View source: R/ctrl_GPCMlasso.R
ctrl_GPCMlasso | R Documentation |
Control parameters for penalty terms and for tuning the fitting algorithm.
ctrl_GPCMlasso( log.lambda = TRUE, lambda = NULL, l.lambda = 50, lambda.min = 0.1, 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 )
log.lambda |
Should the grid of tuning parameters be created on a log scale? |
lambda |
Optional argument to specify a vector of tuning parameters. If |
l.lambda |
Specifies the length of the grid of tuning parameters. |
lambda.min |
Minimal value used if the grid of tuning parameters is created automatically. |
adaptive |
Should adaptive lasso be used? Default is |
weight.penalties |
Should penalties be weightes accoreding to the number of penalty term and the number of parameters
corresponding to one pair between item and covariate. Only relevant if both |
ada.lambda |
Size of tuning parameter for Ridge-regularized estimation of parameters used for adaptive weights. |
ada.power |
By default, 1st power of absolute values of Ridge-regularized estimates are used. Could be changed to squared values by |
Q |
Number of nodes to be used in Gauss-Hermite quadrature. |
lambda2 |
Tuning parameter for ridge penalty on all coefficients except sigma/slope parameters. Should be small, only used to stabilize results. |
cvalue |
Internal parameter for the quadratic approximation of the L1
penalty. Should be sufficiently small. For details see
|
trace |
Should the trace of the progress (current tuning parameter) be printed? |
folds |
Number of folds for cross-validation. Only relevant if |
cores |
Number of cores to be used parallel when fitting the model. |
null_thresh |
Threshold which is used to distinguih between values equal and unequal to zero. |
gradtol |
Parameter to tune optimization accuracy, for details see |
steptol |
Parameter to tune optimization accuracy, for details see |
iterlim |
Parameter to tune optimization accuracy, for details see |
precision |
Number of decimal places used to round coefficient estimates. |
all.dummies |
Should (in case of factors with more than 2 categories) the dummy variables for all categories be included in the design matrix? If |
Gunther Schauberger
gunther.schauberger@tum.de
Schauberger, Gunther and Mair, Patrick (2019): A Regularization Approach for the Detection of Differential Item Functioning in Generalized Partial Credit Models, Behavior Research Methods, https://link.springer.com/article/10.3758/s13428-019-01224-2
data(tenseness_small) ## formula for simple model without covariates form.0 <- as.formula(paste("cbind(",paste(colnames(tenseness_small)[1:5],collapse=","),")~0")) ###### ## fit simple RSM where loglikelihood and score function are evaluated parallel on 2 cores rsm.0 <- GPCMlasso(form.0, tenseness_small, model = "RSM", control= ctrl_GPCMlasso(cores=2)) rsm.0 ## Not run: ## formula for model with covariates (and DIF detection) form <- as.formula(paste("cbind(",paste(colnames(tenseness_small)[1:5],collapse=","),")~.")) ###### ## fit GPCM model with 10 different tuning parameters gpcm <- GPCMlasso(form, tenseness_small, model = "GPCM", control = ctrl_GPCMlasso(l.lambda = 10)) gpcm plot(gpcm) pred.gpcm <- predict(gpcm) trait.gpcm <- trait.posterior(gpcm) ###### ## fit RSM, detect differential step functioning (DSF) rsm.DSF <- GPCMlasso(form, tenseness_small, model = "RSM", DSF = TRUE, control = ctrl_GPCMlasso(l.lambda = 10)) rsm.DSF plot(rsm.DSF) ## create binary data set tenseness_small_binary <- tenseness_small tenseness_small_binary[,1:5][tenseness_small[,1:5]>1] <- 2 ###### ## fit and cross-validate Rasch model set.seed(1860) rm.cv <- GPCMlasso(form, tenseness_small_binary, model = "RM", cv = TRUE, control = ctrl_GPCMlasso(l.lambda = 10)) rm.cv plot(rm.cv) ## End(Not run)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.