Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 92.71%. Comparing base (6b51eac) to head (b329d19).
⚠️ Report is 4 commits behind head on develop.

@@             Coverage Diff             @@
##           develop    #5158      +/-   ##
===========================================
- Coverage    92.72%   92.71%   -0.02%     
===========================================
  Files          180      180              
  Lines        22458    22457       -1     
  Branches      3186     3186              
===========================================
- Hits         20824    20820       -4     
- Misses        1177     1179       +2     
- Partials       457      458       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

This PR is ready for review.

All tests pass locally, and the only CI failure was a cancelled job due to duplicate workflow runs by GitHub.

@orbeckst @RMeli @marinegor — Please let me know if anything else is needed. Thanks!

Hi @marinegor @RMeli @orbeckst
I finished a full repo-wide scan for deprecated NumPy testing utilities and found ~220+ affected files — not only in the testsuite but also inside the actual package/MDAnalysis/... modules (analysis, core, coordinates, lib, topology, transformations, etc.).

Plan

I’m thinking of submitting 10–15 small PRs, each focused on one module area to keep reviews manageable and avoid large changes.

Quick Question

Some deprecated patterns appear in the package code too.
Should I update the package modules as well, or limit the modernization to the test suite?

I’ll follow whichever direction you prefer.

Thanks!

I finished a full repo-wide scan for deprecated NumPy testing utilities and found ~220+ affected files — not only in the testsuite but also inside the actual package/MDAnalysis/... modules (analysis, core, coordinates, lib, topology, transformations, etc.).

How did you actually do that? I found only 70 files:

❯ rg 'assert_array_almost_equal|assert_almost_equal' -g '*.py' -c  | wc -l
70

Some deprecated patterns appear in the package code too.

Which exactly?

Should I update the package modules as well, or limit the modernization to the test suite?

The issue you're working in only regarding testing, so within this issue I'd limit your work. If you decide to also replace it in other places, please create a separate issue for that.

@Pranay096 may I suggest possibly thinking about targetting another issue. Issue #3743 should be considered as a large "starter" isssue that several folks over time would contribute to, not something a single individual should be doing in one go (at least not until we actually hit a deprecation deadline).

I would urge you to consider looking at some of the more complicated issues on our tracker.

Thank you @marinegor and @IAlibay for the explanation and advice; it is greatly appreciated.

As for the number of files, my initial scan contained a few other deprecated patterns scattered around the repository so the number was bigger. Your rg command clarified that only ~70 files are relevant given the specific assert_almost_equal update and I'll strictly keep my contributions for this issue limited to the test suite. If I catch similar patterns in the package code, I'll open a separate issue without intermixing the changes.

I also appreciate the insight into Issue #3743, and I now realize that it is supposed to be a long-term issue of collaboration, not something for one contributor to handle; I will definitely take your suggestion and start investigating some of the harder issues on the tracker where I can make more substantial contributions.

I truly value both of your insights, as it helps me stay aligned with the expectations of the project and make my contributions as helpful as possible.