JP-3664: Testing changes for numpy 2.0 compatibility

[tapastro]

Resolves JP-3664

Closes #8580

This draft exists to remove jwst's upper pin on numpy, as well as provides a spot to test changes to coron3 unit tests that appear to be incompatible with np2.0.

Checklist for PR authors (skip items if you don't have permissions or they are not applicable)

• added entry in CHANGES.rst within the relevant release section
• updated or added relevant tests
• updated relevant documentation
• added relevant milestone
• added relevant label(s)
• ran regression tests, post a link to the Jenkins job below.
  How to run regression tests on a PR
• All comments are resolved
• Make sure the JIRA ticket is resolved properly

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 66.99%. Comparing base (03ca10f) to head (b7358ab).
Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #8718      +/-   ##
==========================================
- Coverage   76.92%   66.99%   -9.94%     
==========================================
  Files         498      377     -121     
  Lines       45810    38156    -7654     
==========================================
- Hits        35240    25561    -9679     
- Misses      10570    12595    +2025     

Flag	Coverage Δ
nightly	?

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

[melanieclarke]

#8812 fixed most of the coron unit tests with numpy 2, but there's still one check that's barely out of spec on the Linux tests.

I think changing these lines to atol=1e-4, rtol=1e-5 should work:

jwst/jwst/coron/tests/test_coron.py

Lines 144 to 145 in 217d660

	npt.assert_allclose(aligned, truth_aligned, atol=1e-5)
	npt.assert_allclose(shifts, truth_shifts, atol=1e-5)

[zacharyburnett]

started a regtest: https://github.com/spacetelescope/RegressionTests/actions/runs/11555475241/job/32161141720

[tapastro]

Reviving this PR, starting a fresh set of tests: https://github.com/spacetelescope/RegressionTests/actions/runs/12717188152

From comments on the JP ticket by David, I think we're ready to test and review this PR.

Change log fixed

[melanieclarke]

Regression test results look mostly as expected, but there is a unit test for extract_1d (marked 'slow', so it doesn't run in CI) that failed: jwst/extract_1d/tests/test_extract_1d_step.py::test_extract_niriss_soss_96. I'm looking into it, but I haven't been able to reproduce it locally yet.

Update: looks like an optimization instability for the pathological testing data. I sent in a PR to fix the test with slightly less improbable input data: #9069