Fix bug in random function of HalfStudent

[Dhruvanshu-Joshi]

What is this PR about?
In this PR, we aim to solve the issue 6657. Originally what was passed as sigma is now passed by keyword as scale.

Checklist

• Explain important implementation details 👆
• Make sure that the pre-commit linting/style checks pass.
• Link relevant issues (preferably in nice commit messages)
• Are the changes covered by tests and docstrings?
• Fill out the short summary sections 👇

Major / Breaking Changes

• ...

New features

• ...

Bugfixes

• Fixes the bug as found in Discourse_link

Documentation

• ...

Maintenance

• ...

---

📚 Documentation preview 📚: https://pymc--6658.org.readthedocs.build/en/6658/

[ricardoV94]

Oops #6654 seems to not be working, the changes in this PR should have triggered the tests.

[ricardoV94]

The tests not being triggered should be fixed once #6659 is merged

[ricardoV94]

@Dhruvanshu-Joshi If you rebase from main and push again the tests should run

[codecov]

Codecov Report

Merging #6658 (9c9c686) into main (5d68bf3) will increase coverage by 0.01%.
The diff coverage is 100.00%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6658      +/-   ##
==========================================
+ Coverage   91.97%   91.98%   +0.01%     
==========================================
  Files          94       94              
  Lines       15942    15942              
==========================================
+ Hits        14663    14665       +2     
+ Misses       1279     1277       -2     

Impacted Files	Coverage Δ
pymc/distributions/continuous.py	97.70% <100.00%> (ø)

... and 2 files with indirect coverage changes

[ricardoV94]

No test failed which means we need to tweak the HalfStudentT random test to cover the change (it should fail if the change hadn't been done or reverted)

[ricardoV94]

Since this is blocking a release I went ahead and added a test.

Apparently the HalfSTudentT random method was not being tested at all!

[Dhruvanshu-Joshi]

Hey @ricardoV94 . I was working on the test but was trying a different approach wherein I tried to generate the sample using pm.HalfStudentT rv and compare the mean and std of the sample with the theoretical mean and std. I am still pretty naive at generating tests and will go through the codebase again to have a better grasp and be quick to complete the task so the it does not block any future release. Thank You for your help.

[ricardoV94]

@Dhruvanshu-Joshi There was absolutely nothing wrong with the time you were taking!

In the end the test solution was easy: just use the framework we have in place to test the random methods.

I assumed this was already being used for the current RV but it didn't! I didn't give you the proper advice.

If anything it was my fault :)

[Dhruvanshu-Joshi]

Thanks @ricardoV94 . I'll go through the framework to better improve my understanding🙂

[ricardoV94]

There are more details in https://www.pymc.io/projects/docs/en/latest/contributing/implementing_distribution.html#adding-tests-for-the-new-randomvariable