Refactor transformations to rely on scipy directly.

[kyleaoman]

Rotations were inconsistent with the scipy rotation system even though that's how they were asked to be defined. For example:

rotation = scipy.spatial.transform.Rotation.from_matrix(...)  # ... is some rotation matrix
coordinates_before = sg.gas.coordinates
sg.rotate(rotation)
# glossing over Rotation.apply stripping cosmo attributes & comparison of arrays with !=, conceptually:
sg.gas.coordinates != rotation.apply(coordinates_before)  # the inverse was actually applied!

I've refactored the whole coordinate transformation backend to rely much more on scpiy.spatial.transform. Unfortunately it's not directly compatible with unyt_array or cosmo_array, so a little bit of wrapping is still needed, but much of the previous internal ambiguity around .dot(transformation) (or should it be .dot(transformation.T?) has been removed.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (e6e7429) to head (936cabd).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##              main       #86   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            7         7           
  Lines         2289      2286    -3     
  Branches       260       262    +2     
=========================================
- Hits          2289      2286    -3     

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.