Installation#

You can install the software with pip

python3 -m pip install cardiac-geometries

Alternatively you can install it directly from Github

python3 -m pip install git+https://github.com/ComputationalPhysiology/cardiac_geometries

However, to actually use the software you need to also install gmsh with python bindings and to convert the mesh into FEniCS format you also need to install FEniCS.

Install using Docker#

Docker is a simple alternative if you don’t want to go through the hazel of installing both FEniCS and Gmsh. We provide a pre-build Docker image with FEniCS and Gmsh installed at orgs/scientificcomputing, which should be compatible with both AMD64 and ARM64 architectures. You can run the image interactively using the command

docker run --rm -v $PWD:/home/shared -w /home/shared -it ghcr.io/scientificcomputing/fenics-gmsh:2023-03-01a

If you want to get an image with cardiac-gemometries installed you can use the Docker image from the GitHub repo

docker run --rm -v $PWD:/home/shared -w /home/shared -it ghcr.io/computationalphysiology/cardiac_geometries:latest

Installing python bindings for gmsh#

If you are using Mac and have installed gmsh with Homebrew, i.e

brew install gmsh

then it will also install python bindings. In my case there is a file called gmsh.py located in /opt/homebrew/lib. In order to make python see this file, you can just update the PYTHONPATH environment variable, e.g

export PYTHONPATH=/opt/homebrew/lib:$PYTHONPATH

Similar hacks might be possible for Linux as well