Specify number of cores for jump and ramp_fit

[mwregan2]

This PR addresses the problem that for multiprocessing the number of cores to use can only be a fraction of the total number of cores. This is fine for small machines but on large clusters even 1/4 can be larger than someone is allowed to use.
The PR adds the flexibility to specify explicitly the number of cores the user wants to use. Changes were made for both Jump and Ramp fitting and new methods were written with unit tests.
No code changes are needed in the JWST repo only in the documentation.

Checklist

• added entry in CHANGES.rst (either in Bug Fixes or Changes to API)
• updated relevant tests
• updated relevant documentation
• updated relevant milestone(s)
• added relevant label(s)

[codecov]

Codecov Report

Patch coverage: 98.59% and project coverage change: +1.24% 🎉

Comparison is base (0c7cb32) 74.11% compared to head (72a7e83) 75.36%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #183      +/-   ##
==========================================
+ Coverage   74.11%   75.36%   +1.24%     
==========================================
  Files          33       33              
  Lines        6100     6039      -61     
==========================================
+ Hits         4521     4551      +30     
+ Misses       1579     1488      -91     

Files Changed	Coverage Δ
tests/test_twopoint_difference.py	100.00% <ø> (+8.56%)	⬆️
src/stcal/jump/jump.py	66.33% <93.33%> (+3.40%)	⬆️
src/stcal/ramp_fitting/gls_fit.py	64.39% <100.00%> (+0.16%)	⬆️
src/stcal/ramp_fitting/ols_fit.py	80.17% <100.00%> (+0.03%)	⬆️
src/stcal/ramp_fitting/utils.py	91.03% <100.00%> (+0.41%)	⬆️
tests/test_jump.py	94.60% <100.00%> (+6.14%)	⬆️
tests/test_ramp_fitting.py	89.57% <100.00%> (+0.23%)	⬆️

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

[kmacdonald-stsci]

I'd like to request the ramp fit test be moved to a different location per my specific comment above. Other than that, everything looks good.

[hbushouse]

The CI failures are caused by unit tests in the jwst/ramp_fitting package trying to load the now obsolete function compute_slices (it's been renamed in this PR). Does the jwst PR contain updates to fix this?

[hbushouse]

OK, this is looking better. Just need to address the 2 errors coming from the code style CI test (removed unused imports from test modules).

[hbushouse]

Still need to have the style check errors fixed before we can merge:

tests/test_jump.py:3:24: F401 [] astropy.io.fits imported but unused
tests/test_twopoint_difference.py:3:24: F401 [] astropy.io.fits imported but unused

[hbushouse]

I've fixed the code style errors.

[hbushouse]

Finally all looks good.