Improve aca coord transforms

[taldcroft]

Description

This is a major update to the ACA pixel row,col <=> angle yag,zag transformations.

This includes:

• For pixels_to_yagzag use the 2020 calibration coefficents by default.
• For yagzag_to_pixels use Newton's method to accurately invert pixels_to_yagzag using a good starting point.
• Correctly account for the ACA temperature for both ground (CXC) and flight transformations.

This PR uses numba-accelerated routines to improve performance and accuracy.

TODO:

• Documentation updates (docs build is now clean and docstrings look good)
• Document performance improvements here in the validate notebooks.
• Integration testing, e.g. impact on proseco / sparkles or other packages.
  • Acceptable impact as seen in Update test regress vals for chandra_aca coord improvements sparkles#228 and Use chandra_aca 2020 transforms with T_aca = T_ccd + 41 C proseco#414.

Fixes #152

Interface impacts

• Default t_aca=35.0 (from previous 20.0) is likely going to cause regressions (albeit benign) in places. This will change some answers but for the better in the modern era.
• Default allow_bad change to True could result in not raising an exception. Likely not impacting.

Testing

Unit tests

• Mac

(ska3) ➜  chandra_aca git:(improve-aca-coord-transforms) git rev-parse --short HEAD
06d34eb
(ska3) ➜  chandra_aca git:(improve-aca-coord-transforms) pytest
====================================== test session starts =======================================
platform darwin -- Python 3.13.11, pytest-9.0.2, pluggy-1.6.0
rootdir: /Users/aldcroft/git
configfile: pytest.ini
plugins: anyio-4.12.1, timeout-2.4.0
collected 268 items                                                                              

chandra_aca/tests/test_aca_image.py .....................                                  [  7%]
chandra_aca/tests/test_attitude.py ....................................................... [ 28%]
......                                                                                     [ 30%]
chandra_aca/tests/test_dark_model.py ............                                          [ 35%]
chandra_aca/tests/test_dark_subtract.py .........                                          [ 38%]
chandra_aca/tests/test_drift.py ..............................................             [ 55%]
chandra_aca/tests/test_manvr_mon_images.py .......                                         [ 58%]
chandra_aca/tests/test_maude_decom.py .........................                            [ 67%]
chandra_aca/tests/test_planets.py .................                                        [ 73%]
chandra_aca/tests/test_psf.py ...                                                          [ 75%]
chandra_aca/tests/test_residuals.py .....                                                  [ 76%]
chandra_aca/tests/test_star_probs.py .................................                     [ 89%]
chandra_aca/tests/test_transform.py .............................                          [100%]

====================================== 268 passed in 40.36s ======================================

Independent check of unit tests by Jean

• Linux on ska3/aca

jeanconn-kady> pytest
======================================================================= test session starts =======================================================================
platform linux -- Python 3.13.11, pytest-9.0.2, pluggy-1.6.0
rootdir: /proj/sot/ska/jeanproj/git
configfile: pytest.ini
plugins: anyio-4.12.1, timeout-2.4.0
collected 280 items                                                                                                                                               

chandra_aca/tests/test_aca_image.py .....................                                                                                                   [  7%]
chandra_aca/tests/test_all.py ...............s.........................                                                                                     [ 22%]
chandra_aca/tests/test_attitude.py .............................................................                                                            [ 43%]
chandra_aca/tests/test_dark_model.py ............                                                                                                           [ 48%]
chandra_aca/tests/test_dark_subtract.py .........                                                                                                           [ 51%]
chandra_aca/tests/test_drift.py ..............................................                                                                              [ 67%]
chandra_aca/tests/test_manvr_mon_images.py .......                                                                                                          [ 70%]
chandra_aca/tests/test_maude_decom.py .........................                                                                                             [ 79%]
chandra_aca/tests/test_planets.py .................                                                                                                         [ 85%]
chandra_aca/tests/test_psf.py ...                                                                                                                           [ 86%]
chandra_aca/tests/test_residuals.py .....                                                                                                                   [ 88%]
chandra_aca/tests/test_star_probs.py .................................                                                                                      [100%]

============================================================ 279 passed, 1 skipped in 68.00s (0:01:07) ============================================================
jeanconn-kady> git rev-parse HEAD
e2599b48b01a2e494189c177418aa0851519fb2f

Functional tests

No functional testing.

[taldcroft]

@jeanconn - I have completed all TODOs and planned work for this PR so it is ready for final review.

[jeanconn]

It looked to me like the proseco tests are failing again which I didn't completely understand in the context of the chandra_aca changes since they last passed.

[taldcroft]

@jeanconn - I'm looking into proseco today, but those regressions are independent of whether this PR is correct by itself. Of course you can review / approve with a lien on having a good proseco patch.

[taldcroft]

@jeanconn - are you good with this now that proseco is passing and you approved that update PR?

[jeanconn]

What are your thoughts on sot/sparkles#228 ?

[jeanconn]

Actually, nevermind, I've got a new failure in sparkles so I'll go fix that.

[jeanconn]

OK, sparkles tests re-passing so that could use review.

[taldcroft]

I just approved the sparkles PR.