pymatchit-causal: Propensity Score Matching and Causal Inference in Python

pymatchit-causal: Propensity Score Matching and Causal Inference in Python pymatchit-causal provides a robust, "R-style" workflow for matching and observational data preprocessing in Python. It allows researchers and data scientists to mitigate selection bias by balancing covariates between treated and control groups, bringing the functionality of the standard R package MatchIt to the Python ecosystem. Release v0.5.0 Highlights: Advanced Matching Algorithms: This release introduces Full Matching, Genetic Matching, and Cardinality Matching, complementing the existing Nearest Neighbor, Exact, Subclassification, and Coarsened Exact Matching (CEM) methods.Expanded Distance Estimators: Addition of Covariate Balancing Propensity Score (CBPS) optimization, alongside new machine learning estimators (Decision Trees, AdaBoost, Lasso, Ridge, ElasticNet), and the ability to seamlessly use user-supplied propensity scores.Publication-Ready Diagnostics: A completely overhauled visual diagnostic suite featuring visually aligned Love Plots, ECDF Plots, Propensity Density Plots, and the introduction of new Jitter and QQ Plots for comprehensive verification of match quality and covariate balance.Enhanced support for advanced configurations, including targeting ATE vs. ATT estimands, exact variable matching prior to propensity evaluation (mahvars, exact, constraints), and discarding units outside common support. For extensive documentation, API references, and example workflows, please visit the GitHub repository at https://github.com/jtuenner/pymatchit