JP-3483: fix atoca failures due to sparse matrix solver

[emolter]

Resolves JP-3483

Closes #8110

This PR addresses a bug where the ATOCA algorithm's matrix solver would sometimes fail on good data, likely because the approximations in spsolve were inappropriate. The fix allows the code to try using a least-squares solver instead when these failures occur.

Checklist for maintainers

• 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
• Make sure the JIRA ticket is resolved properly

Jenkins run here. I don't think the failures have anything to do with this change.

[codecov]

Codecov Report

Attention: 7 lines in your changes are missing coverage. Please review.

Comparison is base (7386db3) 75.15% compared to head (354bfd0) 75.14%.
Report is 1 commits behind head on master.

Files	Patch %	Lines
jwst/extract_1d/soss_extract/atoca.py	22.22%	7 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #8273      +/-   ##
==========================================
- Coverage   75.15%   75.14%   -0.02%     
==========================================
  Files         470      470              
  Lines       38604    38611       +7     
==========================================
  Hits        29013    29013              
- Misses       9591     9598       +7     

Flag	Coverage Δ		*Carryforward flag
nightly	77.38% <ø> (-0.01%)	⬇️	Carriedforward from 7386db3

*This pull request uses carry forward flags. Click here to find out more.

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

[tapastro]

One python version issue, I'm testing the regressions locally on py3.9 but I think my suggestion should work.

[hbushouse]

Looks good to me, other than the issue with having it work with py 3.9.

[hbushouse]

@emolter Need to resolve conflicts so that the CI tests can run. At that point another regtest run would be good.

[emolter]

New Jenkins regression tests are here

[hbushouse]

Looks good!