As part of our NSF-funded initiative, we are working with two pilot projects that highlight different aspects of the statistical Python ecosystem:
YAGLM: A Python package for generalized linear models (GLM), offering modern statistical methodology with various loss functions, regularizers, and solvers. GLMs are a flexible and powerful generalization of ordinary linear regression and cover many statistical models widely used in applications. Beyond the basic LASSO/ridge penalties, the package supports structured penalties such as the nuclear norm as well as the group, exclusive, fused, and generalized lasso and non-convex penalties. It comes with a variety of tuning parameter selection methods including: cross-validation, information criteria that have favorable model selection properties, and degrees of freedom estimators.
ISLP: A companion Python package for the textbook “An Introduction to Statistical Learning with Applications in Python,” closely following the examples in its hugely successful
Rversion. Besides just providing labs, it provides Pythonic design matrix building, a simple implementation of Bayesian Additive Regression Trees and object oriented stepwise model selection.