Adding StickBreakingWeights distribution class

[larryshamalama]

This pull request adds a class for truncated stick-breaking weights which are vital in the construction of Dirichlet Processes, which I hope to work on soon. See section 1.1 in Ishwaran and James (2001). My previous attempt as a PR is #4971 .

Some tests are failing, will have to resolve them first.

CC: @ricardoV94 @AustinRochford @fonnesbeck

[larryshamalama]

Two sets of tests not passing locally in test_distributions.py and test_distributions_moments.py. The test in test_distributions_random.py seems okay (from my perspective)

[codecov]

Codecov Report

Merging #5200 (7885768) into main (d52655d) will increase coverage by 0.01%.
The diff coverage is 93.33%.

@@            Coverage Diff             @@
##             main    #5200      +/-   ##
==========================================
+ Coverage   80.36%   80.38%   +0.01%     
==========================================
  Files          89       89              
  Lines       14808    14867      +59     
==========================================
+ Hits        11901    11951      +50     
- Misses       2907     2916       +9     

Impacted Files	Coverage Δ
pymc/distributions/__init__.py	100.00% <ø> (ø)
pymc/distributions/multivariate.py	75.57% <93.33%> (+1.44%)	⬆️
pymc/distributions/continuous.py	96.93% <0.00%> (-0.50%)	⬇️

[larryshamalama]

Question about test_distributions.py

[ricardoV94]

Looks great. Left some comments bellow, mostly concerning tests. Glad to see this PR entering the polishing phase!

[larryshamalama]

December 20 work in progress

[larryshamalama]

Two new sets of tests + question in make_node

[ricardoV94]

Looks great! Left some comments, let me know if anything is confusing

[ricardoV94]

LGTM. Awesome work @larryshamalama

[larryshamalama]

As always, thanks so much for all your suggestions and help @ricardoV94

[twiecki]

Would be great to have an example NB in pymc-examples that shows how to use this.