ctrl_GPCMlasso: Control function for GPCMlasso
In GPCMlasso: Regularized Explanatory Generalized Partial Credit Models

View source: R/ctrl_GPCMlasso.R

ctrl_GPCMlasso

R Documentation

Control function for GPCMlasso

Description

Control parameters for penalty terms and for tuning the fitting algorithm.

Usage

ctrl_GPCMlasso(
  log.lambda = TRUE,
  lambda = NULL,
  l.lambda = 50,
  lambda.min = 0.1,
  penalize.main.effects = FALSE,
  fuse.per.variable = FALSE,
  scale.covariate.effects = TRUE,
  include.item.effects = TRUE,
  adaptive = TRUE,
  weight.penalties = TRUE,
  ada.lambda = 1e-04,
  ada.power = 1,
  Q = 15,
  lambda2 = 1e-04,
  cvalue = 1e-05,
  trace = TRUE,
  folds = 10,
  cores = 25,
  null_thresh = 0.01,
  gradtol = 1e-06,
  steptol = 1e-06,
  iterlim = 500,
  precision = 3,
  all.dummies = FALSE,
  ctrl.gpcm = list()
)

Arguments

`log.lambda`	Logical. Should the grid of tuning parameters be created on a log scale?
`lambda`	Optional vector of tuning parameters. If `NULL`, a grid of length `l.lambda` is created automatically.
`l.lambda`	Length of the tuning parameter grid.
`lambda.min`	Minimal value used for the automatically created tuning parameter grid.
`penalize.main.effects`	Logical. Should global covariate effects be penalized?
`fuse.per.variable`	Logical. Should item-specific effects referring to the same covariate be fused across items?
`scale.covariate.effects`	Logical. If `TRUE`, covariate effects are multiplied by item discrimination parameters as in the original GPCMlasso parameterization. If `FALSE`, covariate effects enter the linear predictor directly.
`include.item.effects`	Logical. If `TRUE`, item-specific covariate effects are included. If `FALSE`, only global covariate effects are estimated.
`adaptive`	Logical. Should adaptive lasso weights be used?
`weight.penalties`	Logical. Should penalties be weighted according to the number of penalty terms and corresponding parameters?
`ada.lambda`	Tuning parameter for ridge-regularized estimation of adaptive weights.
`ada.power`	Power used for adaptive weights.
`Q`	Number of nodes used in Gauss-Hermite quadrature.
`lambda2`	Ridge penalty on all coefficients except slope parameters.
`cvalue`	Internal parameter for the quadratic approximation of the `L_1` penalty.
`trace`	Logical. Should progress information be printed?
`folds`	Number of folds for cross-validation.
`cores`	Number of cores used for parallel computation.
`null_thresh`	Threshold used to distinguish zero from non-zero values.
`gradtol`	Parameter for optimization accuracy. See `nlm`.
`steptol`	Parameter for optimization accuracy. See `nlm`.
`iterlim`	Parameter for optimization accuracy. See `nlm`.
`precision`	Number of decimal places used to round coefficient estimates in selected output methods.
`all.dummies`	Logical. Should all dummy variables be included for factors with more than two categories?
`ctrl.gpcm`	List of control arguments for the initial `gpcm` fit used to obtain starting values.

Value

A list of control parameters used by GPCMlasso.

Author(s)

Gunther Schauberger
gunther.schauberger@tum.de

References

Schauberger, G. and Mair, P. (2020). A regularization approach for the detection of differential item functioning in generalized partial credit models. Behavior Research Methods, 52, 279–294. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.3758/s13428-019-01224-2")}

Examples

ctrl <- ctrl_GPCMlasso()
str(ctrl)

ctrl_small <- ctrl_GPCMlasso(
  l.lambda = 5,
  cores = 1,
  trace = FALSE
)

ctrl_fusion <- ctrl_GPCMlasso(
  l.lambda = 5,
  cores = 1,
  penalize.main.effects = TRUE,
  include.item.effects = TRUE,
  fuse.per.variable = TRUE,
  scale.covariate.effects = FALSE,
  trace = FALSE
)

GPCMlasso documentation built on Sept. 8, 2026, 5:08 p.m.