mxPenaltyZap: mxPenaltyZap

View source: R/MxPenalty.R

mxPenaltyZapR Documentation

mxPenaltyZap

Description

Fix any free parameters within epsilon of zero to zero. These parameters are no longer estimated. Remove all MxPenalty objects from the model. This is envisioned to be used after using mxPenaltySearch to locate the best penalty hyperparameters and apply penalties to model estimation. While penalties can simplify a model, they also bias parameters toward zero. By re-estimating the model after using mxPenaltyZap, parameters that remain free are likely to exhibit less bias.

Usage

mxPenaltyZap(model, silent = FALSE)

Arguments

model

an MxModel

silent

whether to suppress diagnostic output


OpenMx documentation built on Nov. 8, 2023, 1:08 a.m.