Edit environment.yml + Add unit-tests for particle_metadata.ipynb

[ninamiolane]

This PR:

• cleans environment.yml by removing unnecessary github import of simSPI
• adds test_notebooks.py file to test jupyter notebooks
• tests particle_metadata.ipynb

Note that download_and_upload_with_osf.ipynb is not tested as the jupyter notebook is currently running for too long.

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[ninamiolane]

There is a problem in this PR, as the unit-tests of ioSPI are taking over an hour, due to the installation of conda dependencies. (see screenshot below)

We should investigate the source of the problem, and add a fix so that the unit-tests do not take as long, because unit-tests that take over an hour will prevent many contributors from adding code to the library, as it will be very frustrating.

The conda dependencies are in the environment.yml file at the root of the repository. The long unit-tests have first been observed in PR #75 , but not in other recent PRs, thus it should come from additions made in this PR: the install of jupyter or the install of osfclient maybe?

[ninamiolane]

Installing jupyter (via pip) in geomstats' workflows does not take as long: https://github.com/geomstats/geomstats/blob/a7bb7e80802727312221e3d18f86bbb2c40a4825/.github/workflows/test.yml#L47

jupyter is in doc requirements: https://github.com/geomstats/geomstats/blob/a7bb7e80802727312221e3d18f86bbb2c40a4825/setup.cfg#L56

[artajmir3]

FYI, I've encountered the same issue in PR #83 exactly after the commit that added jupyter to enviroment.yml, and there is no osfclient in that PR. Hence, I think the issue is probably with the installation of Jupiter.

[ninamiolane]

@artajmir3 thanks for the insight! Yes, we are not sure why this is taking so long. Let us know if you find out how to solve this jupyter issue? We need to solve it before we can merge this PR and your PR :)

[codecov[bot]]

Codecov Report

Merging #75 (9513d5e) into master (0ee2cb5) will not change coverage. The diff coverage is n/a.

@@           Coverage Diff           @@
##           master      #75   +/-   ##
=======================================
  Coverage   98.66%   98.66%           
=======================================
  Files           5        5           
  Lines         223      223           
=======================================
  Hits          220      220           
  Misses          3        3           

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 0ee2cb5...9513d5e. Read the comment docs.

[bongjinkoo]

The same path problem occurred when installing jupyter with pip. Also, subprocess.check_call() didn't work. So I changed it to subprocess.run() and it works now. I think the DeppSource error can be ignored.