in given sets of individuals — ts_diversity","title":"Calculate diversity in given sets of individuals — ts_diversity","text":"Calculate diversity given sets individuals","code":""},{"path":"https://www.slendr.net/reference/ts_diversity.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate diversity in given sets of individuals — ts_diversity","text":"","code":"ts_diversity(   ts,   sample_sets,   mode = c(\"site\", \"branch\", \"node\"),   windows = NULL,   span_normalise = TRUE )"},{"path":"https://www.slendr.net/reference/ts_diversity.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate diversity in given sets of individuals — ts_diversity","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set). simple vector provided, interpreted .list(sample_sets), meaning given statistic calculated individual separately. mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) added automatically) span_normalise Divide result span window? Default TRUE, see tskit documentation detail.","code":""},{"path":"https://www.slendr.net/reference/ts_diversity.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate diversity in given sets of individuals — ts_diversity","text":"set individuals either single diversity value vector diversity values (one window)","code":""},{"path":"https://www.slendr.net/reference/ts_diversity.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate diversity in given sets of individuals — ts_diversity","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # collect sampled individuals from all populations in a list sample_sets <- ts_samples(ts) %>%   split(., .$pop) %>%   lapply(function(pop) pop$name)  # compute diversity in each population based on sample sets extracted # in the previous step ts_diversity(ts, sample_sets) %>% .[order(.$diversity), ] #> # A tibble: 4 × 2 #>   set    diversity #>   <chr>      <dbl> #> 1 CH    0          #> 2 AFR   0.0000004  #> 3 NEA   0.00000933 #> 4 EUR   0.0000661"},{"path":"https://www.slendr.net/reference/ts_draw.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot a graphical representation of a single tree — ts_draw","title":"Plot a graphical representation of a single tree — ts_draw","text":"function first obtains SVG representation tree calling draw_svg method tskit renders bitmap image R. many optional keyword arguments draw_svg method can provided automatically passed method behind scenes.","code":""},{"path":"https://www.slendr.net/reference/ts_draw.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot a graphical representation of a single tree — ts_draw","text":"","code":"ts_draw(   x,   width = 1000,   height = 1000,   labels = FALSE,   sampled_only = TRUE,   title = NULL,   ... )"},{"path":"https://www.slendr.net/reference/ts_draw.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot a graphical representation of a single tree — ts_draw","text":"x single tree extracted ts_tree width, height Pixel dimensions rendered bitmap labels Label node individual name? sampled_only individuals explicitly sampled simplification labeled? relevant situations sampled individuals can among ancestral nodes. title Optional title figure ... Keyword arguments tskit draw_svg function.","code":""},{"path":"https://www.slendr.net/reference/ts_draw.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot a graphical representation of a single tree — ts_draw","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/ts_draw.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot a graphical representation of a single tree — ts_draw","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract the first tree in the tree sequence and draw it tree <- ts_tree(ts, i = 1)  # ts_draw accepts various optional arguments of tskit.Tree.draw_svg ts_draw(tree, time_scale = \"rank\")"},{"path":"https://www.slendr.net/reference/ts_edges.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","title":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","text":"Extract spatio-temporal edge annotation table given tree tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_edges.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","text":"","code":"ts_edges(x)"},{"path":"https://www.slendr.net/reference/ts_edges.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","text":"x Tree object generated ts_phylo slendr tree sequence object produced ts_load, ts_recapitate, ts_simplify, ts_mutate","code":""},{"path":"https://www.slendr.net/reference/ts_edges.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","text":"Data frame sf type containing times nodes start-end coordinates edges across space","code":""},{"path":"https://www.slendr.net/reference/ts_edges.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract spatio-temporal edge annotation table from a given tree or tree\nsequence — ts_edges","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract an annotated table with (spatio-)temporal edge information ts_edges(ts) #> # A tibble: 298 × 6 #>    child_node_id parent_node_id child_time parent_time child_pop parent_pop #>            <int>          <int>      <dbl>       <dbl> <fct>     <fct>      #>  1             0             74      70000       70040 NEA       NEA        #>  2             1             74      70000       70040 NEA       NEA        #>  3             2             55      40000       40220 NEA       NEA        #>  4             3             55      40000       40220 NEA       NEA        #>  5             4             34          0         740 AFR       AFR        #>  6             5             28          0          80 AFR       AFR        #>  7             6             26          0          20 AFR       AFR        #>  8             7             27          0          80 AFR       AFR        #>  9             8             27          0          80 AFR       AFR        #> 10             9             36          0        1160 AFR       AFR        #> # ℹ 288 more rows"},{"path":"https://www.slendr.net/reference/ts_eigenstrat.html","id":null,"dir":"Reference","previous_headings":"","what":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","title":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","text":"EIGENSTRAT data produced function can used admixr R package (https://bodkan.net/admixr/).","code":""},{"path":"https://www.slendr.net/reference/ts_eigenstrat.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","text":"","code":"ts_eigenstrat(ts, prefix, chrom = \"chr1\", outgroup = NULL)"},{"path":"https://www.slendr.net/reference/ts_eigenstrat.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","text":"ts Tree sequence object class slendr_ts prefix EIGENSTRAT trio prefix chrom name chromosome EIGENSTRAT snp file (default \"chr1\") outgroup formal, artificial outgroup added? NULL (default), outgroup added. non-NULL character name serve name outgroup ind file.","code":""},{"path":"https://www.slendr.net/reference/ts_eigenstrat.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","text":"Object class EIGENSTRAT created admixr package","code":""},{"path":"https://www.slendr.net/reference/ts_eigenstrat.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Convert genotypes to the EIGENSTRAT file format — ts_eigenstrat","text":"case outgroup formally specified slendr model generated tree sequence data, possible artificially create outgroup sample name specified outgroup argument, carry ancestral alleles (.e. value \"2\" geno file position snp file).","code":""},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"functions present R interface corresponding f-statistics methods tskit.","code":""},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"","code":"ts_f2(   ts,   A,   B,   mode = c(\"site\", \"branch\", \"node\"),   span_normalise = TRUE,   windows = NULL )  ts_f3(   ts,   A,   B,   C,   mode = c(\"site\", \"branch\", \"node\"),   span_normalise = TRUE,   windows = NULL )  ts_f4(   ts,   W,   X,   Y,   Z,   mode = c(\"site\", \"branch\", \"node\"),   span_normalise = TRUE,   windows = NULL )  ts_f4ratio(   ts,   X,   A,   B,   C,   O,   mode = c(\"site\", \"branch\"),   span_normalise = TRUE )"},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"ts Tree sequence object class slendr_ts mode mode calculation (\"sites\" \"branch\") span_normalise Divide result span window? Default TRUE, see tskit documentation detail. windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) specified added automatically. W, X, Y, Z, , B, C, O Character vectors individual names (largely following nomenclature Patterson 2021, see crucial differences tskit ADMIXTOOLS Details)","code":""},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"Data frame statistics calculated given sets individuals","code":""},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"Note order populations f3 statistic implemented tskit (https://tskit.dev/tskit/docs/stable/python-api.html#tskit.TreeSequence.f3) different might expect ADMIXTOOLS, defined Patterson 2012 (see https://academic.oup.com/genetics/article/192/3/1065/5935193 heading \"three-population test introduction f-statistics\", well ADMIXTOOLS documentation https://github.com/DReichLab/AdmixTools/blob/master/README.3PopTest#L5). Specifically, widely used notation introduced Patterson assumes population triplet f3(C; , B), C \"focal\" sample (.e., either outgroup sample tested admixture). contrast, tskit implements f3(; B, C), \"focal sample\" . Although likely confuse many ADMIXTOOLS users, slendr much choice , ts_*() functions designed broadly compatible raw tskit methods.","code":""},{"path":"https://www.slendr.net/reference/ts_f4ratio.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the f2, f3, f4, and f4-ratio statistics — ts_f2","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk and add mutations to it ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # calculate f2 for two individuals in a previously loaded tree sequence ts_f2(ts, A = \"AFR_1\", B = \"EUR_1\") #> # A tibble: 1 × 3 #>   A     B           f2 #>   <chr> <chr>    <dbl> #> 1 AFR_1 EUR_1 0.000024  # calculate f2 for two sets of individuals ts_f2(ts, A = c(\"AFR_1\", \"AFR_2\"), B = c(\"EUR_1\", \"EUR_3\")) #> # A tibble: 1 × 3 #>   A           B                  f2 #>   <chr>       <chr>           <dbl> #> 1 AFR_1+AFR_2 EUR_1+EUR_3 0.0000247  # calculate f3 for two individuals in a previously loaded tree sequence ts_f3(ts, A = \"EUR_1\", B = \"AFR_1\", C = \"NEA_1\") #> # A tibble: 1 × 4 #>   A     B     C            f3 #>   <chr> <chr> <chr>     <dbl> #> 1 EUR_1 AFR_1 NEA_1 -0.000012  # calculate f3 for two sets of individuals ts_f3(ts, A = c(\"AFR_1\", \"AFR_2\", \"EUR_1\", \"EUR_2\"),           B = c(\"NEA_1\", \"NEA_2\"),           C = \"CH_1\") #> # A tibble: 1 × 4 #>   A                       B           C           f3 #>   <chr>                   <chr>       <chr>    <dbl> #> 1 AFR_1+AFR_2+EUR_1+EUR_2 NEA_1+NEA_2 CH_1  0.000126  # calculate f4 for single individuals ts_f4(ts, W = \"EUR_1\", X = \"AFR_1\", Y = \"NEA_1\", Z = \"CH_1\") #> # A tibble: 1 × 5 #>   W     X     Y     Z           f4 #>   <chr> <chr> <chr> <chr>    <dbl> #> 1 EUR_1 AFR_1 NEA_1 CH_1  0.000009  # calculate f4 for sets of individuals ts_f4(ts, W = c(\"EUR_1\", \"EUR_2\"),           X = c(\"AFR_1\", \"AFR_2\"),           Y = \"NEA_1\",           Z = \"CH_1\") #> # A tibble: 1 × 5 #>   W           X           Y     Z           f4 #>   <chr>       <chr>       <chr> <chr>    <dbl> #> 1 EUR_1+EUR_2 AFR_1+AFR_2 NEA_1 CH_1  0.000018  # calculate f4-ratio for a given set of target individuals X ts_f4ratio(ts, X = c(\"EUR_1\", \"EUR_2\", \"EUR_4\", \"EUR_5\"),                A = \"NEA_1\", B = \"NEA_2\", C = \"AFR_1\", O = \"CH_1\") #> # A tibble: 4 × 6 #>   X     A     B     C     O      alpha #>   <chr> <chr> <chr> <chr> <chr>  <dbl> #> 1 EUR_1 NEA_1 NEA_2 AFR_1 CH_1  0.0486 #> 2 EUR_2 NEA_1 NEA_2 AFR_1 CH_1  0.181  #> 3 EUR_4 NEA_1 NEA_2 AFR_1 CH_1  0.0417 #> 4 EUR_5 NEA_1 NEA_2 AFR_1 CH_1  0"},{"path":"https://www.slendr.net/reference/ts_fst.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate pairwise statistics between sets of individuals — ts_fst","title":"Calculate pairwise statistics between sets of individuals — ts_fst","text":"discussion difference \"site\", \"branch\", \"node\" options mode argument, please see tskit documentation https://tskit.dev/tskit/docs/stable/stats.html#sec-stats-mode.","code":""},{"path":"https://www.slendr.net/reference/ts_fst.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate pairwise statistics between sets of individuals — ts_fst","text":"","code":"ts_fst(   ts,   sample_sets,   mode = c(\"site\", \"branch\", \"node\"),   windows = NULL,   span_normalise = TRUE )"},{"path":"https://www.slendr.net/reference/ts_fst.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate pairwise statistics between sets of individuals — ts_fst","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set) mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) specified added automatically. span_normalise Divide result span window? Default TRUE, see tskit documentation detail.","code":""},{"path":"https://www.slendr.net/reference/ts_fst.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate pairwise statistics between sets of individuals — ts_fst","text":"pairwise calculation, either single Fst value vector Fst values (one window)","code":""},{"path":"https://www.slendr.net/reference/ts_fst.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate pairwise statistics between sets of individuals — ts_fst","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # compute F_st between two sets of individuals in a given tree sequence ts ts_fst(ts, sample_sets = list(afr = c(\"AFR_1\", \"AFR_2\", \"AFR_3\"),                               eur = c(\"EUR_1\", \"EUR_2\"))) #> # A tibble: 1 × 3 #>   x     y       Fst #>   <chr> <chr> <dbl> #> 1 afr   eur   0.303"},{"path":"https://www.slendr.net/reference/ts_genotypes.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract genotype table from the tree sequence — ts_genotypes","title":"Extract genotype table from the tree sequence — ts_genotypes","text":"Extract genotype table tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_genotypes.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract genotype table from the tree sequence — ts_genotypes","text":"","code":"ts_genotypes(ts)"},{"path":"https://www.slendr.net/reference/ts_genotypes.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract genotype table from the tree sequence — ts_genotypes","text":"ts Tree sequence object class slendr_ts","code":""},{"path":"https://www.slendr.net/reference/ts_genotypes.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract genotype table from the tree sequence — ts_genotypes","text":"Data frame object class tibble containing genotypes simulated individuals columns","code":""},{"path":"https://www.slendr.net/reference/ts_genotypes.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract genotype table from the tree sequence — ts_genotypes","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk, recapitate it, simplify it, and mutate it ts <- ts_load(slendr_ts, model) %>%   ts_recapitate(Ne = 10000, recombination_rate = 1e-8) %>%   ts_simplify() %>%   ts_mutate(mutation_rate = 1e-8)  # extract the genotype matrix (this could take  a long time consume lots # of memory!) gts <- ts_genotypes(ts) #> 3 multiallelic sites (0.138% out of 2173 total) detected and removed"},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":null,"dir":"Reference","previous_headings":"","what":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"function iterates tree sequence returns IBD tracts pairs individuals nodes","code":""},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"","code":"ts_ibd(   ts,   coordinates = FALSE,   within = NULL,   between = NULL,   squash = FALSE,   minimum_length = NULL,   maximum_time = NULL,   sf = TRUE )"},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"ts Tree sequence object class slendr_ts coordinates coordinates detected IBD tracts reported? FALSE (default), total length shared IBD segments numbers reported. TRUE, coordinates segment returned (note can massive impact memory usage). See details information. within character vector individual names integer vector node IDs indicating set nodes within look IBD segments. list lists character vectors individual names integer vectors node IDs, indicating set nodes look shared IBD segments. squash adjacent IBD segments pairs nodes squashed differ 'genealogical paths' MRCA? Default FALSE. context, see https://github.com/tskit-dev/tskit/issues/2459. option EXPERIMENTAL! minimum_length Minimum length IBD segment return results. useful reducing total amount IBD returned (see Details). maximum_time Oldest MRCA node considered IBD ancestor return IBD segment results. useful reducing total amount IBD returned. sf IBD segments spatial tree sequence analyzed, returned table spatial sf object? Default TRUE.","code":""},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"data frame IBD results (either coordinates IBD segment shared pair nodes, summary statistics total IBD sharing pair)","code":""},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"function considered experimental. full control IBD segment detection tree-sequence data, users can (perhaps, time , ) rely tskit method ibd_segments (see https://tskit.dev/tskit/docs/stable/python-api.html#tskit.TreeSequence.ibd_segments). Iternally, function leverages tskit TreeSequence method ibd_segments. However, note ts_ibd function always returns data frame IBD tracts, provide option iterate individual IBD segments shown official tskit documentation https://tskit.dev/tskit/docs/stable/ibd.html. general, R handles heavy iteration poorly, function attempt serve full wrapper ibd_segments. Unfortunately, distinction \"squashed IBD\" (many consider expected definition IBD) tskit’s IBD defined via distinct genealogical paths (see https://github.com/tskit-dev/tskit/issues/2459 discussion topic), makes meaning filtering parameter ibd_segments() method tskit minimum_length somewhat unintuitive. moment, function argument filters IBD segments tskit level, level squashed IBD segments!","code":""},{"path":"https://www.slendr.net/reference/ts_ibd.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Collect Identity-by-Descent (IBD) segments (EXPERIMENTAL) — ts_ibd","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # find IBD segments between specified Neanderthals and Europeans ts_ibd(   ts,   coordinates = TRUE,   between = list(c(\"NEA_1\", \"NEA_2\"), c(\"EUR_1\", \"EUR_2\")),   minimum_length = 40000 ) #> # A tibble: 34 × 13 #>    node1 node2 length  mrca node1_time node2_time tmrca   left  right name1 #>    <int> <int>  <dbl> <dbl>      <dbl>      <dbl> <dbl>  <dbl>  <dbl> <chr> #>  1     0    16  52762    83      70000          0 20001 228046 280808 NEA_1 #>  2     0    16  45006    83      70000          0 20001 449167 494173 NEA_1 #>  3     0    17  52762    83      70000          0 20001 228046 280808 NEA_1 #>  4     0    18  52762    83      70000          0 20001 228046 280808 NEA_1 #>  5     0    18  45006    83      70000          0 20001 449167 494173 NEA_1 #>  6     0    19  57765    83      70000          0 20001  47447 105212 NEA_1 #>  7     0    19  47935    83      70000          0 20001 105212 153147 NEA_1 #>  8     0    19  46028    83      70000          0 20001 178539 224567 NEA_1 #>  9     0    19  58054    79      70000          0  2340 227185 285239 NEA_1 #> 10     1    16  52762    83      70000          0 20001 228046 280808 NEA_1 #> # ℹ 24 more rows #> # ℹ 3 more variables: name2 <chr>, pop1 <fct>, pop2 <fct>"},{"path":"https://www.slendr.net/reference/ts_load.html","id":null,"dir":"Reference","previous_headings":"","what":"Load a tree sequence file produced by a given model — ts_load","title":"Load a tree sequence file produced by a given model — ts_load","text":"function loads tree sequence file simulated given slendr model. Optionally, tree sequence can recapitated simplified.","code":""},{"path":"https://www.slendr.net/reference/ts_load.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Load a tree sequence file produced by a given model — ts_load","text":"","code":"ts_load(file, model = NULL)"},{"path":"https://www.slendr.net/reference/ts_load.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Load a tree sequence file produced by a given model — ts_load","text":"file path tree-sequence file (either originating slendr model standard non-slendr tree sequence). model Optional slendr_model object produced tree-sequence file. Used adding various annotation data metadata standard tskit tree-sequence object.","code":""},{"path":"https://www.slendr.net/reference/ts_load.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Load a tree sequence file produced by a given model — ts_load","text":"Tree-sequence object class slendr_ts, serves interface point Python module tskit using slendr functions ts_ prefix.","code":""},{"path":"https://www.slendr.net/reference/ts_load.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Load a tree sequence file produced by a given model — ts_load","text":"loading, recapitation simplification performed using Python module pyslim serves link tree sequences generated SLiM tskit module manipulation tree sequence data. steps modelled official pyslim tutorial documentation available : https://tskit.dev/pyslim/docs/latest/tutorial.html. recapitation simplification steps can also performed individually using functions ts_recapitate ts_simplify.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_load.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Load a tree sequence file produced by a given model — ts_load","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load tree sequence generated by a given model ts <- ts_load(slendr_ts, model)  # even tree sequences generated by non-slendr models can be msprime_ts <- system.file(\"extdata/models/msprime.trees\", package = \"slendr\") ts <- ts_load(msprime_ts)  # load tree sequence and immediately simplify it only to sampled individuals # (note that the example tree sequence is already simplified so this operation # does not do anything in this case) ts <- ts_load(slendr_ts, model = model) %>% ts_simplify(keep_input_roots = TRUE)  # load tree sequence and simplify it to a subset of sampled individuals ts_small <- ts_simplify(ts, simplify_to = c(\"CH_1\", \"NEA_1\", \"NEA_2\",                                             \"AFR_1\", \"AFR_2\", \"EUR_1\", \"EUR_2\"))  # load tree sequence, recapitate it and simplify it ts <- ts_load(slendr_ts, model) %>%   ts_recapitate(recombination_rate = 1e-8, Ne = 10000, random_seed = 42) %>%   ts_simplify()  # load tree sequence, recapitate it, simplify it and overlay neutral mutations ts <- ts_load(slendr_ts, model) %>%   ts_recapitate(recombination_rate = 1e-8, Ne = 10000, random_seed = 42) %>%   ts_simplify() %>%   ts_mutate(mutation_rate = 1e-8)  ts #> ╔═════════════════════════╗ #> ║TreeSequence             ║ #> ╠═══════════════╤═════════╣ #> ║Trees          │       70║ #> ╟───────────────┼─────────╢ #> ║Sequence Length│   500000║ #> ╟───────────────┼─────────╢ #> ║Time Units     │    ticks║ #> ╟───────────────┼─────────╢ #> ║Sample Nodes   │       26║ #> ╟───────────────┼─────────╢ #> ║Total Size     │205.3 KiB║ #> ╚═══════════════╧═════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │ 297│ 9.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│  72│ 8.8 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │2206│80.9 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │  85│ 3.8 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   6│ 2.7 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   6│41.0 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │2201│53.8 KiB│          No║ #> ╚═══════════╧════╧════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/ts_metadata.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","title":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","text":"Extract list tree sequence metadata saved SLiM","code":""},{"path":"https://www.slendr.net/reference/ts_metadata.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","text":"","code":"ts_metadata(ts)"},{"path":"https://www.slendr.net/reference/ts_metadata.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","text":"ts Tree sequence object class slendr_ts","code":""},{"path":"https://www.slendr.net/reference/ts_metadata.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","text":"List metadata fields extracted tree-sequence object","code":""},{"path":"https://www.slendr.net/reference/ts_metadata.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract list with tree sequence metadata saved by SLiM — ts_metadata","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract the list of metadata information from the tree sequence ts_metadata(ts) #> $version #> [1] \"slendr_0.8.1.9000\" #>  #> $description #> [1] \"\" #>  #> $sampling #> # A tibble: 13 × 3 #>    name   time pop   #>    <chr> <int> <chr> #>  1 NEA_1 70000 NEA   #>  2 NEA_2 40000 NEA   #>  3 AFR_1     0 AFR   #>  4 AFR_2     0 AFR   #>  5 AFR_3     0 AFR   #>  6 AFR_4     0 AFR   #>  7 AFR_5     0 AFR   #>  8 CH_1      0 CH    #>  9 EUR_1     0 EUR   #> 10 EUR_2     0 EUR   #> 11 EUR_3     0 EUR   #> 12 EUR_4     0 EUR   #> 13 EUR_5     0 EUR   #>  #> $sample_names #>  [1] \"NEA_1\" \"NEA_2\" \"AFR_1\" \"AFR_2\" \"AFR_3\" \"AFR_4\" \"AFR_5\" \"CH_1\"  \"EUR_1\" #> [10] \"EUR_2\" \"EUR_3\" \"EUR_4\" \"EUR_5\" #>  #> $sample_ids #> NULL #>  #> $map #> NULL #>  #> $arguments #> $arguments$BURNIN_LENGTH #> [1] 0 #>  #> $arguments$MAX_ATTEMPTS #> [1] 1 #>  #> $arguments$RECOMB_RATE #> [1] 1e-08 #>  #> $arguments$SEED #> [1] 314159 #>  #> $arguments$SEQUENCE_LENGTH #> [1] 500000 #>  #> $arguments$SIMULATION_LENGTH #> [1] 216667 #>  #>"},{"path":"https://www.slendr.net/reference/ts_mutate.html","id":null,"dir":"Reference","previous_headings":"","what":"Add mutations to the given tree sequence — ts_mutate","title":"Add mutations to the given tree sequence — ts_mutate","text":"Add mutations given tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_mutate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add mutations to the given tree sequence — ts_mutate","text":"","code":"ts_mutate(   ts,   mutation_rate,   random_seed = NULL,   keep_existing = TRUE,   mut_type = NULL )"},{"path":"https://www.slendr.net/reference/ts_mutate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add mutations to the given tree sequence — ts_mutate","text":"ts Tree sequence object class slendr_ts mutation_rate Mutation rate used msprime simulate mutations random_seed Random seed passed msprime's mutate method (NULL, seed generated 0 maximum integer number available) keep_existing Keep existing mutations? mut_type Assign SLiM mutation type neutral mutations? NULL (default), special mutation type used. integer number given, mutations SLiM mutation type integer identifier created.","code":""},{"path":"https://www.slendr.net/reference/ts_mutate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add mutations to the given tree sequence — ts_mutate","text":"Tree-sequence object class slendr_ts, serves interface point Python module tskit using slendr functions ts_ prefix.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_mutate.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add mutations to the given tree sequence — ts_mutate","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  ts <- ts_load(slendr_ts, model) ts_mutate <- ts_mutate(ts, mutation_rate = 1e-8, random_seed = 42)  ts_mutate #> ╔═════════════════════════╗ #> ║TreeSequence             ║ #> ╠═══════════════╤═════════╣ #> ║Trees          │       70║ #> ╟───────────────┼─────────╢ #> ║Sequence Length│   500000║ #> ╟───────────────┼─────────╢ #> ║Time Units     │    ticks║ #> ╟───────────────┼─────────╢ #> ║Sample Nodes   │       26║ #> ╟───────────────┼─────────╢ #> ║Total Size     │204.8 KiB║ #> ╚═══════════════╧═════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │ 298│ 9.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│  73│ 8.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │2246│82.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │  86│ 3.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   5│ 2.6 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   4│38.0 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │2240│54.7 KiB│          No║ #> ╚═══════════╧════╧════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/ts_names.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract names of individuals in a tree sequence — ts_names","title":"Extract names of individuals in a tree sequence — ts_names","text":"Extract names individuals tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_names.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract names of individuals in a tree sequence — ts_names","text":"","code":"ts_names(ts, split = NULL)"},{"path":"https://www.slendr.net/reference/ts_names.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract names of individuals in a tree sequence — ts_names","text":"ts Tree sequence object class slendr_ts split sample names tree sequence split column (population time column)? Default NULL names samples returned single character vector. set \"pop\" \"time\", list character vectors returned, one vector unique \"pop\" \"time\" grouping.","code":""},{"path":"https://www.slendr.net/reference/ts_names.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract names of individuals in a tree sequence — ts_names","text":"vector character sample names. split specified, list vectors returned, one element list per population sampling time.","code":""},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract combined annotated table of individuals and nodes — ts_nodes","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"function combines information table individuals table nodes single data frame can used downstream analyses.","code":""},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"","code":"ts_nodes(x, sf = TRUE)"},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"x Tree sequence object class slendr_ts phylo object extracted ts_phylo sf spatial data returned sf format? FALSE, spatial geometries returned simply x y columns, instead standard POINT data type.","code":""},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"Data frame processed information tree sequence object. model generated data spatial, result returned spatial object class sf.","code":""},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"source data (tables individuals nodes recorded tree sequence generated SLiM) combined single data frame. model generated data spatial, coordinates nodes (pixel-based default SLiM spatial simulations occur raster), coordinates automatically converted explicit spatial object sf class unless spatial = FALSE. See https://r-spatial.github.io/sf/ extensive introduction sf package ways spatial data can processed, analysed, visualised.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_nodes.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract combined annotated table of individuals and nodes — ts_nodes","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract an annotated table with (spatio-)temporal node information ts_nodes(ts) #> # A tibble: 86 × 12 #>    name  pop   node_id  time time_tskit sampled remembered retained alive #>    <chr> <fct>   <int> <dbl>      <dbl> <lgl>   <lgl>      <lgl>    <lgl> #>  1 NEA_1 NEA         0 70000       2334 TRUE    TRUE       TRUE     FALSE #>  2 NEA_1 NEA         1 70000       2334 TRUE    TRUE       TRUE     FALSE #>  3 NEA_2 NEA         2 40000       1334 TRUE    TRUE       TRUE     FALSE #>  4 NEA_2 NEA         3 40000       1334 TRUE    TRUE       TRUE     FALSE #>  5 AFR_1 AFR         4     0          0 TRUE    TRUE       TRUE     TRUE  #>  6 AFR_1 AFR         5     0          0 TRUE    TRUE       TRUE     TRUE  #>  7 AFR_2 AFR         6     0          0 TRUE    TRUE       TRUE     TRUE  #>  8 AFR_2 AFR         7     0          0 TRUE    TRUE       TRUE     TRUE  #>  9 AFR_3 AFR         8     0          0 TRUE    TRUE       TRUE     TRUE  #> 10 AFR_3 AFR         9     0          0 TRUE    TRUE       TRUE     TRUE  #> # ℹ 76 more rows #> # ℹ 3 more variables: pedigree_id <dbl>, pop_id <int>, ind_id <dbl>"},{"path":"https://www.slendr.net/reference/ts_phylo.html","id":null,"dir":"Reference","previous_headings":"","what":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","title":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","text":"Convert tree tree sequence object class phylo","code":""},{"path":"https://www.slendr.net/reference/ts_phylo.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","text":"","code":"ts_phylo(   ts,   i,   mode = c(\"index\", \"position\"),   labels = c(\"tskit\", \"pop\"),   quiet = FALSE )"},{"path":"https://www.slendr.net/reference/ts_phylo.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","text":"ts Tree sequence object class slendr_ts Position tree tree sequence. mode = \"index\", -th tree returned (zero-based indexing tskit), mode = \"position\", tree covering -th base simulated genome returned (, tskit's indexing). mode argument interpreted? Either \"index\" -th tree sequence genealogies, \"position\" along simulated genome. labels stored node labels final phylo object? Options either population name tskit integer node ID (different thing phylo class node integer index). quiet ape's internal phylo validity test printed ?","code":""},{"path":"https://www.slendr.net/reference/ts_phylo.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","text":"Standard phylogenetic tree object implemented R package ape","code":""},{"path":"https://www.slendr.net/reference/ts_phylo.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Convert a tree in the tree sequence to an object of the class phylo — ts_phylo","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>%   ts_recapitate(Ne = 10000, recombination_rate = 1e-8) %>%   ts_simplify()  # extract the 1st tree from a given tree sequence, return ape object tree <- ts_phylo(ts, i = 1, mode = \"index\", quiet = TRUE) tree #>  #> Phylogenetic tree with 26 tips and 24 internal nodes. #>  #> Tip labels: #>   25 (EUR_5), 24 (EUR_5), 23 (EUR_4), 22 (EUR_4), 21 (EUR_3), 20 (EUR_3), ... #> Node labels: #>   84, 28, 27, 30, 31, 32, ... #>  #> Rooted; includes branch lengths.  # extract the tree at a 42th basepair in the given tree sequence tree <- ts_phylo(ts, i = 42, mode = \"position\", quiet = TRUE)  # because the tree is a standard ape phylo object, we can plot it easily plot(tree, use.edge.length = FALSE) ape::nodelabels()"},{"path":"https://www.slendr.net/reference/ts_recapitate.html","id":null,"dir":"Reference","previous_headings":"","what":"Recapitate the tree sequence — ts_recapitate","title":"Recapitate the tree sequence — ts_recapitate","text":"Recapitate tree sequence","code":""},{"path":"https://www.slendr.net/reference/ts_recapitate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Recapitate the tree sequence — ts_recapitate","text":"","code":"ts_recapitate(   ts,   recombination_rate,   Ne = NULL,   demography = NULL,   random_seed = NULL )"},{"path":"https://www.slendr.net/reference/ts_recapitate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Recapitate the tree sequence — ts_recapitate","text":"ts Tree sequence object loaded ts_load recombination_rate constant value recombination rate Ne Effective population size recapitation process demography Ancestral demography passed internally msprime.sim_ancestry() (see msprime's documentation mode detail) random_seed Random seed passed pyslim's recapitate method (NULL, seed generated 0 maximum integer number available)","code":""},{"path":"https://www.slendr.net/reference/ts_recapitate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Recapitate the tree sequence — ts_recapitate","text":"Tree-sequence object class slendr_ts, serves interface point Python module tskit using slendr functions ts_ prefix.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_recapitate.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Recapitate the tree sequence — ts_recapitate","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  ts <- ts_load(slendr_ts, model) %>%   ts_recapitate(recombination_rate = 1e-8, Ne = 10000, random_seed = 42)  ts #> ╔════════════════════════╗ #> ║TreeSequence            ║ #> ╠═══════════════╤════════╣ #> ║Trees          │      70║ #> ╟───────────────┼────────╢ #> ║Sequence Length│  500000║ #> ╟───────────────┼────────╢ #> ║Time Units     │   ticks║ #> ╟───────────────┼────────╢ #> ║Sample Nodes   │      26║ #> ╟───────────────┼────────╢ #> ║Total Size     │70.8 KiB║ #> ╚═══════════════╧════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │ 298│ 9.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│  73│ 8.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │   0│ 1.2 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │  86│ 3.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   6│ 2.7 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   4│39.8 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │   0│16 Bytes│          No║ #> ╚═══════════╧════╧════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/ts_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","title":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","text":"Extract names times individuals interest current tree sequence (either sampled individuals user simplified )","code":""},{"path":"https://www.slendr.net/reference/ts_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","text":"","code":"ts_samples(ts)"},{"path":"https://www.slendr.net/reference/ts_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","text":"ts Tree sequence object class slendr_ts","code":""},{"path":"https://www.slendr.net/reference/ts_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","text":"Table individuals scheduled sampling across space time","code":""},{"path":"https://www.slendr.net/reference/ts_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract names and times of individuals of interest in the current tree sequence\n(either all sampled individuals or those that the user simplified to) — ts_samples","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract the table of individuals scheduled for simulation and sampling ts_samples(ts) #> # A tibble: 13 × 3 #>    name   time pop   #>    <chr> <int> <chr> #>  1 NEA_1 70000 NEA   #>  2 NEA_2 40000 NEA   #>  3 AFR_1     0 AFR   #>  4 AFR_2     0 AFR   #>  5 AFR_3     0 AFR   #>  6 AFR_4     0 AFR   #>  7 AFR_5     0 AFR   #>  8 CH_1      0 CH    #>  9 EUR_1     0 EUR   #> 10 EUR_2     0 EUR   #> 11 EUR_3     0 EUR   #> 12 EUR_4     0 EUR   #> 13 EUR_5     0 EUR"},{"path":"https://www.slendr.net/reference/ts_save.html","id":null,"dir":"Reference","previous_headings":"","what":"Save a tree sequence to a file — ts_save","title":"Save a tree sequence to a file — ts_save","text":"Save tree sequence file","code":""},{"path":"https://www.slendr.net/reference/ts_save.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Save a tree sequence to a file — ts_save","text":"","code":"ts_save(ts, file)"},{"path":"https://www.slendr.net/reference/ts_save.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Save a tree sequence to a file — ts_save","text":"ts Tree sequence object loaded ts_load file File tree sequence saved","code":""},{"path":"https://www.slendr.net/reference/ts_save.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Save a tree sequence to a file — ts_save","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/ts_save.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Save a tree sequence to a file — ts_save","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree sequence ts <- ts_load(slendr_ts, model)  # save the tree-sequence object to a different location another_file <- paste(tempfile(), \".trees\") ts_save(ts, another_file)"},{"path":"https://www.slendr.net/reference/ts_segregating.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","title":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","text":"Calculate density segregating sites given sets individuals","code":""},{"path":"https://www.slendr.net/reference/ts_segregating.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","text":"","code":"ts_segregating(   ts,   sample_sets,   mode = c(\"site\", \"branch\", \"node\"),   windows = NULL,   span_normalise = FALSE )"},{"path":"https://www.slendr.net/reference/ts_segregating.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set). simple vector provided, interpreted .list(sample_sets), meaning given statistic calculated individual separately. mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) added automatically) span_normalise Divide result span window? Default TRUE, see tskit documentation detail.","code":""},{"path":"https://www.slendr.net/reference/ts_segregating.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","text":"set individuals either single diversity value vector diversity values (one window)","code":""},{"path":"https://www.slendr.net/reference/ts_segregating.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate the density of segregating sites for the given sets of individuals — ts_segregating","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # collect sampled individuals from all populations in a list sample_sets <- ts_samples(ts) %>%   split(., .$pop) %>%   lapply(function(pop) pop$name)  ts_segregating(ts, sample_sets) #> # A tibble: 4 × 2 #>   set   segsites #>   <chr>    <dbl> #> 1 AFR          1 #> 2 CH           0 #> 3 EUR        122 #> 4 NEA          7"},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":null,"dir":"Reference","previous_headings":"","what":"Simplify the tree sequence down to a given set of individuals — ts_simplify","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"function convenience wrapper around simplify method implemented tskit, designed work tree sequence data simulated SLiM using slendr R package.","code":""},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"","code":"ts_simplify(   ts,   simplify_to = NULL,   keep_input_roots = FALSE,   keep_unary = FALSE,   keep_unary_in_individuals = FALSE,   filter_nodes = TRUE )"},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"ts Tree sequence object class slendr_ts simplify_to character vector individual names. NULL, explicitly remembered individuals (.e. specified via schedule_sampling function left tree sequence simplification. keep_input_roots history ancestral MRCA samples retained tree sequence? Default FALSE. keep_unary unary nodes preserved simplification? Default FALSE. keep_unary_in_individuals unary nodes preserved simplification associated individual recorded table individuals? Default FALSE. set TRUE keep_unary also TRUE filter_nodes nodes reindexed simplification? Default TRUE. See tskit's documentation Python method simplify()","code":""},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"Tree-sequence object class slendr_ts, serves interface point Python module tskit using slendr functions ts_ prefix.","code":""},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"simplification process used remove redundant information tree sequence retains information necessary describe genealogical history set samples. information simplification works pyslim tskit, see official documentation https://tskit.dev/tskit/docs/stable/python-api.html#tskit.TreeSequence.simplify https://tskit.dev/pyslim/docs/latest/tutorial.html#simplification. clear description difference remembering retaining use techniques implement historical individuals (.e. ancient DNA samples) pyslim documentation https://tskit.dev/pyslim/docs/latest/tutorial.html#historical-individuals.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_simplify.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simplify the tree sequence down to a given set of individuals — ts_simplify","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  ts <- ts_load(slendr_ts, model) ts #> ╔════════════════════════╗ #> ║TreeSequence            ║ #> ╠═══════════════╤════════╣ #> ║Trees          │      70║ #> ╟───────────────┼────────╢ #> ║Sequence Length│  500000║ #> ╟───────────────┼────────╢ #> ║Time Units     │   ticks║ #> ╟───────────────┼────────╢ #> ║Sample Nodes   │      26║ #> ╟───────────────┼────────╢ #> ║Total Size     │68.2 KiB║ #> ╚═══════════════╧════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │ 298│ 9.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│  73│ 8.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │   0│ 1.2 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │  86│ 3.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   5│ 2.6 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   3│37.3 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │   0│16 Bytes│          No║ #> ╚═══════════╧════╧════════╧════════════╝ #>   # simplify tree sequence to sampled individuals ts_simplified <- ts_simplify(ts)  # simplify to a subset of sampled individuals ts_small <- ts_simplify(ts, simplify_to = c(\"CH_1\", \"NEA_1\", \"NEA_2\", \"AFR_1\",                                             \"AFR_2\", \"EUR_1\", \"EUR_2\"))  ts_small #> ╔════════════════════════╗ #> ║TreeSequence            ║ #> ╠═══════════════╤════════╣ #> ║Trees          │      28║ #> ╟───────────────┼────────╢ #> ║Sequence Length│  500000║ #> ╟───────────────┼────────╢ #> ║Time Units     │   ticks║ #> ╟───────────────┼────────╢ #> ║Sample Nodes   │      14║ #> ╟───────────────┼────────╢ #> ║Total Size     │56.0 KiB║ #> ╚═══════════════╧════════╝ #> ╔═══════════╤════╤════════╤════════════╗ #> ║Table      │Rows│Size    │Has Metadata║ #> ╠═══════════╪════╪════════╪════════════╣ #> ║Edges      │ 118│ 3.7 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Individuals│  32│ 4.9 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Migrations │   0│ 8 Bytes│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Mutations  │   0│ 1.2 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Nodes      │  39│ 2.1 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Populations│   5│ 2.6 KiB│         Yes║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Provenances│   4│37.8 KiB│          No║ #> ╟───────────┼────┼────────┼────────────╢ #> ║Sites      │   0│16 Bytes│          No║ #> ╚═══════════╧════╧════════╧════════════╝ #>"},{"path":"https://www.slendr.net/reference/ts_table.html","id":null,"dir":"Reference","previous_headings":"","what":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"function extracts data given tree sequence table. times converted model-specific time units tskit's \"generations backwards\" time direction.","code":""},{"path":"https://www.slendr.net/reference/ts_table.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"","code":"ts_table(ts, table = c(\"individuals\", \"edges\", \"nodes\", \"mutations\"))"},{"path":"https://www.slendr.net/reference/ts_table.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"ts Tree sequence object class slendr_ts table tree sequence table return","code":""},{"path":"https://www.slendr.net/reference/ts_table.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"Data frame information give tree-sequence table (can either table individuals, edges, nodes, mutations).","code":""},{"path":"https://www.slendr.net/reference/ts_table.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"processing analyses, output function ts_nodes might useful, merges information node individual tables one table annotates useful information model configuration data.","code":""},{"path":[]},{"path":"https://www.slendr.net/reference/ts_table.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Get the table of individuals/nodes/edges/mutations from the tree sequence — ts_table","text":"","code":"# load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk and add mutations to it ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # get the 'raw' tskit table of individuals ts_table(ts, \"individuals\") #> # A tibble: 73 × 9 #>    ind_id  time pedigree_id  pop_id alive remembered retained sampled time_tskit #>     <dbl> <dbl>       <dbl> <int[1> <lgl> <lgl>      <lgl>    <lgl>    <dbl[1d]> #>  1      0   -10    16023346       0 TRUE  TRUE       TRUE     TRUE             0 #>  2      1   -10    16023350       1 TRUE  TRUE       TRUE     TRUE             0 #>  3      2   -10    16023352       1 TRUE  TRUE       TRUE     TRUE             0 #>  4      3   -10    16023353       1 TRUE  TRUE       TRUE     TRUE             0 #>  5      4   -10    16023355       1 TRUE  TRUE       TRUE     TRUE             0 #>  6      5   -10    16023359       1 TRUE  TRUE       TRUE     TRUE             0 #>  7      6   -10    16024625       3 TRUE  TRUE       TRUE     TRUE             0 #>  8      7   -10    16025081       3 TRUE  TRUE       TRUE     TRUE             0 #>  9      8   -10    16025520       3 TRUE  TRUE       TRUE     TRUE             0 #> 10      9   -10    16026684       3 TRUE  TRUE       TRUE     TRUE             0 #> # ℹ 63 more rows  # get the 'raw' tskit table of edges ts_table(ts, \"edges\") #> # A tibble: 298 × 5 #>       id child parent   left  right #>    <dbl> <int>  <int>  <dbl>  <dbl> #>  1     0     6     26      0 500000 #>  2     1    13     26      0 500000 #>  3     2     7     27      0 500000 #>  4     3     8     27      0 500000 #>  5     4    12     27      0 500000 #>  6     5     5     28      0 500000 #>  7     6    10     28      0 500000 #>  8     7    11     29 306056 500000 #>  9     8    26     29 306056 500000 #> 10     9    27     30      0 500000 #> # ℹ 288 more rows  # get the 'raw' tskit table of nodes ts_table(ts, \"nodes\") #> # A tibble: 86 × 5 #>    node_id    ind_id pop_id  time time_tskit #>      <int> <int[1d]>  <int> <dbl>      <dbl> #>  1       0        25      2 70010       2334 #>  2       1        25      2 70010       2334 #>  3       2        43      2 40010       1334 #>  4       3        43      2 40010       1334 #>  5       4         1      1   -10          0 #>  6       5         1      1   -10          0 #>  7       6         2      1   -10          0 #>  8       7         2      1   -10          0 #>  9       8         3      1   -10          0 #> 10       9         3      1   -10          0 #> # ℹ 76 more rows  # get the 'raw' tskit table of mutations ts_table(ts, \"mutations\") #> # A tibble: 2,246 × 5 #>       id  site  node     time time_tskit #>    <dbl> <int> <int>    <dbl>      <dbl> #>  1     0     0    83 3626483.    120883. #>  2     1     1    83 1766659.     58889. #>  3     2     2    83 2476958.     82566. #>  4     3     3    83 4356462.    145216. #>  5     4     4    83  876159.     29206. #>  6     5     5    83 3848814.    128294. #>  7     6     6    83 2554081.     85136. #>  8     7     7    83  976093.     32537. #>  9     8     8    32 3622571.    120753. #> 10     9     9    32  104626.      3488. #> # ℹ 2,236 more rows"},{"path":"https://www.slendr.net/reference/ts_tajima.html","id":null,"dir":"Reference","previous_headings":"","what":"Calculate Tajima's D for given sets of individuals — ts_tajima","title":"Calculate Tajima's D for given sets of individuals — ts_tajima","text":"discussion difference \"site\" \"branch\" options mode argument, please see tskit documentation https://tskit.dev/tskit/docs/stable/stats.html#sec-stats-mode","code":""},{"path":"https://www.slendr.net/reference/ts_tajima.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Calculate Tajima's D for given sets of individuals — ts_tajima","text":"","code":"ts_tajima(ts, sample_sets, mode = c(\"site\", \"branch\", \"node\"), windows = NULL)"},{"path":"https://www.slendr.net/reference/ts_tajima.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Calculate Tajima's D for given sets of individuals — ts_tajima","text":"ts Tree sequence object class slendr_ts sample_sets list (optionally named list) character vectors individual names (one vector per set). simple vector provided, interpreted .list(sample_sets), meaning given statistic calculated individual separately. mode mode calculation (\"sites\" \"branch\") windows Coordinates breakpoints windows. first coordinate (0) last coordinate (equal ts$sequence_length) added automatically)","code":""},{"path":"https://www.slendr.net/reference/ts_tajima.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Calculate Tajima's D for given sets of individuals — ts_tajima","text":"set individuals either single Tajima's D value vector Tajima's D values (one window)","code":""},{"path":"https://www.slendr.net/reference/ts_tajima.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Calculate Tajima's D for given sets of individuals — ts_tajima","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model) %>% ts_mutate(mutation_rate = 1e-8, random_seed = 42)  # calculate Tajima's D for given sets of individuals in a tree sequence ts ts_tajima(ts, list(eur = c(\"EUR_1\", \"EUR_2\", \"EUR_3\", \"EUR_4\", \"EUR_5\"),                    nea = c(\"NEA_1\", \"NEA_2\"))) #> # A tibble: 2 × 2 #>   set       D #>   <chr> <dbl> #> 1 eur   -1.16 #> 2 nea    2.18"},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"Extract data frame coordinates ancestry tracts given tree sequence.","code":""},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"","code":"ts_tracts(   ts,   census,   squashed = TRUE,   source = NULL,   target = NULL,   quiet = FALSE )"},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"ts Tree sequence object class slendr_ts census Census time. See documentation linked Details information. slendr-specific tree sequence provided ts, census time expected given slendr model-specific time units, must correspond gene-flow event encoded model. squashed ancestry tracts squashed (.e., continuous tracts can traced different ancestral nodes merged)? Default TRUE. FALSE, effectively continuous ancestry tracts split individual segments, assigned specific ancestral node ID (recorded column ancestor_id). source source population extract tracts ? NULL (default), ancestry tracts populations contributing gene flow census time reported. Otherwise, ancestry tracts specified source populations extracted. Note option ignored non-slendr tree sequences! target Similar purpose source , except filters tracts discovered target population(s) quiet default summary output tspop Python package silenced? Default FALSE.","code":""},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"data frame containing coordinates ancestry tracts","code":""},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"function implements R-friendly interface algorithm extracting ancestry tracts provided Python module tspop https://tspop.readthedocs.io/en/latest/ developed Georgia Tsambos. Please make sure cite paper describes algorithm detail: https://academic.oup.com/bioinformaticsadvances/article/3/1/vbad163/7429395. technical details, see also tutorial : https://tspop.readthedocs.io/en/latest/basicusage.html. general, using function slendr-generated tree sequence, please aware output changes slightly get running pure tspop.get_pop_ancestry() Python. First, ts_tracts() populates output data frame additional metadata (names individuals populations). Additionally, slendr models, specifically designed return ancestry tracts originating ancestral population contributed ancestry gene-flow event started specific time (.e., scheduled model via gene_flow()) function. return every single ancestry tracts present tree sequence every single sample node (every single potential ancestry population) tspop.get_pop_ancestry() Python method. said, run tree sequence originate slendr simulation, behavior ts_tracts() identical underlying tspop.get_pop_ancestry(). current version slendr, ts_tracts() works slendr/msprime sequences slendr/SLiM tree sequences. Support slendr-generated SLiM tree sequences development. Tracts tree sequences originating non-slendr msprime SLiM simulations restricted way , mentioned previous paragraph, ts_tracts() situation effectively reduces standard tspop.get_pop_ancestry() call.","code":""},{"path":"https://www.slendr.net/reference/ts_tracts.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract ancestry tracts from a tree sequence (EXPERIMENTAL) — ts_tracts","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env(quiet = TRUE)  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_msprime.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(file = slendr_ts, model = model)  # extract Neanderthal ancestry tracts (i.e. those corresponding to the # census event at the gene-flow time at 55000 kya as scheduled by # the simulation which produced the tree sequence) nea_tracts <- ts_tracts(ts, census = 55000, source = \"NEA\") #>  #> PopAncestry summary #> Number of ancestral populations: \t4 #> Number of sample chromosomes: \t\t26 #> Number of ancestors: \t\t\t763 #> Total length of genomes: \t\t130000000.000000 #> Ancestral coverage: \t\t\t120000000.000000 #>  nea_tracts #> # A tibble: 42 × 8 #>    name  node_id pop   source_pop    left   right length source_pop_id #>    <chr>   <dbl> <fct> <fct>        <dbl>   <dbl>  <dbl>         <dbl> #>  1 EUR_1      16 EUR   NEA          44049  101210  57161             2 #>  2 EUR_1      16 EUR   NEA         244041  280650  36609             2 #>  3 EUR_1      16 EUR   NEA        4636048 4736790 100742             2 #>  4 EUR_1      16 EUR   NEA        4756344 4850072  93728             2 #>  5 EUR_1      17 EUR   NEA         106906  268610 161704             2 #>  6 EUR_1      17 EUR   NEA        1420672 1432252  11580             2 #>  7 EUR_1      17 EUR   NEA        3438894 3511366  72472             2 #>  8 EUR_1      17 EUR   NEA        3624298 3682434  58136             2 #>  9 EUR_1      17 EUR   NEA        4139989 4239794  99805             2 #> 10 EUR_1      17 EUR   NEA        4318138 4351616  33478             2 #> # ℹ 32 more rows"},{"path":"https://www.slendr.net/reference/ts_tree.html","id":null,"dir":"Reference","previous_headings":"","what":"Get a tree from a given tree sequence — ts_tree","title":"Get a tree from a given tree sequence — ts_tree","text":"information optional keyword arguments see tskit documentation: https://tskit.dev/tskit/docs/stable/python-api.html#-treesequence-class","code":""},{"path":"https://www.slendr.net/reference/ts_tree.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Get a tree from a given tree sequence — ts_tree","text":"","code":"ts_tree(ts, i, mode = c(\"index\", \"position\"), ...)"},{"path":"https://www.slendr.net/reference/ts_tree.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Get a tree from a given tree sequence — ts_tree","text":"ts Tree sequence object class slendr_ts Position tree tree sequence. mode = \"index\", -th tree returned (zero-based indexing tskit), mode = \"position\", tree covering -th base simulated genome returned (, tskit's indexing). mode argument interpreted? Either \"index\" -th tree sequence genealogies, \"position\" along simulated genome. ... Additional keyword arguments accepted tskit.TreeSequence.tskit.TreeSequence.at_index methods","code":""},{"path":"https://www.slendr.net/reference/ts_tree.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Get a tree from a given tree sequence — ts_tree","text":"Python-reticulate-based object class tskit.trees.Tree","code":""},{"path":"https://www.slendr.net/reference/ts_tree.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Get a tree from a given tree sequence — ts_tree","text":"","code":"check_dependencies(python = TRUE, quit = TRUE) # dependencies must be present  init_env() #> The interface to all required Python modules has been activated.  # load an example model with an already simulated tree sequence slendr_ts <- system.file(\"extdata/models/introgression_slim.trees\", package = \"slendr\") model <- read_model(path = system.file(\"extdata/models/introgression\", package = \"slendr\"))  # load the tree-sequence object from disk ts <- ts_load(slendr_ts, model)  # extract the zero-th tree in the tree sequence tree <- ts_tree(ts, i = 0)  # extract the tree at a position in the tree sequence tree <- ts_tree(ts, i = 100000, mode = \"position\")"},{"path":"https://www.slendr.net/reference/ts_vcf.html","id":null,"dir":"Reference","previous_headings":"","what":"Save genotypes from the tree sequence as a VCF file — ts_vcf","title":"Save genotypes from the tree sequence as a VCF file — ts_vcf","text":"Save genotypes tree sequence VCF file","code":""},{"path":"https://www.slendr.net/reference/ts_vcf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Save genotypes from the tree sequence as a VCF file — ts_vcf","text":"","code":"ts_vcf(ts, path, chrom = NULL, individuals = NULL)"},{"path":"https://www.slendr.net/reference/ts_vcf.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Save genotypes from the tree sequence as a VCF file — ts_vcf","text":"ts Tree sequence object class slendr_ts path Path VCF file chrom Chromosome name written CHROM column VCF individuals character vector individuals tree sequence. missing, individuals present tree sequence saved.","code":""},{"path":"https://www.slendr.net/reference/ts_vcf.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Save genotypes from the tree sequence as a VCF file — ts_vcf","text":"return value, called side effects","code":""},{"path":"https://www.slendr.net/reference/world.html","id":null,"dir":"Reference","previous_headings":"","what":"Define a world map for all spatial operations — world","title":"Define a world map for all spatial operations — world","text":"Defines either abstract geographic landscape (blank containing user-defined landscape) using real Earth cartographic data Natural Earth project (https://www.naturalearthdata.com).","code":""},{"path":"https://www.slendr.net/reference/world.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Define a world map for all spatial operations — world","text":"","code":"world(   xrange,   yrange,   landscape = \"naturalearth\",   crs = NULL,   scale = c(\"small\", \"medium\", \"large\") )"},{"path":"https://www.slendr.net/reference/world.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Define a world map for all spatial operations — world","text":"xrange Two-dimensional vector specifying minimum maximum horizontal range (\"longitude\" using real Earth cartographic data) yrange Two-dimensional vector specifying minimum maximum vertical range (\"latitude\" using real Earth cartographic data) landscape Either \"blank\" (blank abstract geography), \"naturalearth\" (real Earth geography) object class sf defining abstract geographic features world crs EPSG code coordinate reference system use spatial operations. CRS assumed default (NULL), implying abstract landscape tied real-world geographic region (landscape = \"blank\" landscape custom-defined geographic landscape), implying WGS-84 (EPSG 4326) coordinate system real Earth landscape defined (landscape =   \"naturalearth\"). scale Natural Earth geographic data used (.e. landscape =   \"naturalearth\"), parameter determines resolution data used. value \"small\" corresponds 1:110m data provided package, values \"medium\" \"large\" correspond 1:50m 1:10m respectively downloaded internet. Default value \"small\".","code":""},{"path":"https://www.slendr.net/reference/world.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Define a world map for all spatial operations — world","text":"Object class slendr_map, encodes standard spatial object class sf additional slendr-specific attributes requested x-range y-range.","code":""},{"path":"https://www.slendr.net/reference/world.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Define a world map for all spatial operations — world","text":"","code":"# create a blank abstract world 1000x1000 distance units in size blank_map <- world(xrange = c(0, 1000), yrange = c(0, 1000), landscape = \"blank\")  # it is possible to construct custom landscapes (islands, corridors, etc.) island1 <- region(\"island1\", polygon = list(c(10, 30), c(50, 30), c(40, 50), c(0, 40))) island2 <- region(\"island2\", polygon = list(c(60, 60), c(80, 40), c(100, 60), c(80, 80))) island3 <- region(\"island3\", center = c(20, 80), radius = 10) archipelago <- island1 %>% join(island2) %>% join(island3)  custom_map <- world(xrange = c(1, 100), c(1, 100), landscape = archipelago)  # real Earth landscapes can be defined using freely-available Natural Earth # project data and with the possibility to specify an appropriate Coordinate # Reference System, such as this example of a map of Europe  real_map <- world(xrange = c(-15, 40), yrange = c(30, 60), crs = \"EPSG:3035\") #> Reading layer `ne_110m_land' from data source  #>   `/private/var/folders/70/b_q2zdh116b9pfg29p03sx600000gn/T/RtmpqG5m0G/naturalearth/ne_110m_land.shp'  #>   using driver `ESRI Shapefile' #> Simple feature collection with 127 features and 3 fields #> Geometry type: POLYGON #> Dimension:     XY #> Bounding box:  xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513 #> Geodetic CRS:  WGS 84"},{"path":"https://www.slendr.net/news/index.html","id":"slendr-091","dir":"Changelog","previous_headings":"","what":"slendr 0.9.1","title":"slendr 0.9.1","text":"new helper function get_env() now returns name built-slendr Python environment (without activating ). (#162ccc) clear_env() now new argument = (TRUE|FALSE) allows deleting slendr Python environments. Previously, function always removed recent environment, lead accumulation potentially large number slendr environments time. (#8707b9) plot_model() new optional argument samples = – set result sampling schedule created schedule_sampling() – visualize counts samples recorded given time-point. (#d72ac5) msprime dependency slendr updated version 1.3.1. result, setup_env() re-run update internal slendr Python environment. (#dcb83d)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-090","dir":"Changelog","previous_headings":"","what":"slendr 0.9.0","title":"slendr 0.9.0","text":"CRAN release: 2024-02-08 full support running SLiM msprime simulations slendr analyzing tree sequences using tskit interface implemented. Please note Windows support still rather experimental – internal slendr test suite currently assumes SLiM installed using msys2 system described section 2.3.1 SLiM manual means installing SLiM (via conda) might require additional adjustments. fallback option form slim_path= argument slim() function can used non-standard SLiM installation circumstances. convenience, please add path directory containing slim.exe binary PATH variable editing C:/Users/<username>/Documents/.Renviron file accordingly. See relevant section Windows installation slendr documentation additional information. Feedback Windows functionality bug reports highly appreciated via GitHub issues! Many thanks @GKresearch @rdinnager huge help making Windows port happen! (PR #149) trivial change made slendr’s SLiM back-end script fixing issue introduced SLiM 4.1 upgrade (see changelog version 0.8.1 ). expected lead different simulation outputs two versions slendr (0.8.2 vs 0.8.1) SLiM (4.1 vs 4.0.1) used. (PR #148) msprime internal dependency slendr updated 1.3.0, Python 3.12. result, loading slendr, users prompted re-run setup_env() make sure dedicated slendr Python environment fully updated. time, prevents failing installation (least) M1 macOS using pip. (#5ce212, #a210d4)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-081","dir":"Changelog","previous_headings":"","what":"slendr 0.8.1","title":"slendr 0.8.1","text":"CRAN release: 2024-01-15 Fixed issue apparent contradiction time direction models range expansion scheduled within time interval together associated “locked-” changes population size time interval. (#d2a29e) introduction tspop installable via pip (see changelog previous version) caused GLIBCXX-related errors conda pip dependencies related pandas Python package. work around issue, setup_env() longer installs pandas conda regardless setting pip = TRUE|FALSE parameter. Instead, pandas installed via pip single step tspop installed. (#cbe960) WARNING: SLiM 4.1 just released includes couple backwards incompatible changes related implementation spatial maps prevent current version slendr’s slim() function working correctly. rely functionality provided slim() function, use SLiM 4.0. (Note want multiple versions SLiM system, can either use slim_path = argument slim() specify $PATH required version SLiM ~/.Renviron file just like normal circumstances). Porting slendr SLiM 4.1 worked .","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-080","dir":"Changelog","previous_headings":"","what":"slendr 0.8.0","title":"slendr 0.8.0","text":"CRAN release: 2023-12-07 order support new ts_tracts() function backed tspop module (see item ), new slendr Python environment required. , users run setup_env() get required Python dependencies now installed internal virtual environment named Python-3.11_msprime-1.2.0_tskit-0.5.6_pyslim-1.0.4_tspop-0.0.2. (#b5330c) Experimental support tspop link-ancestors algorithm detection ancestry tracts form new slendr function ts_tracts(). works slendr-generated msprime tree sequences “pure” msprime SLiM tree sequences (slendr-generated SLiM tree sequences), tested toy models. Note: tspop Python module published conda. order set new slendr Python environment, run setup_env(pip = TRUE) make sure Python dependencies installed pip instead conda. (PR #145) Updated Python dependencies (bugfix pyslim release v1.0.4 tskit v0.5.6, latter due broken jsonschema dependency tskit). (#001ee5) Experimental support manually created spatial tree sequences. (PR #144)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-072","dir":"Changelog","previous_headings":"","what":"slendr 0.7.2","title":"slendr 0.7.2","text":"CRAN release: 2023-08-08 new function ts_names() added, avoiding need extremely frequent (, unfortunately, cumbersome) trick getting named lists individual symbolic names ts_samples(ts) %>% split(., .[[split]]) %>% lapply([[, \"name\") confusing experienced R users. (#7db6ea) Fixed broken concatenation symbolic sample names tree-sequence statistic functions, provided unnamed single-element lists character vectors. (#b3c650) plot_model() now argument file =, making possible save visualization model without actually opening plotting device. can useful particularly working remote server, order avoid often slow X11 rendering. (#e60078) plot_model() now argument order = allowing override default -order ordering populations along x-axis. (#7a10ea)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-071","dir":"Changelog","previous_headings":"","what":"slendr 0.7.1","title":"slendr 0.7.1","text":"CRAN release: 2023-07-14 Starting release, *spatial* simulation data analysis functionality slendr conditional presence R geospatial packages sf, stars, rnaturalearth system. means users able install slendr (use non-spatial functionality) even without R packages installed. said, nothing really changes practice: spatial features slendr just one install.packages(c(\"sf\", \"stars\", \"rnaturalearth\")) away! difference slendr doesn’t try installation, users instructed (needed) package loaded. (#7a10ea) spatial dependencies present spatial slendr function called regardless (world(), move(), etc.), error message printed information install spatial dependencies via install.packages() . ? ’s true main reason slendr’s existence ability simulate spatio-temporal data realistic landscapes via SLiM. However, practice, “average” uses slendr wild (classrooms!) rely traditional, non-spatial interface, spatial features used comparatively rarely moment (except cutting-edge exploratory research). Given setting spatial dependencies can bit hurdle, decided make dependencies optional, rather force every user go process installation whether need spatial features . function check_dependencies() now exported can used check whether slendr Python environment () SLiM () present. useful software building upon slendr, normal users can freely ignore . (#6ae6ce) path file tree sequence loaded now tracked internally via attr(<tree sequence>, \"path\") attribute. Note implemented purposes clean large-scale simulation studies (facilitated demografr) mostly internal feature, considered experimental. (#f181a2) Attempts resize population right time split (led issues simulations) now prevented. (#f181a2) Fix minor issue preventing sampling msprime population right time creation. (#aea231)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-070","dir":"Changelog","previous_headings":"","what":"slendr 0.7.0","title":"slendr 0.7.0","text":"CRAN release: 2023-06-26 emergency upgrade match latest pyslim 1.0.3 due serious bug recapitation. See extensive discussion process identification bug eventual fix. brief summary practical consequences bug, see thread pyslim’s developer formal announcement . change require re-run setup_env() order update slendr’s Python internals creating new internal Python virtual environment. (#45539a) potential issue parent population scheduled removal daughter population splits now caught moment daughter population() call rather simulation slim() run. (#0791b5) function plot_model() new argument gene_flow=<TRUE|FALSE> determines whether gene-flow arrows visualized . (#104aa6) possibility perform recapitation, simplification, mutation tree sequence right inside call ts_load() (providing recapitate = TRUE, simplify = TRUE, mutate = TRUE, together arguments) now removed. motivation change realization benefit things like ts_load(\"<path>\", recapitate = TRUE, Ne = ..., recombination_rate = ...) ts_load(\"<path>\") %>% ts_recapitate(Ne = ..., recombination_rate = ...), frequent confusion recapitate = TRUE switches forgotten user. slendr teaching material actively used research codebases know use latter, explicit, pipeline approach anyway, one example reduncancy harm good. (#ad82ee) Note: Loading library(slendr) prompt message “legacy packages maptools, rgdal, rgeos, underpinning sp package, just loaded, retire October 2023. […].” internal business packages used slendr unfortunately silenced slendr’s side. ’s reason panic, can safely ignore . Apologies unnecessary noise.","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-060","dir":"Changelog","previous_headings":"","what":"slendr 0.6.0","title":"slendr 0.6.0","text":"CRAN release: 2023-06-07 relatively large update, unfortunately released haste due retirement rgdal package – significant dependency entire spatial R ecosystem phased effort move towards modern low-level geospatial architecture. Although slendr depend rgdal, many dependencies used (won’t short term, hence push remove rgdal dependency). significant update addition IBD functionality tskit, described . However, large part functionality extensively tested considered extremely experimental stage. like use , might safer either wait later release IBD functionality stable, use underlying, battle-tested Python implementation tskit. ts_ibd() now returns ID number MRCA node pair nodes sharing given IBD segment, well TMRCA node. (#7e2825) Trivial parameter errors caught population() calls rather simulation (solving minor issues discovered via big simulation runs development demografr). (#e33373) Fix error plotting exponential resizes last “present”. (#4c49a4) ts_ibd() longer gives obscure error = provided named list individuals’ names (instead expected unnamed list). names list elements used way, error happens somewhere deeply R->Python translation layer inside reticulate ’s need users concern . (#7965e4) Population size parameters times now explicitly converted integer numbers. internal, formal change (conversion happening implicitly inside SLiM engine anyway) now explicitly stated, also documentation relevant function. (#b7e89e) Population names now restricted strings also valid Python identifiers. Although restriction needed msprime back end slendr (SLiM), makes sense keep things tidy unified. fixes msprime crashing ValueError: population name must valid Python identifier. (#4ef518) layout algorithm plot_model() improved significantly. (PR #135). new optional argument run = added slim() msprime(). set TRUE (default), engines operate usual way. set FALSE, simulation run functions simply print command-line command execute engine question (returning CLI command invisibly). (#2e5b85) following start-note longer shown upon calling library(slendr): Users call init_env() manually activate Python environment slendr (see note version 0.5.0 extended explanation). ts_simplify() now accepts optional arguments keep_unary keep_unary_in_individuals (see official tskit docs detail) (#1b2112) Fix ts_load() failing load slendr-produced tree sequences simplified smaller set sampled individuals (reported ). issue caused incompatible sizes sampling table (always form used simulation) table individuals stored tree sequence simplification (potentially containing smaller set individuals original sampling table). fix , slendr tree sequence objects now track information individuals regarded “samples” (.e. symbolic names) maintained simplification, serialization loading, used slendr’s internal machinery join operations. (PR #137) Metadata summary ts_nodes() results longer printed whenever typed R console. Instead, summary can obtained explicit call summary() ts_nodes() tables. (#01af51 ts_tree() ts_phylo() now extract trees based tskit’s zero-based indexing #554e13. ts_simplify() now accepts filter_nodes = TRUE|FALSE, behavior tskit’s method #f07ffed.","code":"NOTE: Due to Python setup issues on some systems which have been     causing trouble particularly for novice users, calling library(slendr)     no longer activates slendr's Python environment automatically.     In order to use slendr's msprime back end or its tree-sequence     functionality, users must now activate slendr's Python environment     manually by executing init_env() after calling library(slendr).     (This note will be removed in the next major version of slendr.)"},{"path":"https://www.slendr.net/news/index.html","id":"slendr-051","dir":"Changelog","previous_headings":"","what":"slendr 0.5.1","title":"slendr 0.5.1","text":"CRAN release: 2023-03-09 minor release implements emergency fix CRAN warning suddenly popped latest CRAN checks. (#5600a4) new function ts_ibd() added, representing R interface tskit method TreeSequence.ibd_segments(). However, note ts_ibd() returns IBD results data frame (optionally, spatially annotated sf data frame). function operate around iteration, Python counterpart tskit. next major version slendr, function considered experimental. (PR #123)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-050","dir":"Changelog","previous_headings":"","what":"slendr 0.5.0","title":"slendr 0.5.0","text":"CRAN release: 2023-02-02 Minor breaking change! Python environments slendr longer automatically activated upon calling library(slendr)! Using coalescent msprime back end slendr’s tree-sequence functions now requires making explicit call new function init_env() library(slendr) executed. (PR #102) Motivation change: small proportion users experiencing issues broken conda environments various issues Python virtual environments general. ’s hard guess frequent , experience workshops courses suggests perhaps 1 20 users experiencing Python issues hindered ability use slendr .(Fun fact: first user-submitted GitHub issue upon releasing first version slendr R package … Python virtual environment issue). Explanation: Activating Python environments automatically upon calling library(slendr) popular feature hid away complexities R-Python interface powers slendr’s tree-sequence functionality. particularly convenient many slendr users, particularly experience Python . Unfortunately, cases Python virtual environments tskit/msprime/pyslim user’s system ended corrupted (anything else Python level got broken), automatic Python environment activation performed library(slendr) call failed slendr even loaded. Sadly, completely pulled rug slendr nothing done perspective (issue happened low-level layer embedded-Python slendr ’ve loaded R). Solving issues difficult experienced users, many slendr users experience Python , never used conda, don’t understand concept “Python virtual environments” R-Python interface works. ! , slendr R package. Splitting Python virtual environment activation step init_env() function means library(slendr) now always succeeds (regardless potential underlying Python issues user’s sytem), making much easier diagnose fix Python problems R package loaded. , recap: library(slendr) longer activates slendr’s isolated Python virtual environment. order simulate tree sequences analyse using interface tskit, necessary call init_env(). function performs Python-activation steps library(slendr) used call automagically earlier slendr versions. change scripts necessary. Related previous point: slendr now requires Python 3.11, msprime 1.2.0, tskit 0.5.4, pyslim 1.0.1, keep recent releases Python dependencies. , presents hassle user, thing required re-running setup_env(). (PR #112). named list provided sample_sets = argument oneway statistic function, names used set column resulting data frame even single samples used. (#2a6781) now possible non-spatial populations otherwise spatial model. course, plotting models map, spatial components model plotted slendr give warning. absolutely sure users intends , slendr also give warning running compile_model() models like . Please consider option experimental time-hard predict edge cases might break (unit tests documentation tests passing though). Feedback welcome. (PR #112). now possible label groups samples slendr’s tskit interface functions make data frames statistics results readable. example, running ts_f3(ts, = c(\"p1_1\", \"p1_2\", \"p1_3\"), B = c(\"p2_1\", \"p2_3\"), C = c(\"p3_1\", \"p3_2\", \"p3_\")) resulted following data-frame output: gets unwieldy rather quickly, especially dozens hundreds samples grouped together populations. new syntax allows following shortcut via customised group names leveraging standard named list functionality R: readable line tskit-interface functions slendr used functionality via sample_sets = argument (ts_divergence(), ts_diversity(), etc.). (#ac5e484) default state parent = argument population() now NULL instead \"ancestor\". prevents silly surprising clashes situation population’s name really “ancestor”. change internally populations ancestral, splits data frame element slendr model object includes population carries formal “ancestral parent population” \"__pop_is_ancestor\" instead just \"ancestor\". Note internal implementation detail something particularly involve user. Still, somehow using slendr’s internal data structures, keep mind. (#f8a39a2)","code":"> ts_f3(ts, A = c(\"p1_1\", \"p1_2\", \"p1_3\", \"p1_4\", \"p1_5\"),             B = c(\"p2_1\", \"p2_2\", \"p2_3\"),             C = c(\"p3_1\", \"p3_2\", \"p3_3\", \"p3_4\"))  # A tibble: 1 × 4   A                        B              C                         f3   <chr>                    <chr>          <chr>                  <dbl> 1 p1_1+p1_2+p1_3+p1_4+p1_5 p2_1+p2_2+p2_3 p3_1+p3_2+p3_3+p3_4 0.000130 > ts_f3(ts, A = list(group_one = c(\"p1_1\", \"p1_2\", \"p1_3\", \"p1_4\", \"p1_5\")),             B = list(group_two = c(\"p2_1\", \"p2_2\", \"p2_3\")),             C = list(group_three = c(\"p3_1\", \"p3_2\", \"p3_3\", \"p3_4\"))) # A tibble: 1 × 4   A         B         C                 f3   <chr>     <chr>     <chr>          <dbl> 1 group_one group_two group_three 0.000130"},{"path":"https://www.slendr.net/news/index.html","id":"slendr-040","dir":"Changelog","previous_headings":"","what":"slendr 0.4.0","title":"slendr 0.4.0","text":"CRAN release: 2022-09-30 msprime() function now makes sure given slendr model can fully coalesce single common ancestor population. Previously, multiple ancestral populations created parent = \"ancestor\" cause infinite simulation plugged msprime() backend. (#095b124) initial size population emerges split another population now printed population history summary R console. (#6525bf3) couple fixes support loading, processing, plotting “manually” created tree sequences implemented (see ). sure practically useful, ’s important able load even “pure” tree sequences simulators SLiM msprime. set unit tests added, making sure minimalist nodes & edges table can loaded, well nodes & edges & individuals, plus tables populations sites & mutations. PRs extensive unit tests bug reports tree sequences failing load appreciated! code handling cases “manually-created” tree sequences missing individual table, missing populations table, etc. seems especially brittle moment (#79adf14). -1 value missing value indicator used tskit now replaced R-like NA various tree-sequence tables (annotated slendr original tskit ) (#79adf14). Relative paths now expanded ts_save() (#382e0b7). slendr models can now optionally compiled without serialization disk. works msprime() coalescent back end much faster cases huge number simulations needs run non-serialized models, msprime() now calls back end engine directly R-Python interface (rather command line) output tree sequences saved disk, rather passed Python-R interface directly memory (PR #112). Deprecated argument sampling = functions slim() msprime() now permanently removed favour samples = argument (#0757b6e). Avoid unnecessary array type tskit results returned via reticulate. Numeric vectors (columns data frames numerical results) obtained way simple R numeric vector (#5101b39). One-way multi-way statistics results now returned simple numerical vectors. Previously, results returned type array despite “looking” vectors (values returned R reticulate-Python layer), caused unnecessary annoyances type-conversions R side things even intended (#403df3b). Computing population genetic statistics named samples present tree sequence (likely typos) now correctly caught reported error (#da7e0bb).","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-030","dir":"Changelog","previous_headings":"","what":"slendr 0.3.0","title":"slendr 0.3.0","text":"CRAN release: 2022-08-19 SLiM 4.0 now required running simulations slim() engine. want run slendr simulations SLiM (spatial non-spatial), need upgrade SLiM installation. SLiM 3.7.1 version longer supported upcoming new slendr spatial features depend SLiM 4.x maintaining two functionally identical yet syntactically different back ends feasible (PR #104). time SLiM 4.0 release, new versions Python modules msprime, tskit pyslim also released. fact, able work SLiM 4.0 tree sequences properly, Python modules must upgraded well. Next time load library(slendr), prompted setup new updated Python environment can easily running setup_env(). Experimental support running coalescent msprime simulations analysing tree-sequence data using tskit Windows platform now implemented (PR #102).","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-020","dir":"Changelog","previous_headings":"","what":"slendr 0.2.0","title":"slendr 0.2.0","text":"CRAN release: 2022-08-09 slendr now CRAN! Big changes way tree-sequence outputs handled slendr default. See comment extended description examples change. (PR #100). Briefly, simulation functions slim() msprime() now return tree-sequence object default (can switched setting load = FALSE), avoiding need always run ts <- ts_load(model) previously. time, parameter output = can now used slim() msprime() specify location tree-sequence file saved (temporary file default). slendr’s tree-sequence R interface tskit Python module generalized load, process, analyze tree sequences non-slendr models! means users can use slendr R package even analyzing tree sequences coming standard msprime SLiM scripts, including spatial capabilities available slendr tree sequences far. Please note generalization still rather experimental might corner cases tree sequence msprime SLiM script load properly leads errors. happens, please open GitHub issue script question attached. (PR #91) Removed functions function arguments originally deprecated renaming phase pre-preprint refactoring. affects compile, boundary, dispersal, expand, geneflow, plot.slendr, plot_graph, read, sampling, shrink. Similarly, deprecated dir argument compile_model now path, geneflow argument compile_model now gene_flow, _dist suffix removed competition_dist, mate_dist, dispersal_dist. get error missing function function argument code used work ancient version slendr, . (#985b451) setting isolated Python environment using setup_env(), slendr now makes decision whether install Python dependencies using pip (critical osx-arm64 conda msprime/tskit unfortunately currently broken) conda (every platform). can still influenced user using pip = <TRUE|FALSE> argument, now change default behavior ARM64 Mac. (#54a413d) name default slendr Python environment now shortened even , redundant _pandas prefix now dropped. Users notified upon calling library(slendr) new environment created. OK, ’s bug. (#54a413d) format default slendr Python environment now msprime-<version>_tskit-<version>_pyslim-<version>_pandas, dropping slendr_ prefix. paves way towards future non-slendr tskit R package, share Python environment slendr (R packages go hand hand). isn’t really user-facing change, except calling setup_env() suggests creating new Python environment library(slendr) appear slendr environment yet present. Calling setup_env() creating new Python environment scratch solve problem. (#eb05180) xrange yrange parameters world() now enforced two-dimensional numeric vectors, avoiding unnecessary issues misspecified longitude/latitude (#df95369) argument sampling = slim() msprime() now renamed samples = (#adf4e0d). automated setup_env() function creating dedicated mini Python environments slendr now installs packages using pip default. Reason: rate conda failures dependency conflicts (even trivial case installing nothing msprime + tskit + pyslim + pandas) high rely . option use conda package installations setup_env() still , users must explicitly call setup_env(pip = FALSE) get behavior. Note conda still used means install Python ! change affects way Python modules installed dedicated slendr Python environment, installation Python . (#81be1a7) name automatically created slendr-specific Python environment now composed names versions Python modules installed. makes possible naturally upgrade slendr Python dependencies case tskit / msprime / pyslim folks upgrade packages. case, slendr user upgrades slendr package (new version requires newer versions Python modules), slendr simply recommend create new Python environment without additional effort part. (#81be1a7) code setup_env() simplified bare essentials. Now serves way auto-setup dedicated, isolated Python installation slendr environment. interface install Python modules custom-defined Python environment created outside R removed functionality necessary – custom environments can easily activated calling reticulate::use_virtualenv reticulate::use_condaenv. (#30f24b9) Python users want use custom Python environments msprime, tskit, pyslim, can silence suggestion use setup_env() printed library(slendr) call setting options(slendr.custom_env = TRUE). (#30f24b9) argument sim_length = now renamed simulation_length =. accepted moment using old name simply inform user future deprecation. (#56491fb) Extensive set runnable examples including figures built-pre-compiled example model added documentation. (#395df62c)","code":""},{"path":"https://www.slendr.net/news/index.html","id":"slendr-010","dir":"Changelog","previous_headings":"","what":"slendr 0.1.0","title":"slendr 0.1.0","text":"First numbered version slendr celebrate bioRxiv preprint. 🥳 🎉","code":""}]
diff --git a/docs/sitemap.xml b/docs/sitemap.xml
index 506c7fbb8..43518c83b 100644
--- a/docs/sitemap.xml
+++ b/docs/sitemap.xml
@@ -87,6 +87,9 @@
   <url>
     <loc>https://www.slendr.net/reference/gene_flow.html</loc>
   </url>
+  <url>
+    <loc>https://www.slendr.net/reference/get_env.html</loc>
+  </url>
   <url>
     <loc>https://www.slendr.net/reference/index.html</loc>
   </url>