HOW TO SETUP WARPX JUPYTER DEV ENVIRONMENT

WARPX

All physics simulation code is pulled from the WarpX repository.

JUPYTER ACCOUNT + SSH SETUP

Before setting up the development environment, create a jupyter account and connect to perlmutter via nersc.

WARPX INSTALLATION

First, clone the WarpX repository -

git clone git@github.com:ECP-WarpX/WarpX.git

OR

git clone https://github.com/ECP-WarpX/WarpX.git

Build and compile using cmake -

cmake .

cmake -S . -B build

cmake --build build -j 4

Setup python on perlmutter -

module load python/3.9

python3 -m pip wheel -v .

To set up as a python module -

export PYTHONUSERBASE=/YOUR_PATH/WarpXPython

python setup.py install --user

CONDA ENVIRONMENT SETUP

######################### #########################

Start here if you only need python pywarpx on jupyter

######################### #########################

Load python version 3.9 as a module -

module load python/3.9

Create a warpx conda environment -

conda create -n warpx -c conda-forge warpx

Activate the environment -

conda activate warpx

Configure your environment according to WarpX documentation -

conda config --set solver libmamba

mamba install --yes -c conda-forge python ipykernel ipympl matplotlib numpy pandas yt openpmd-viewer openpmd-api h5py fast-histogram dask dask-jobqueue pyarrow

python -m ipykernel install --user --name warpx --display-name warpx

echo -e '#!/bin/bash\nmodule load python\nsource activate warpx\nexec "$@"' > $HOME/.local/share/jupyter/kernels/warpx/kernel-helper.sh

chmod a+rx $HOME/.local/share/jupyter/kernels/warpx/kernel-helper.sh

KERNEL_STR=$(jq '.argv |= ["{resource_dir}/kernel-helper.sh"] + .' $HOME/.local/share/jupyter/kernels/warpx/kernel.json | jq '.argv[1] = "python"')

echo ${KERNEL_STR} | jq > $HOME/.local/share/jupyter/kernels/warpx/kernel.json

conda install --channel conda-forge yt

conda update -y -n base conda