JP-4313: Use explicit read_times in JWST model in ramp fit

[melanieclarke]

Toward JP-4313

For coordination with spacetelescope/jwst#10534:
Use explicit read times for JWST data if available. This is intended to support new multistripe modes with repeated in-frame reads of the same detector area.

Tasks

• update or add relevant tests
• update relevant docstrings and / or docs/ page
• Does this PR change any API used downstream? (if not, label with no-changelog-entry-needed)
  • write news fragment(s) in changes/: echo "changed something" > changes/<PR#>.<changetype>.rst (see changelog readme for instructions)
  • if your change breaks existing functionality, also add a changes/<PR#>.breaking.rst news fragment
• run regression tests with this branch installed ("git+https://github.com/<fork>/stcal@<branch>")
  • jwst regression test
  • romancal regression test

[codecov]

Codecov Report

❌ Patch coverage is 50.00000% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 91.79%. Comparing base (a04ec58) to head (b2da8cd).
⚠️ Report is 2 commits behind head on main.

Files with missing lines	Patch %	Lines
src/stcal/ramp_fitting/ramp_fit.py	50.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #546      +/-   ##
==========================================
- Coverage   91.80%   91.79%   -0.01%     
==========================================
  Files          63       63              
  Lines        8855     8859       +4     
==========================================
+ Hits         8129     8132       +3     
- Misses        726      727       +1     

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[melanieclarke]

Regtests with jwst on main:
https://github.com/spacetelescope/RegressionTests/actions/runs/26599529754

Regtests with romancal on main:
https://github.com/spacetelescope/RegressionTests/actions/runs/26599567262

Regtests with the jwst branch + stdatamodels update:
https://github.com/spacetelescope/RegressionTests/actions/runs/26588854461

All passing.

[schlafly]

I probably need to review better what romancal does, but I'm a little confused. The old version took read_pattern from the input model and converted it to a list of lists and put it at read_pattern on ramp_data. The new version no longer looks for read pattern at all and instead looks for read_times, to my eyes basically redefining read_pattern -> read_times for models but not for the ramp data. Why can't Webb use the old name? I would have naively expected romancal to break since it uses read_pattern as the name
https://github.com/spacetelescope/romancal/blob/main/romancal/ramp_fitting/ramp_fit_step.py#L259
and haven't yet figured out why it still works with the new name.

[melanieclarke]

I probably need to review better what romancal does, but I'm a little confused. The old version took read_pattern from the input model and converted it to a list of lists and put it at read_pattern on ramp_data. The new version no longer looks for read pattern at all and instead looks for read_times, to my eyes basically redefining read_pattern -> read_times for models but not for the ramp data. Why can't Webb use the old name? I would have naively expected romancal to break since it uses read_pattern as the name https://github.com/spacetelescope/romancal/blob/main/romancal/ramp_fitting/ramp_fit_step.py#L259 and haven't yet figured out why it still works with the new name.

Romancal doesn't use create_ramp_fit_class, it creates the ramp fit class directly. The ramp data class takes a list of lists of read times as the "read_pattern", so even if romancal did use create_ramp_fit_class, it wouldn't be correct to directly read meta.exposure.read_pattern into the ramp class read_pattern attribute.

I'd like for Webb to use meta.exposure.read_times instead of meta.exposure.read_pattern in the datamodel because we need to directly store a list of lists of read times, not a frame averaging pattern.

At some point, I think it would make sense to move create_ramp_fit_class from stcal to jwst to make it clearer that we each need to set up our own ramp fit classes, but I didn't want to do it here.