Workflow for building HUC-based meshes in python.

Installation and Setup

All code is in python3. Recommended is to download anaconda3, as this makes it fairly easy to get all the required packages.

Note that this package expects you to place the top-level directory in your PYTHONPATH:

export PYTHONPATH=`pwd`:${PYTHONPATH}

Required Packages

Standard packages needed (should be included in all distributions?) include argparse and subprocess, and for testing, pytest and dist_utils. Standard math packages include numpy, matplotlib, and scipy.

GIS work uses packages: fiona, rasterio, and shapely. The packages for fiona and rasterio from conda-forge seem to work better?

conda create -n ats_meshing -c conda-forge -c defaults python=3 ipython numpy matplotlib scipy meshpy basemap basemap-data-hires fiona rasterio shapely ipykernel requests

Mesh generation uses Triangle, which can be wrapped for python using meshpy. This one is a little more difficult, as there seem to be no good distributions of it?

Note this depends upon boost python, which seems to come by default with anaconda?

For now, build from source:

 export ANACONDA_DIR=/path/to/your/anaconda
 export BOOST_ROOT=${ANACONDA_DIR}
 git clone https://github.com/ecoon/meshpy.git
 cd meshpy
 ./configure.py --python-exe=${ANACONDA_DIR}/bin/python --boost-inc-dir=${ANACONDA_DIR}/include  --boost-lib-dir=${ANACONDA_DIR}/lib --boost-python-libname=boost_python3  --disable-static --enable-shared --cxxflags=-stdlib=libc++ --ldflags=-stdlib=libc++
 make
 python setup.py install

Note to look closely at the result of the configure command -- it errors frequently and has trouble finding Boost?

Check your python installation:

 $> ipython
 import numpy as np
 import matplotlib as plt
 import scipy
 import rasterio
 import fiona
 import shapely
 import meshpy.triangle

Alternate Setup

Alternative, use a docker container.

On a mac: docker build -t ats-meshing -f ./Dockerfile ./

 brew install socat
 nohup socat TCP-LISTEN:6000,reuseaddr,fork UNIX-CLIENT:\"$DISPLAY\"
 IP=$(ifconfig en0 | grep inet | awk '$1=="inet" {print $2}')
 docker run -it --name container1 --net=host --env DISPLAY=${IP}:0  -v ~/codes/ats/ats-meshing/:/ats ats-meshing
 cd ats-meshing

A first example

A good way to get started is to simply run go:

python3 bin/mesh_hucs.py -c 060102080102

This downloads a HUC file (not small, so takes a minute) then extracts, smooths, and triangulates some HUCs in it.

A set of examples

Basic triangulation of an existing HUC 12:

python3 bin/mesh_hucs.py -c 060102080102

Triangulate a HUC 10, ensuring that HUC12 edges are included:

python3 bin/mesh_hucs.py -c 0601020801

Triangulate with refinement, grading the max area to higher res near the river:

python3 bin/mesh_hucs.py --refine-distance 0 1000 1000 10000 -c 060102080102

Triangulate a given user-provided shapefile, for instance a subwatershed from the Coweeta basin:

python3 bin/mesh_shape.py --refine-distance 10 100 1000 1000 --hint=06 --center --shape-index=0 data/hydrologic_units/others/Coweeta/coweeta_subwatersheds.shp