inv.repam {survPen}

Reverses the initial reparameterization for stable evaluation of the log determinant of the penalty matrix

Description

Transforms the final model by reversing the initial reparameterization performed by repam. Derives the corrected version of the Bayesian covariance matrix

Usage

inv.repam(model, X.ini, S.pen.ini)

Arguments

model

survPen object, see survPen.fit for details

X.ini

initial design matrix (before reparameterization)

S.pen.ini

initial penalty matrices

Value

survPen object with standard parameterization