A Bragg Pulse Simulator for He3-He4

Dev env setup

I use miniconda to manage the python packages for this project, please have it installed in order to follow this setup guide. I am using python 3.11 (as of Oct 2024) since this is the only version that somewhat satisfies all dependencies. At some point, I would like to move to python 3.12 to utilise the new typing features. I'm also using VSCode for multi view, extensions, copilot, etc. For an identical dev experience, consider using the VSCode profile py311_h34sim.code-profile, and workspace py311_h34sim.code-workspace

Create the conda environment

Please run one of the following command.

conda env create --file env-hist-macos-arm64.yml --prefix ./envs/py311_he34sim
conda env create --file env-hist-intel.yml --prefix ./envs/py311_he34sim
conda env create --file env-hist-noarch.yml --prefix ./envs/py311_he34sim

By default, I want to create the environment in the project directory ./ so it doesn't clutter my home directory. But if you want to store packages for this environment in the conda default directory, use --name py311_he34sim instead of --prefix.

macos-armos uses the Apple Accelerate framework, intel uses MKL for hardware accelerated linear algebra computations.

Don't forget to activate the newly created environment.

VSCode's jupyter page should auto detect the new environment.

conda activate ./envs/py311_he34sim 

and deactivate once done

conda deactivate 

To see all your current conda environments:

conda env list 

you should see * with path to this repo

Update the conda environment configuration file

conda env export --from-history > env-hist.yml

and manually remove the name and prefix fields (? maybe it's fine without removing them? since I explicitly set the environment's name already), then I like to manually update the macos-arm64 and intel configurations.

To export a list of packages actually installed, run:

conda env export > env.yml

Remove/Delete/Uninstall this conda environment

conda env remove --prefix ./envs/py311_he34sim

Git and jupyter

(Why do I have *.ipynb in .gitignore.?)

Because jupyter's file is rather large with too much not-git-useful data, I am using jupytext to convert .ipynb files to .py for better git readability.

jupytext --to py twoParticle_v.a.6.ipynb

jupytext --to ipynb twoParticle_v.a.6.py

Because I'm using VSCode so I have to do these manual conversions, if you use jupyter notebook or lab, you can setup a pairing between .ipynb and .py files, and then can convert automatically, see doc here.

Random command logs needed to get shit working

which ffmpeg
# and then update os.environ["IMAGEIO_FFMPEG_EXE"]

conda install matplotlib --update-deps --force-reinstall

conda install opencv --update-deps --force-reinstall

conda install numpy --update-deps --force-reinstall

conda install 'numpy>=2.0' --update-deps --force-reinstall