I am the developer of PyCEBox, the Python Individual Conditional Expectation Toolbox. Individual conditional expectation plots are a tool for visualizing the relationship between features and precitions for black box machine learning models. PyCEBox uses Pandas and Matplotlib to produce individual conditional expectation plots for any machine learning model that provides a scikit-learn-like prediction interface.
I contributed to a project that ported the book Statistical Rethinking: A Bayesian Course with Examples in R and Stan from R and Stan to Python and PyMC3. I ported Chapter 8, Markov Chain Monte Carlo and Chapter 11, Monsters and Mixtures.