Installation¶

This page describes how to install APPFL on a machine independent of operating systems.

Conda environment¶

We highly recommend to create new conda environment and install the required packages for APPFL.

conda create -n APPFL python=3.10
conda activate APPFL

For most users, including data scientists, this simple installation is sufficient for running the package.

pip install pip --upgrade
pip install "appfl[mpi,examples]"

Note

If you do not need to use MPI for simulations, then you can install the package without the mpi option: pip install "appfl[examples]"".

If you want to even minimize the installation of package dependencies, you can use:

pip install appfl

Note

torch may need to be updated manually to support CUDA. Please check GPU support in PyTorch.

Code developers and contributors may want to work on the local repository. To set up the development environment,

git clone --single-branch --branch main https://github.com/APPFL/APPFL.git
cd APPFL
pip install -e ".[dev,mpi,examples]"

Note

If you do not need to use MPI for simulations, then you can install the package without the mpi option: pip install -e ".[dev,examples]".

On Ubuntu, if the installation process failed, you can try:

sudo apt install libopenmpi-dev,libopenmpi-bin,libopenmpi-doc