grass-py3-pdal

Repo which contains a Dockerfile to compile GRASS GIS 7.8 (release branch, grass78) with Python-3 and PDAL support

Installation

docker pull mundialis/grass-py3-pdal

Usage

Note: This docker image contains a link from the current grass7x start script to a generic grass script:

Show version

docker run --rm mundialis/grass-py3-pdal grass --version

This will display the version and copyright statement.

Use convenient alias to start GRASS GIS on command line

In order to simplify the use through docker, consider to define an alias. Note that this variant will remove the docker container immediately after usage in order to not increasingly consume dist space. The alias is useful both for a single command or for executing a script:

# the alias does the following
# - runs docker container and removes it  at exit
# - mounts current directory into docker container under /data/
# - runs GRASS GIS inside with current user/group, in order to avoid writing out data as root user
# - uses /tmp/ in docker container to avoid write permission issues
alias grass_docker="docker run -v $PWD:/data --user $(id --user):$(id --group) -e "HOME=/tmp/" --rm mundialis/grass-py3-pdal grass"

# run dockerized GRASS GIS via alias
grass_docker --version

Show detailed revision number

# using alias from above
grass_docker --tmp-location EPSG:4326 --exec g.version -rge

Run a PDAL test with an included LAZ file

The simple.laz file (source: PDAL test data) is included in this Docker image under /tmp/. It can be scanned by GRASS GIS using the r.in.pdal addon (also included):

# using alias from above
grass_docker -c epsg:4326 --tmp-location --exec r.in.pdal -sg input=/tmp/simple.laz output=dummy
 Starting GRASS GIS...
 Creating new GRASS GIS location <tmploc>...
 Cleaning up temporary files...
 Executing <r.in.pdal -sg input=/tmp/simple.laz output=dummy> ...
 n=853535.43 s=848899.7 w=635619.85 e=638982.55 t=586.38 b=406.59
 Execution of <r.in.pdal -sg input=/tmp/simple.laz output=dummy> finished.
 Cleaning up temporary files...

Executing a script: zonal statistics

Instead of a single GRASS GIS command also a sequence, stored in a text file, can be executed. We first download to geospatial datasets for our zonal statistics analysis:

# download North Carolina, USA, elevation raster map to current directory
wget https://apps.mundialis.de/workshops/osgeo_ireland2017/north_carolina/elev_state_500m.tif
gdalinfo elev_state_500m.tif

# download Raleigh, NC USA, ZIP code vector map (Geopackage) to current directory
wget https://apps.mundialis.de/workshops/osgeo_ireland2017/north_carolina/zipcodes_wake.gpkg
ogrinfo -al -so zipcodes_wake.gpkg

Now we have two spatial datasets downloaded. The spatial reference system of these North Carolina dataset files is EPSG:32119, NAD83 / North Carolina (see also https://epsg.io/32119).

Next we write a shell script to analyse these datasets in order to compute the minimum, average, maximum elevation as well as 1st and 3rd quartiles per ZIP code area using v.what.vect. Store the following code block as a text file named grass_zonal_stats.sh in the current directory:

# note that the current directory is mounted under /data/ in the docker container
#  import datasets into GRASS GIS
r.import input=/data/elev_state_500m.tif output=elev_state_500m
#  the vector import performs minor topological cleaning
v.import input=/data/zipcodes_wake.gpkg output=zipcodes_wake

# check metadata
r.info elev_state_500m
v.info zipcodes_wake

# set computational region to vector map and pixel geometry to raster map
#  see https://grasswiki.osgeo.org/wiki/Computational_region
g.region vector=zipcodes_wake align=elev_state_500m -p

# enrich attribute table of 'zipcodes_wake' with additional columns (results of zonal statistics)
v.rast.stats map=zipcodes_wake raster=elev_state_500m column_prefix=elev method=minimum,maximum,average,first_quartile,third_quartile

# export result outside of GRASS GIS/docker image to current directory
v.out.ogr input=zipcodes_wake output=/data/zipcodes_wake_elev_stats.gpkg

The saved shell script is then run via docker-GRASS GIS as follows:

# note: via alias the current directory is mounted under /data/ in the docker container
grass_docker -c epsg:32119 --tmp-location --exec bash /data/grass_zonal_stats.sh

After successful completion the resulting vector maps zipcodes_wake_elev_stats.gpkg is found in the current directory.

# check metadata of new map
ogrinfo -al -so zipcodes_wake_elev_stats.gpkg

# look at it e.g. in QGIS
qgis zipcodes_wake_elev_stats.gpkg

Using Python

The script below uses the Python API of GRASS GIS to make use of functionality without explicitely starting an interactive GRASS GIS session. Moreover, we use the grass-session Python interface here.

In this example, we import of the "admin0" level vector data from Natural Earth. The Python script downloads and imports the vector data including a topological cleaning on the fly.

# download sample script
wget https://neteler.gitlab.io/grass-gis-analysis/grass_session_vector_import.py
# run it (we explicitely define the GRASS GIS start script name here)
export GRASSBIN=grass ; python3 grass_session_vector_import.py