j042 / gustav-alpha

alpha gustav

gustav-alpha

Complete snapshots of gustav project developments. Last update: 2021.08.31

Setup guide using conda

Info:

conda is recommended, but it is possible to setup without conda - use pip and fufill SplineLib's requirements.
If you'd like to give conda a try, you could visit: https://docs.conda.io/en/latest/miniconda.html

Installation on RWTH Compute Cluster

mkdir miniconda3
cd miniconda3
wget https://repo.anaconda.com/miniconda/Miniconda3-py39_4.10.3-Linux-x86_64.sh
bash Miniconda3-py39_4.10.3-Linux-x86_64.sh -u

The script will prompt you for the directory, where miniconda is supposed to be installed and will per default suggest the current directory. The -u option is required for installing into an already existing directory.

Setup (local and on RWTH Cluster)

Clone and get ready.

git clone [e-mail]:j042/gustav.git
cd gustav
export GUS=$PWD

Create a new environment and install packages
pip:
Note: pip tends to work better than conda in our case

conda create -n gus python=3.9         # Tested with 3.7, 3.8 and 3.9
conda activate gus
pip install cython numpy scipy vedo matplotlib "meshio[all]" optimesh

conda:
Note: still needs some pip. vedo tends to fail, if installed using conda

conda create -n gus python=3.9
conda activate gus
conda install -c anaconda cython
conda install -c anaconda numpy
conda install -c anaconda scipy
conda install -c conda-forge vedo  
conda install -c conda-forge matplotlib
conda install -c conda-forge meshio
pip install optimesh

Totally OPTIONAL, there's only one function that returns trimesh for your convenience:

conda install -c conda-forge scikit-image shapely rtree pyembree
pip install trimesh[all]

SplineLib Dependency: SplineLib requires cmake-3.19 or higher. If your system does not fulfill the requirements, you could install requirements using conda for your gus environment:

conda install -c conda-forge cmake      # installs cmake-3.20 (last checked: 2021.06.25)

Install third party modules:
Note: If you want to work on the library, replace python3 setup.py install with python3 setup.py develop to make your life easier.

# the following two commands are only required on the RWTH Cluster
###########################
module switch intel gcc
module load cmake
###########################

cd third_party/triangle
python3 setup.py install

cd $GUS
cd third_party/tetgen
python3 setup.py install

cd $GUS
cd third_party/splinelibpy
python3 setup.py install

cd $GUS
python3 setup.py develop

Setup on RWTH Compute Cluster without conda

Load the required modules

module switch intel gcc/10
module load python/3.8.7 cmake

Extend the PYTHONPATH variable with the directory where the additional packages will be installed

export PYTHONPATH=$PYTHONPATH:$HOME/.local/lib/python3.8/site-packages

Install the required python dependencies

pip install cython numpy scipy vedo matplotlib "meshio[all]" optimesh

Install all required third party software

cd third_party/triangle
python3 setup.py install --prefix $HOME/.local/

cd ../tetgen
python3 setup.py install --prefix $HOME/.local/

cd ../splinelibpy
python3 setup.py install --prefix $HOME/.local/

Install the software

cd $GUS
python3 setup.py develop --prefix $HOME/.local/