Use PSF astrometry in tweakreg.

[schlafly]

This PR fixes two issues with tweakreg's accuracy. The first is that we at some point unintentionally switched from using the PSF astrometry (x_psf, y_psf) to the segment centroid astrometry (xcentroid, ycentroid) in tweakreg. This leads to much worse astrometric accuracy. The second is that I had left in a fudge factor in the tweakreg regression test, so that we did not notice when this occurred. This PR switches back to using the PSF astrometry and removes the fudge factor.

Regression tests will fail due to changes in the astrometry. Regression tests are here:
https://github.com/spacetelescope/RegressionTests/actions/runs/12713885744

Three failures in steps that use tweakreg. Note that in these cases the median absolute errors decrease and the shifts become closer to zero arcseconds (when the simulated WCS is correct) or to 1" in dec (when the simulated WCS is intentionally offset by 1 arcsecond in ra / dec, in the tweakreg regression test).

We will need to okify these regression differences when merging.

Tasks

• request a review from someone specific, to avoid making the maintainers review every PR
• add a build milestone, i.e. 24Q4_B15 (use the latest build if not sure)
• Does this PR change user-facing code / API? (if not, label with no-changelog-entry-needed)
  • write news fragment(s) in changes/: echo "changed something" > changes/<PR#>.<changetype>.rst (see below for change types)
  • update or add relevant tests
  • update relevant docstrings and / or docs/ page
  • Regression test: https://github.com/spacetelescope/RegressionTests/actions/runs/12713885744 - [ ] Do truth files need to be updated ("okified")?
    • after the reviewer has approved these changes, run okify_regtests to update the truth files
• if a JIRA ticket exists, make sure it is resolved properly

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 1 line in your changes missing coverage. Please review.

Project coverage is 28.60%. Comparing base (f538ba0) to head (f36a1dd).
Report is 2 commits behind head on main.

Files with missing lines	Patch %	Lines
romancal/tweakreg/tweakreg_step.py	0.00%	1 Missing ⚠️

❗ There is a different number of reports uploaded between BASE (f538ba0) and HEAD (f36a1dd). Click for more details.

HEAD has 5 uploads less than BASE

Flag	BASE (f538ba0)	HEAD (f36a1dd)
	6	1

Additional details and impacted files

@@             Coverage Diff             @@
##             main    #1578       +/-   ##
===========================================
- Coverage   77.94%   28.60%   -49.35%     
===========================================
  Files         115      112        -3     
  Lines        7622     7576       -46     
===========================================
- Hits         5941     2167     -3774     
- Misses       1681     5409     +3728     

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

[ddavis-stsci]

Looks reasonable to me

[mairanteodoro]

Looks good to me.
I haven't checked it yet, but don't we have to update the docs too?

[schlafly]

Thanks, yes, I should update this line of the docs.

romancal/docs/roman/tweakreg/README.rst

Line 23 in f538ba0

either ``'x'`` and ``'y'`` or ``'xcentroid'`` and ``'ycentroid'`` columns which

I'll do that this evening and merge.