CausalML in 60 Minutes
TL;DR: Learn how to estimate heterogeneous treatment effects using CausalML in 60 minutes with hands-on examples including individual-level causal inference on CDC diabetes health data.
Tutorial in 30 Seconds#
CausalML is an open-source Python library from Uber for causal machine learning, providing a suite of meta-learner algorithms to estimate individualized treatment effects from observational data.
Key capabilities:
- Meta-learners: S, T, X, R, and DR-Learner algorithms for heterogeneous treatment effect estimation
- Uplift modeling: Identify who benefits most from a treatment or intervention
- Robustness checks: Placebo tests, sensitivity analysis, and estimator comparisons built in
- Scikit-learn compatible: Integrates with any sklearn-compatible base learner
This tutorial's goal is to show you in 60 minutes:
- The basic API of CausalML (an open-source library for causal machine learning)
- Concrete examples of using CausalML to estimate who benefits most from physical activity using CDC BRFSS diabetes health data
Official References#
Tutorial Content#
This tutorial includes all the code, notebooks, and Docker containers in tutorials/CausalML_Diabetes_Study
README.md: Instructions and setup for the tutorial environment- A Docker system to build and run the environment using our standardized approach
CausalML.API.ipynb: Tutorial notebook focusing on the CausalNavigator API and meta-learner configurationsCausalML.example.ipynb: Advanced end-to-end causal inference example- Loads and preprocesses the CDC BRFSS diabetes dataset (250,000+ respondents)
- Checks causal assumptions (overlap/positivity) using propensity score analysis
- Estimates individualized treatment effects (CATE) using the X-Learner
- Visualizes heterogeneity across age, income, and health status subgroups
- Validates results with placebo tests, estimator comparisons, and sensitivity analysis
causalml_utils.py: Utility functions and theCausalNavigatorwrapper class