arpes

arpes copied to clipboard

1. Move example data to a separate data repository and use fetch patterns. We still need the support code here but the new repository is in place and data migrated.
2. Move features behind feature gates checked at runtime. This has enabled moving heavy requirements behind extra requirements flags.
3. Remove really optional things like cvxpy from standard builds.
4. Use conda-pack to uninstall conda after solving the environment during Docker builds. This will need to be unified with whatever the Ernstorfer group is doing.

Hi @Tommaso-Pincelli,

You can find here the branch I mentioned by email.

If tests fail it's because of something which has been fixed on the main branch, so don't worry about that.