Add MvNormal moment

[yadav-sachin]

Add moment and tests for the below distribution as part of #5078

• pymc.distributions.multivariate.MvNormal

[ricardoV94]

@yadav-sachin Thanks for opening the PR. I left some comments above. The MvNormal might be a tough nut to crack, as it is not always straightforward to figure out what the right shape should be. Perhaps @michaelosthege will be able to give you some tips

[Sayam753]

@yadav-sachin The conflicts need to be resolved. Otherwise the changes look good to me 🚀

[codecov]

Codecov Report

Merging #5171 (2595160) into main (38f2632) will increase coverage by 0.03%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #5171      +/-   ##
==========================================
+ Coverage   78.07%   78.10%   +0.03%     
==========================================
  Files          88       88              
  Lines       14176    14182       +6     
==========================================
+ Hits        11068    11077       +9     
+ Misses       3108     3105       -3     

Impacted Files	Coverage Δ
pymc/distributions/multivariate.py	71.72% <100.00%> (+0.24%)	⬆️
pymc/backends/report.py	91.60% <0.00%> (+2.09%)	⬆️

[Sayam753]

You can run pre-commit run --all-files locally. Refer to PyMC Python Style Guide in case pre-commit is not configured.

[yadav-sachin]

Let me know if any more changes are required. Thanks for the suggestions and reviews.

[yadav-sachin]

@ricardoV94 Two checks are failing on this pull request. What should I do to resolve them?

[michaelosthege]

I think that one failing test is flaky, related to #4771, so don't worry about it.

But I just checked this on my phone, so maybe look into it and double check if it may involve the MvNormal moment

[ricardoV94]

I think that one failing test is flaky, related to #4771, so don't worry about it.

But I just checked this on my phone, so maybe look into it and double check if it may involve the MvNormal moment

Shouldn't matter, because we are not using moments by default yet

[ricardoV94]

I think we can drop these 3 as well

[ricardoV94]

Thanks!