weightederm - Weighted Empirical Risk Minimization for Changepoint Regression

R interface to the 'weightederm' package for 'Python', which provides 'scikit-learn'-style estimators for offline change point regression (data segmentation) via weighted empirical risk minimization. Supports least-squares, Huber, and logistic losses with fixed or cross-validated numbers of change points. Wraps 'Python' via 'reticulate'. Arpino and Venkataramanan (2026) <doi:10.48550/arXiv.2604.11746>.