Tests for various files with generic graph

src/Test/Test.jl

@@ -8,7 +8,7 @@ module Test
 
 using Graphs
 
-export GenericEdge, GenericGraph, GenericDiGraph
+export GenericEdge, GenericGraph, GenericDiGraph, generic_graph
 
 """
     GenericEdge <: Graphs.AbstractEdge
@@ -37,7 +37,7 @@ struct GenericGraph{T} <: Graphs.AbstractGraph{T}
 end
 
 function GenericGraph(elist::Vector{Graphs.SimpleGraphEdge{T}}) where {T<:Integer}
-    GenericGraph{T}(SimpleGraph(elist))
+    return GenericGraph{T}(SimpleGraph(elist))
 end
 
 """
@@ -50,8 +50,18 @@ struct GenericDiGraph{T} <: Graphs.AbstractGraph{T}
     g::SimpleDiGraph{T}
 end
 
+"""
+    generic_graph(g::Union{SimpleGraph, SimpleDiGraph})
+
+Return either a GenericGraph or GenericDiGraph that wraps a copy of g.
+"""
+function generic_graph(g::Union{SimpleGraph,SimpleDiGraph})
+    g = copy(g)
+    return is_directed(g) ? GenericDiGraph(g) : GenericGraph(g)
+end
+
 function GenericDiGraph(elist::Vector{Graphs.SimpleDiGraphEdge{T}}) where {T<:Integer}
-    GenericDiGraph{T}(SimpleDiGraph(elist))
+    return GenericDiGraph{T}(SimpleDiGraph(elist))
 end
 
 Graphs.is_directed(::Type{<:GenericGraph}) = false

src/degeneracy.jl

@@ -324,6 +324,10 @@ Int64[]
 ```
 """
 function k_corona(g::AbstractGraph, k; corenum=core_number(g))
+
+    # TODO k_corona does not correctly work for all AbstractGraph as
+    # it relies on induced_subgraph, which does not work on any AbstractGraph
+
     kcore = k_core(g, k)
     kcoreg = g[kcore]
     kcoredeg = degree(kcoreg)

src/operators.jl

@@ -1,3 +1,6 @@
+# TODO most of the operators here do not work with any AbstractGraph yet
+# as they require cloning and modifying graphs.
+
 """
     complement(g)
 
@@ -402,25 +405,16 @@ end
 # """Provides multiplication of a graph `g` by a vector `v` such that spectral
 # graph functions in [GraphMatrices.jl](https://github.com/jpfairbanks/GraphMatrices.jl) can utilize Graphs natively.
 # """
-function *(g::Graph, v::Vector{T}) where {T<:Real}
-    length(v) == nv(g) || throw(ArgumentError("Vector size must equal number of vertices"))
-    y = zeros(T, nv(g))
-    for e in edges(g)
-        i = src(e)
-        j = dst(e)
-        y[i] += v[j]
-        y[j] += v[i]
-    end
-    return y
-end
-
-function *(g::DiGraph, v::Vector{T}) where {T<:Real}
+function *(g::AbstractGraph, v::Vector{T}) where {T<:Real}
     length(v) == nv(g) || throw(ArgumentError("Vector size must equal number of vertices"))
     y = zeros(T, nv(g))
     for e in edges(g)
         i = src(e)
         j = dst(e)
         y[i] += v[j]
+        if !is_directed(g)
+            y[j] += v[i]
+        end
     end
     return y
 end
@@ -481,7 +475,7 @@ julia> size(g, 3)
 1
 ```
 """
-size(g::Graph, dim::Int) = (dim == 1 || dim == 2) ? nv(g) : 1
+size(g::AbstractGraph, dim::Int) = (dim == 1 || dim == 2) ? nv(g) : 1
 
 """
     sum(g)

src/traversals/bfs.jl

@@ -59,7 +59,11 @@ function _bfs_parents(g::AbstractGraph{T}, source, neighborfn::Function) where {
     end
     while !isempty(cur_level)
         @inbounds for v in cur_level
-            @inbounds @simd for i in neighborfn(g, v)
+            # TODO we previously used @simd on the loop below, but this would fail
+            # if the result of neighorfn(g, v) would not implement firstindex
+            # If @simd really has a performance advantage, then maybe we make
+            # two different cases here.
+            @inbounds for i in neighborfn(g, v)
                 if !visited[i]
                     push!(next_level, i)
                     parents[i] = v

test/connectivity.jl

@@ -7,7 +7,7 @@
     add_edge!(gx, 8, 9)
     add_edge!(gx, 10, 9)
 
-    for g in testgraphs(gx)
+    for g in test_generic_graphs(gx)
         @test @inferred(!is_connected(g))
         cc = @inferred(connected_components(g))
         label = zeros(eltype(g), nv(g))
@@ -20,18 +20,18 @@
         @test cclab[8] == [8, 9, 10]
         @test length(cc) >= 3 && sort(cc[3]) == [8, 9, 10]
     end
-    for g in testgraphs(g6)
+    for g in test_generic_graphs(g6)
         @test @inferred(is_connected(g))
     end
 
     g10 = SimpleDiGraph(4)
     add_edge!(g10, 1, 3)
     add_edge!(g10, 2, 4)
-    for g in testdigraphs(g10)
+    for g in test_generic_graphs(g10)
         @test @inferred(is_bipartite(g))
     end
     add_edge!(g10, 1, 4)
-    for g in testdigraphs(g10)
+    for g in test_generic_graphs(g10)
         @test @inferred(is_bipartite(g))
     end
 
@@ -45,7 +45,7 @@
             end
             if !has_edge(g, i, j)
                 add_edge!(g, i, j)
-                @test @inferred(is_bipartite(g))
+                @test @inferred(is_bipartite(GenericDiGraph(g)))
             end
         end
     end
@@ -66,7 +66,7 @@
     add_edge!(h, 7, 6)
     add_edge!(h, 8, 4)
     add_edge!(h, 8, 7)
-    for g in testdigraphs(h)
+    for g in test_generic_graphs(h)
         @test @inferred(is_weakly_connected(g))
         scc = @inferred(strongly_connected_components(g))
         scc_k = @inferred(strongly_connected_components_kosaraju(g))
@@ -77,15 +77,27 @@
         @test length(wcc) == 1 && length(wcc[1]) == nv(g)
     end
 
-    function scc_ok(graph)
+    function scc_ok(graph::GenericDiGraph)
         # Check that all SCC really are strongly connected
+
+        # TODO it might be better if we did not have to unwrap the GenericDiGraph
+        # (and we somehow might prevent this in the future) but currently the methods
+        # used in this utility test function do not work with GenericDiGraph yet.
+        graph = graph.g
+
         scc = @inferred(strongly_connected_components(graph))
         scc_as_subgraphs = map(i -> graph[i], scc)
         return all(is_strongly_connected, scc_as_subgraphs)
     end
 
-    function scc_k_ok(graph)
+    function scc_k_ok(graph::GenericDiGraph)
         # Check that all SCC really are strongly connected
+
+        # TODO it might be better if we did not have to unwrap the GenericDiGraph
+        # (and we somehow might prevent this in the future) but currently the methods
+        # used in this utility test function do not work with GenericDiGraph yet.
+        graph = graph.g
+
         scc_k = @inferred(strongly_connected_components_kosaraju(graph))
         scc_k_as_subgraphs = map(i -> graph[i], scc_k)
         return all(is_strongly_connected, scc_k_as_subgraphs)
@@ -97,7 +109,7 @@
     add_edge!(h, 4, 2)
     add_edge!(h, 2, 3)
     add_edge!(h, 1, 3)
-    for g in testdigraphs(h)
+    for g in test_generic_graphs(h)
         @test scc_ok(g)
         @test scc_k_ok(g)
     end
@@ -107,14 +119,14 @@
     add_edge!(h2, 2, 4)
     add_edge!(h2, 4, 3)
     add_edge!(h2, 1, 3)
-    for g in testdigraphs(h2)
+    for g in test_generic_graphs(h2)
         @test scc_ok(g)
         @test scc_k_ok(g)
     end
 
     # Test case for empty graph
     h = SimpleDiGraph(0)
-    for g in testdigraphs(h)
+    for g in test_generic_graphs(h)
         scc = @inferred(strongly_connected_components(g))
         scc_k = @inferred(strongly_connected_components_kosaraju(g))
         @test length(scc) == 0
@@ -123,7 +135,7 @@
 
     # Test case for graph with one vertex
     h = SimpleDiGraph(1)
-    for g in testdigraphs(h)
+    for g in test_generic_graphs(h)
         scc = @inferred(strongly_connected_components(g))
         scc_k = @inferred(strongly_connected_components_kosaraju(g))
         @test length(scc) == 1 && scc[1] == [1]
@@ -139,7 +151,7 @@
     add_edge!(h, 2, 3)
     add_edge!(h, 2, 1)
 
-    for g in testdigraphs(h)
+    for g in test_generic_graphs(h)
         scc = @inferred(strongly_connected_components(g))
         scc_k = @inferred(strongly_connected_components_kosaraju(g))
         @test length(scc) == 2
@@ -159,7 +171,7 @@
     add_edge!(h, 3, 5)
     add_edge!(h, 5, 6)
     add_edge!(h, 6, 4)
-    for g in testdigraphs(h)
+    for g in test_generic_graphs(h)
         scc = @inferred(strongly_connected_components(g))
         scc_k = @inferred(strongly_connected_components_kosaraju(g))
         @test length(scc) == 1 && sort(scc[1]) == [1:6;]
@@ -171,7 +183,7 @@
     add_edge!(h, 2, 3)
     add_edge!(h, 3, 1)
     add_edge!(h, 4, 1)
-    for g in testdigraphs(h)
+    for g in test_generic_graphs(h)
         scc = @inferred(strongly_connected_components(g))
         scc_k = @inferred(strongly_connected_components_kosaraju(g))
         @test length(scc) == 2 && sort(scc[1]) == [1:3;] && sort(scc[2]) == [4]
@@ -200,7 +212,7 @@
     add_edge!(h, 11, 12)
     add_edge!(h, 12, 10)
 
-    for g in testdigraphs(h)
+    for g in test_generic_graphs(h)
         scc = @inferred(strongly_connected_components(g))
         scc_k = @inferred(strongly_connected_components_kosaraju(g))
         @test length(scc) == 4
@@ -224,44 +236,50 @@
     fig1[[3, 4, 9, 10, 11, 13, 18, 19, 22, 24]] = [
         0.5, 0.4, 0.1, 1.0, 1.0, 0.2, 0.3, 0.2, 1.0, 0.3
     ]
-    fig1 = SimpleDiGraph(fig1)
+    fig1 = GenericDiGraph(SimpleDiGraph(fig1))
     scc_fig1 = Vector[[2, 5], [1, 3, 4]]
 
     # figure 2 example
     fig2 = spzeros(5, 5)
     fig2[[3, 10, 11, 13, 14, 17, 18, 19, 22]] .= 1
-    fig2 = SimpleDiGraph(fig2)
+    fig2 = GenericDiGraph(SimpleDiGraph(fig2))
 
     # figure 3 example
     fig3 = spzeros(8, 8)
     fig3[[
         1, 7, 9, 13, 14, 15, 18, 20, 23, 27, 28, 31, 33, 34, 37, 45, 46, 49, 57, 63, 64
     ]] .= 1
-    fig3 = SimpleDiGraph(fig3)
+    fig3 = GenericDiGraph(SimpleDiGraph(fig3))
     scc_fig3 = Vector[[3, 4], [2, 5, 6], [8], [1, 7]]
     fig3_cond = SimpleDiGraph(4)
     add_edge!(fig3_cond, 4, 3)
     add_edge!(fig3_cond, 2, 1)
     add_edge!(fig3_cond, 4, 1)
     add_edge!(fig3_cond, 4, 2)
+    fig3_cond
 
     # construct a n-number edge ring graph (period = n)
     n = 10
     n_ring = cycle_digraph(n)
     n_ring_shortcut = copy(n_ring)
     add_edge!(n_ring_shortcut, 1, 4)
+    n_ring = GenericDiGraph(n_ring)
+    n_ring_shortcut = GenericDiGraph(n_ring_shortcut)
 
     # figure 8 example
     fig8 = spzeros(6, 6)
     fig8[[2, 10, 13, 21, 24, 27, 35]] .= 1
-    fig8 = SimpleDiGraph(fig8)
+    fig8 = GenericDiGraph(SimpleDiGraph(fig8))
 
     @test Set(@inferred(strongly_connected_components(fig1))) == Set(scc_fig1)
     @test Set(@inferred(strongly_connected_components(fig3))) == Set(scc_fig3)
 
     @test @inferred(period(n_ring)) == n
     @test @inferred(period(n_ring_shortcut)) == 2
 
+    # TODO condensation currently returns a SimpleDiGraph, even if the input graph
+    # is a GenericDiGraph, so we compare with a SimpleDiGraph in this test,
+    # but one should think, if the condensation should not also be a GenericDiGraph
     @test @inferred(condensation(fig3)) == fig3_cond
 
     @test @inferred(attracting_components(fig1)) == Vector[[2, 5]]
@@ -270,7 +288,7 @@
     g10dists = ones(10, 10)
     g10dists[1, 2] = 10.0
     g10 = star_graph(10)
-    for g in testgraphs(g10)
+    for g in test_generic_graphs(g10)
         @test @inferred(neighborhood_dists(g, 1, 0)) == [(1, 0)]
         @test length(@inferred(neighborhood(g, 1, 1))) == 10
         @test length(@inferred(neighborhood(g, 1, 1, g10dists))) == 9
@@ -280,8 +298,8 @@
         @test length(@inferred(neighborhood(g, 2, -1))) == 0
     end
     g10 = star_digraph(10)
-    for g in testdigraphs(g10)
-        @test @inferred(neighborhood_dists(g10, 1, 0, dir=:out)) == [(1, 0)]
+    for g in test_generic_graphs(g10)
+        @test @inferred(neighborhood_dists(g, 1, 0, dir=:out)) == [(1, 0)]
         @test length(@inferred(neighborhood(g, 1, 1, dir=:out))) == 10
         @test length(@inferred(neighborhood(g, 1, 1, g10dists, dir=:out))) == 9
         @test length(@inferred(neighborhood(g, 2, 1, dir=:out))) == 1
@@ -301,7 +319,7 @@
     ##@test !@inferred(isgraphical([2]))
 
     # Test simple digraphicality
-    sdg = SimpleDiGraph(10, 90)
+    sdg = GenericDiGraph(SimpleDiGraph(10, 90))
     @test @inferred(isdigraphical(indegree(sdg), outdegree(sdg)))
     @test !@inferred(isdigraphical([1, 1, 1], [1, 1, 0]))
     @test @inferred(isdigraphical(Integer[], Integer[]))
@@ -314,14 +332,14 @@
 
     # 1116
     gc = cycle_graph(4)
-    for g in testgraphs(gc)
+    for g in test_generic_graphs(gc)
         z = @inferred(neighborhood(g, 3, 3))
         @test (z == [3, 2, 4, 1] || z == [3, 4, 2, 1])
     end
 
     gd = SimpleDiGraph([0 1 1 0; 0 0 0 1; 0 0 0 1; 0 0 0 0])
     add_edge!(gd, 1, 4)
-    for g in testdigraphs(gd)
+    for g in test_generic_graphs(gd)
         z = @inferred(neighborhood_dists(g, 1, 4))
         @test (4, 1) ∈ z
         @test (4, 2) ∉ z