I just merged my PR for more poisson rates statistics for one and two samples
https://www.statsmodels.org/dev/stats.html#rates
I'm slowly converging to a pattern for one and two sample statistics and corresponding inference
like basid hypothesis tests, confidence intervals, equivalence tests, noninferiority tests as byproduct, non-equivalence tests, power functions, (sample size functions not in PR), tolerance intervals, ...
It still needs more review, and enhancements (but 2700 lines are enough for this PR)
Problem for Poisson rates was that the methods are spread out over many articles with small variations in the methods and no good overview article that I found.
I found some parts in R, spread over many packages, each having a few different pieces.
I'm planning to refactor and cleanup proportions to a similar pattern before the next release
Josef