Fig. 7
From: ecpc: an R-package for generic co-data models for high-dimensional prediction
Data analysis example: a Estimated prior variance contributions of each co-data source, before multiplying with the co-data specific weight. Note that the p values are shown on the log-scale in Settings 2 and 3, to clearly show the non-zero peaks at the smallest p values; b corresponding prediction performance on the validation set for 20 or 50 spline basis functions. The settings correspond to different co-data models: (1) no constrains; (2) positive constrained shape; (3) positive and monotonically constrained shape