thresHolding: Simple threshold selection for some sets of variables...
In EMLasso: Fit (logistic) LASSO when there is missing data in the predictors

Description Usage Arguments Value Note Author(s) Examples

Simple threshold selection for some sets of variables selected by EMLasso

  thresHolding(cvob, coefsUsed = NULL, reps = 500,
    checkThres = seq(0, 1, length.out = 101),
    ds = cvob$glmnet.fit$result[[1]]$ds,
    out = cvob$glmnet.fit$result[[1]]$out,
    wts = rep(1, nrow(ds)),
    imputeDs2FitDsProperties = cvob$glmnet.fit$imputeDs2FitDsProperties,
    ..., niceNames = c(names(coefsUsed)),
    newWindow = FALSE, showWanted = c(0.5, 0.75),
    showThres = 10, verbosity = 0)

`cvob`	result of a call to `crossValidate.EMLassoGLoMo`
`coefsUsed`	list holding vectors of names of columns (after conversion). Thresholds are selected for each of the sets of variables.
`reps`	number of repeats of imputation / crossvalidation
`checkThres`	set of values between 0 and 1 where the TPR / FPR are evaluated
`ds`	dataset (defaults to the one passed to `EMLasso`)
`out`	outcome vector (defaults to the one passed to `EMLasso`)
`wts`	weight vector per observation (defaults to the one passed to `EMLasso`)
`imputeDs2FitDsProperties`	see `imputeDs2FitDs` and `EMLasso`
`...`	passed on to `collectImputationModels`
`niceNames`	display names for each of the sets of variabels, to be used n the titles of the plots
`newWindow`	if `TRUE` (not the default), a new window is created for the plots
`showWanted`	if a vector of two numbers, the first is indicated in a green vertical line, and the second as a green horizontal line
`showThres`	see `plotROCFromRepPredProb`
`verbosity`	The higher this value, the more levels of progress and debug information is displayed (note: in R for Windows, turn off buffered output)

a list, holding the following items:

Thresholds are selected by finding the highest TPR with a maximal FPR of 50

Nick Sabbe nick.sabbe@ugent.be

y<-rbinom(nrow(iris), 1, 0.5)
iris.cpy<-randomNA(iris, n=0.1)
iris.emlognet<-EMLasso(ds=numdfr(iris.cpy), out=y,
	lambdas=c(0.03,0.002,0.0003), nrOfSamplesPerMDRow=7, verbosity=0,
	convergenceChecker=convergenceCheckCreator(minIt=5, maxIt=10))
sfStop()
iris.cv.emlognet<-crossValidate(iris.emlognet, verbosity=0)
thresHolding(iris.cv.emlognet, coefsUsed=list(onlySepL=c("Sepal.Length"), onlySpecies=c("Speciesversicolor", "Speciesvirginica")), reps=10, verbosity=2)