polarization covariances (all)

[xzackli]

This PR will contribute some codes to compute a subset of spin-2 covariances. BB-related covariances will likely still require some aid from simulations. In this PR I'll focus on cleaning up the TTTT code enough to write legible EEEE and TETE.

[xzackli]

Tests appear to be failing because of a pixell issue upstream.

[xgarrido]

This is due to pillow version 10.0 (required by pixell) XingangPan/DragGAN#152 I don't really know how to proceed since it's mainly a pixell issue (we can downgrade pillow within pspy unit tests)

[xgarrido]

Actually pixell never faces this relatively recent issue since the latest github action they run is 3 weeks old and was still using pillow 9.5

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 49.39759% with 42 lines in your changes are missing coverage. Please review.

Project coverage is 29.43%. Comparing base (0c1f57a) to head (1df919a).
Report is 29 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master      #65      +/-   ##
==========================================
- Coverage   33.65%   29.43%   -4.23%     
==========================================
  Files          19       19              
  Lines        2095     2161      +66     
==========================================
- Hits          705      636      -69     
- Misses       1390     1525     +135     

Files	Coverage Δ
pspy/so_cov.py	7.94% <ø> (-18.13%)	⬇️
pspy/so_mcm.py	51.27% <92.10%> (-6.81%)	⬇️
pspy/so_map_preprocessing.py	12.37% <13.33%> (-1.92%)	⬇️

... and 2 files with indirect coverage changes

[xzackli]

@xgarrido @thibautlouis can I merge perhaps tomorrow, with a promise that tests will pass?

[xzackli]

@xgarrido @zatkins2 this is ready to merge!

[xgarrido]

Adding @thibautlouis to the conversation since he is the grand master of pspy