causalsim - Simulation-Ready Causal Data Generating Processes

Construct, simulate, and evaluate causal data generating processes (DGPs) with known ground truth. Designed for benchmarking causal estimators, studying confounding and treatment-effect heterogeneity, and building reproducible teaching examples. Covariate roles (confounder, effect modifier, noise) and heterogeneous treatment effects are first-class concepts in the API, and estimator performance is summarised with bias, root mean squared error, confidence-interval coverage, and power.