Skip to contents

Calculate the penalised loglikelihood of a smooth.spline using the integrated second derivative. The likelihood consists of 1) the (weighted) residuals sum of squares, 2) a penalty term (integrated second derivative = total curvature). The smaller the penalised loglikelihood, the better the fit as the residuals and penalty on roughness are minimised. Adapted from aroma.light::likelihood.smooth.spline.

Usage

loglik_smooth_spline(fittedSmoothSpline)

Arguments

fittedSmoothSpline

A fitted smooth.spline

Value

The penalised loglikelihood.