Installation

cesm-hawc requires Python 3.11 or later.

Recommended: conda environment¶

To run the simulator (cesm-hawc run, the [sim] tier), install cesm-hawc into the conda environment defined by the project’s environment.yml. It installs sasktran2 and the scientific stack from conda-forge, and hawcsimulator and aliprocessing from PyPI, in versions that work together.

curl -O https://raw.githubusercontent.com/torimcd/cesm-hawc/main/environment.yml
conda env create -f environment.yml
conda activate hawc_env
pip install "cesm-hawc[sim]"

mamba and micromamba accept the same commands in place of conda.

Base tier only¶

Extracting WACCM columns and saving simulator-ready inputs (cesm-hawc save-inputs) needs no simulator, and installs with pip in any Python 3.11+ environment:

pip install cesm-hawc

From source¶

For development, clone the repository and install it in editable mode into the same conda environment:

git clone https://github.com/torimcd/cesm-hawc
cd cesm-hawc
conda env create -f environment.yml
conda activate hawc_env
pip install -e ".[sim,dev]"

See Development for running the tests and building these docs.

Checking the install¶

cesm-hawc --help
python -c "import cesm_hawc; print(cesm_hawc.__version__)"

For the [sim] tier, also check that the simulator imports:

python -c "import sasktran2, hawcsimulator; print('sim tier OK')"

On first use¶

The [sim] tier builds several Mie scattering databases the first time it runs, and caches them on disk. Expect the first simulation to take noticeably longer than later ones. The CLI builds these caches once in the main process before starting worker processes.

HPC clusters¶

See Running on HPC for the Alliance Canada setup script and the SLURM job script.