Catlearn - SQUID

Catlearn - SQUID#

Welcome to the installation procedure of catlearn on the SQUID supercomputer.

Testing and examples

The following instructions assume you have completed the initialization procedure described in conda-squid.

Environment preparation#

Load the required modules#

We load the intel compilers that may be needed for some of the packages we will install. The BaseCPU/2023 module contains the intel compilers, MPI, and MKL libraries.

module load BaseCPU/2023

Create a conda environment#

conda create -n catlearn python=3.10 pip

# Check if the environment is created
conda env list | grep catlearn

Catlearn Installation#

Download the Catlearn repository#

Note

An official release of catlearn is avaiable, however, it is very old. Many updates and fixes have been made since the last release and is available on the github repository. I created a time-stamped clone of the repository on GitHub. We will use this instead of the official release.

# Make a directory for catlearn
mkdir -p $WORK/apps/catlearn && cd $WORK/apps/catlearn

# Download and extract the unofficial release
wget https://github.com/kimrojas/CatLearn/archive/refs/tags/v0.6.2.tar.gz
tar zxvf v0.6.2.tar.gz

Install catlearn#

# Activate the catlearn environment
conda activate catlearn

# Install catlearn
pip install -v ./CatLearn-0.6.2