RM_DIF_lbfgs: Rasch model with DIF via lbfgs (only L1)

In submission-user/cmlDIFlasso: DIF-detection with Lasso-methods in a cML-Framework

Description

optimizes loss_function with DIF-part for certain tuning-parameter lambda with lbfgs

Usage

RM_DIF_lbfgs(
  data,
  covariates,
  starting_betas = rep(0, ncol(data)),
  starting_deltas = matrix(rep(0, ncol(data) * ncol(covariates)), ncol(covariates),
    ncol(data)),
  lambda = 0,
  ...
)

Arguments

data	matrix of responses
covariates	Matrix of covariate values of all persons (n x num. of covs)
starting_betas	beta_intercepts starting values for all items (vector)
starting_deltas	starting values for DIF-parameters as matrix (ncovs x nitems)
lambda	penalization parameter
...	other arguments to be passed to lbfgs

Value

result of optim-optimization for DIF-model via cML-loss-function for a certain penalisation-type (for now only all.interactions) and lambda value

submission-user/cmlDIFlasso documentation built on Dec. 23, 2021, 6:40 a.m.