Installation

Installation#

Using pip#

Warning

FRIDOM is in an early development stage and the latest version on PyPI might not be the most recent one. It is recommended to install FRIDOM from the source code repository (see below).

FRIDOM can be installed using pip:

pip install fridom["<backend>"]

The backend can be one of the following:

  • jax-cuda: JAX backend for GPU acceleration with CUDA support. (recommended)

  • jax: JAX backend without GPU acceleration.

  • cupy: CuPy backend with GPU acceleration (fridom is not optimized for CuPy).

For example, to install FRIDOM with the JAX backend and CUDA support, run:

pip install fridom["jax-cuda"]

If no backend is specified, only numpy will be available as a backend.

Building from source#

To install FRIDOM from the source code repository, clone the repository in your desired directory and install the package using pip:

git clone https://github.com/Gordi42/FRIDOM
cd FRIDOM
pip install -e '.[jax-cuda]'

This will install FRIDOM with the JAX backend and CUDA support. To install FRIDOM with a different backend, replace jax-cuda with the desired backend (see above).

Note

The -e flag installs FRIDOM in editable mode, which means that changes to the source code will be reflected in the installed package.

Note

It is recommended to install FRIDOM in a virtual environment to avoid conflicts with other packages. This can for example be done with conda by running the following code before installing FRIDOM:

conda create -y --name fridom python=3.12
conda activate fridom

Optional dependencies#

  • xarray: To convert data to xarray datasets for easier plotting. An installation guide can be found here.

  • imageio: To create animations. It can be installed with pip install "imageio[ffmpeg]".

  • mpi4py: To run simulations in parallel using MPI (currently only with the numpy backend). An installation guide can be found at this link.

Installation on special systems#

Levante (DKRZ)#

If you plan to run FRIDOM on levante at DKRZ with GPU acceleration, make sure to do the installation on a gpu node. A gpu node can be requested with the following command:

salloc -p gpu --gpus=1 --account=projectname

where projectname is the name of your project. After you have been assigned a gpu node, you can install FRIDOM using the above instructions.

Contents