Updating docstrings of distributions

[vitaliset]

What is this PR about?

In response to issue #5459. Me and @pibieta updated docstrings for some distribution classes inside the multivariate.py module, namely:

• LKJCholeskyCov
• LKJCorr
• MatrixNormal
• KroneckerNormal
• CAR

Attention points

• Inside the KroneckerNormal class, the docstring types of covs, chols and evds were left as list of arrays and list of tuples because the proper type is being developed as @OriolAbril told us.
• Inside the KroneckerNormal class the type of sigma should be float as far as I understand but we were not sure because the old docstring called it a variable.
• Inside the CAR class, there was a Numpy matrix input that we changed to ndarray of int. Does this make sense?
• Inside the MatrixNormal class there were some nxn array that we changed to nxn tensor_like of float. Does that make sense or we should put something else?

Docs / Maintenance

updated docstrings to follow the numpydoc standard

#DataUmbrellaPyMCSprint
cc: @reshamas

[codecov]

Codecov Report

Merging #5998 (2575662) into main (f9e0dfd) will decrease coverage by 0.06%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##             main    #5998      +/-   ##
==========================================
- Coverage   89.36%   89.29%   -0.07%     
==========================================
  Files          73       72       -1     
  Lines       13253    12884     -369     
==========================================
- Hits        11843    11505     -338     
+ Misses       1410     1379      -31     

Impacted Files	Coverage Δ
pymc/distributions/multivariate.py	92.03% <ø> (ø)
pymc/backends/base.py	86.44% <0.00%> (-0.74%)	⬇️
pymc/sampling_jax.py	96.96% <0.00%> (ø)
pymc/step_methods/__init__.py	100.00% <0.00%> (ø)
pymc/step_methods/mlda.py
pymc/distributions/continuous.py	97.89% <0.00%> (+<0.01%)	⬆️
pymc/sampling.py	82.48% <0.00%> (+0.07%)	⬆️
pymc/util.py	77.64% <0.00%> (+0.40%)	⬆️
pymc/model_graph.py	95.45% <0.00%> (+1.75%)	⬆️
pymc/__init__.py	100.00% <0.00%> (+31.57%)	⬆️

[vitaliset]

Regarding the W parameter of CAR and the sigma parameter of KroneckerNormal, do you think we should do something @OriolAbril?

[OriolAbril]

Also, if you look at the preview, the M and N are rendered in pink, which is not good. Can you add them to the numpydoc_xref_ignore set? This will make sure they are shown in bold black font.

[twiecki]

Thanks @vitaliset!