Examples¶
Practical examples demonstrating DoubleML PySpark capabilities for causal inference at scale.
Interactive Notebooks¶
- Business example: coupon targeting: Does a coupon campaign pay for itself, and for which customers? Naive comparison, regression and DoubleML side by side, effects by segment, and a targeting rule that turns the loss into a profit. Runs on serverless in a few minutes.
- PLR ATE: Estimation of Average Treatment Effects with PLR.
- IRM ATE: Estimation of Average Treatment Effects with IRM.
- PLR CATE: Estimation of Conditional Average Treatment Effects with PLR.
- IRM CATE: Estimation of Conditional Average Treatment Effects with IRM.
Each example includes:
- Data setup: loading and preparing synthetic or real-world datasets.
- Model configuration: specifying learners, parameters, and estimation settings.
- Results interpretation: understanding outputs, confidence intervals, and causal conclusions.
Start with the business example to see what the method delivers in practice, then use PLR ATE for a closer look at the estimators.