Python package containing a couple of functions for Bayesian A/B and Multivariate testing. It can be used to:
-
Run a Bayesian analysis on an AB/multivariate test that you have run, returning also a Bayesian credible intervals for the relative uplift;
-
Calculate the minimum sample size required for a Bayesian AB/multivariate test, useful for designing future tests
The package is especially useful when doing repeated tests (as it avoids the multiple testing bias) and when you want to incorporate knowledge/results coming from previous tests.