uber/causalml

Implements multiple meta-learner algorithms (S-learner, T-learner, X-learner, R-learner, doubly robust methods) and tree-based approaches for heterogeneous treatment effect estimation from both experimental and observational data. Provides a unified interface for CATE estimation with built-in support for feature selection, multi-treatment scenarios, and cost optimization. Integrates with scikit-learn estimators and supports both traditional ML and deep learning backends for flexible model composition.

5,758 stars and 70,222 monthly downloads. Used by 1 other package. Actively maintained with 8 commits in the last 30 days. Available on PyPI.