Changes in version 0.4.1 (2026-11-18) - Added method references to DESCRIPTION and removed an unnecessary global scipen option change from the guide vignette in response to CRAN review. Changes in version 0.4.0 - Reworked fitting with finite coefficient bounds to use projected proximal gradient, enforcing the bounds throughout optimization and supporting the L1 penalty with soft-thresholding. - Expanded tests and documentation for bounded fitting and optimizer behavior. Changes in version 0.3.3 - Update function documentations - Fix parallel processing bug when running on machines with more than 128 cores Changes in version 0.3.2 - Added cv_capnet() support for parallel execution via PSOCK cluster. - Added family argument to capnet(), cv_capnet(), and walk_capnet() supporting Gaussian, binomial, Poisson, and Gamma families. - Added plot.cv_capnet() heatmap and slice visualizations for cross-validation error surfaces. - Added plot.capnet_path() for coefficient path plots along a hyperparameter dimension. - Fixed contribution cap gradient scale-invariance under standardization. Changes in version 0.3.0 - Added walk_capnet() for walk-forward evaluation with per-step contribution cap enforcement. - Added coef_path() for fitting models along a hyperparameter grid. - Added capnet_violations() to diagnose which (row, feature) pairs exceed their cap. - Added plot_redistribution() to visualize coefficient and contribution shifts across models. Changes in version 0.2.0 - Added cv_capnet() for K-fold cross-validation over alpha and lambda. - Added simulate_capnet_data() for generating synthetic data with known contribution structure. Changes in version 0.1.0 - Initial release. Core capnet() estimator with OWL-QN optimization for elastic net plus contribution cap penalty.