causalml-advanced

This repository contains a collection of Jupyter notebooks focused on causal inference for both experimental and observational data. Each notebook provides examples using simulated data.

The notebooks are organized as follows:

1. Selection Bias: This notebook shows how selection bias arise from observational data.
2. CUPED: This notebook shows how to use regression adjustment and CUPED with heterogeneous treatment effects.
3. Multiple Randomized Designs: This notebook demonstrates how to analyze marketplaces where SUTVA is violated.
4. Power Analysis: This notebook shows how to conduct power analysis for randomized experiments.
5. Instrumental Variables: This notebook shows how to use instrumental variables to estimate causal effects and the case of weak instruments.
6. Regression Discontinuity Design: This notebooks is based on Chapter 17 of the book Causal ML. It shows RDD with a sharp design with and without covariates.
7. Difference-in-Differences: This notebook shows how to use two-way fixed effects estimator in the traditional 2x2 setting and the problems with staggered adoption.
8. Modern DiD Approaches: This notebook uses Callaway and Sant'Anna (2020) to show how to estimate ATT with staggered adoption.
9. Synthetic Control: Estimation of treatment effects using synthetic control methods.
10. Matching: IN PROGRESS