Always use astropy's sigma clipping and default to string-based functions

[astrofrog]

This switches the sigma clipping to always use the astropy sigma clipping, and default to using strings for the statistics/functions to use by default as this provides the best performance. This requires bumping the minimum version of astropy to 4.3 which I hope should be ok now that LTS is now 5.x?

I haven't done extensive testing beyond the built-in test suite and my own use case, but for my case the sigma clipping became almost 3x faster.

• For new contributors: Did you add yourself to the "Authors.rst" file?
• Did you add an entry to the "Changes.rst" file?

Fixes #793

I also think #753 could be closed with this merged.

Perhaps #624 would also be resolved by this? (but make sure you change the functions for sigma clipping to be strings)

[astrofrog]

I wasn't sure what major version this should go in in the changelog or what to milestone it as?

[mwcraig]

Sorry for the delay @astrofrog, and thanks for opening -- this is either 2.4 if use_astropy gives a deprecation warning or 3.0 if it deosn't. I'm inclined towards 2.4, and will push a check for use_astropy by Monday, then merge and release.

Might even get to it today....

[codecov]

Codecov Report

Merging #794 (88f1d6e) into main (4ae834d) will increase coverage by 0.04%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #794      +/-   ##
==========================================
+ Coverage   97.43%   97.48%   +0.04%     
==========================================
  Files           9        9              
  Lines        1402     1390      -12     
==========================================
- Hits         1366     1355      -11     
+ Misses         36       35       -1     

Impacted Files	Coverage Δ
ccdproc/combiner.py	94.62% <100.00%> (+0.12%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[mwcraig]

Thanks @astrofrog, merging!

[astrofrog]

@mwcraig - just to check, are you planning a release soon to include this change? (just to know whether to work around this for now or whether to just wait).

[mwcraig]

Would a release by early next week work for you? I should be able to get to it by then....

[astrofrog]

Yes sounds good! Thank you!

[mwcraig]

@astrofrog -- I released 2.4.0, which includes this PR, last night. It is on PyPI and should be on conda-forge shortly, announcement coming later today or tomorrow.

[astrofrog]

Thank you!