Turn "which statistical test do I use?" into Bayesian estimation problems, with PyMC3 as the workhorse.
You've got data, and now you want to analyze it with Python. Do you do the t-test? The chi-squared test? How do you decide? In this PyCon 2017 talk, I show you how to take common statistical decision problems, formulate them as Bayesian estimation problems, and use PyMC3 as the workhorse. Four worked examples cover parameter estimation and case/control (A/B) comparisons. The talk is math-light and code-heavy, and the accompanying notebooks give you a template for your own Bayesian analyses.
Notebooks are available here.