Issue with loading the datasets in readthedocs about scikit-matter HOT 4 CLOSED

I am not sure where this comes from, the npz files are included in the installation, checked with

python setup.py bdist_egg | grep npz

Output

copying build/lib/skcosmo/datasets/data/degenerate_CH4_manifold.npz -> build/bdist.linux-x86_64/egg/skcosmo/datasets/data
copying build/lib/skcosmo/datasets/data/csd-1000r.npz -> build/bdist.linux-x86_64/egg/skcosmo/datasets/data

The test should also fail if this weren't true.
When I built the doc locally it also works (I made a new environment to test it).

I am not sure how the github page process exactly works, and where to look. Do you have any idea?

from scikit-matter.

I am not sure how the github page process exactly works, and where to look.

The doc is not build on github pages but readthedocs.org. If you give me your username there I can give you access to their build log.

from scikit-matter.

I think the error in the doc is because the necessary .npz files are in a packed python .egg archive so they cannot be accessed as a directory, I could change the type of installation using e.g. pip so no egg archive is used.

How can I change the building commands of readthedocs? (the ones in https://readthedocs.org/projects/scikit-cosmo/builds/13459127/)
I read about a .readthedocs.yaml in https://docs.readthedocs.io/en/stable/config-file/v2.html but couldn't find it so I guess we are using the default yml file?
One would need to add in the yml file method and pip https://docs.readthedocs.io/en/stable/config-file/v2.html#packages
Something like

version: 2

python:
  version: 3.7
  install:
    - method: pip
      path: .

But I am really not sure about this. One could make a branch and activate it in the readthedocs to play around with this yml file, but I have not the rights to push directly branches to the skcosmo.
EDIT: I have the rights to push directly branches

from scikit-matter.

Addressed by #96

from scikit-matter.