pkmon

We implement two least-squares estimators under k-monotony constraint using a method based on the Support Reduction Algorithm from Groeneboom et al (2008) doi:10.1111/j.1467-9469.2007.00588.x. The first one is a projection estimator on the set of k-monotone discrete functions. The second one is a projection on the set of k-monotone discrete probabilities. This package provides functions to generate samples from the spline basis from Lefevre and Loisel (2013) doi:10.1239/jap/1378401239, and from mixtures of splines.