The kofm() function computes the "kStar" multipliers and the diagonal "d" Matrices used by the unr.ridge() function. While these calculations are extensive, they are greatly simplified by using interpolation between "knots" on "piecewise linear-spline" functions.
kofm(muobj, p, dMSE, delmax = 0.999999)
A "numeric" value that is >= 0 and <= p.
The integer number of non-constant x-variables used in defining the linear model being fitted to ill-conditioned (intercorrelated, confounded) data. Note that p must also be rank of the given X-matrix.
Maximum Likelihood estimates of minimum MSE risk Shrinkage Delta-Factors.
Maximum allowed value for Shrinkage delta-factors that is strictly less than 1. (default = 0.999999, which prints as 1 when rounded to fewer than 6 decimal places.)
An output list of two objects:
The k-factor needed to multiply each given dMSE-factor.
The (p x p) diagonal matrix of requested shrinkage-factors.
Obenchain RL. (2020) Ridge TRACE Diagnostics. https://arxiv.org/abs/2005.14291
Obenchain RL. (2020) The Unrestricted Shrinkage Path: Technical Details. "unrPath.pdf" http://localcontrolstatistics.org