lognetProbabilityReusable: Function to help calculate predicted probability of a lognet...
In EMLasso: Fit (logistic) LASSO when there is missing data in the predictors

Description Usage Arguments Value Author(s) See Also Examples

View source: R/lognetProbabilityReusable.R

Method that provides information that can be reused to predict probabilities from the same lognet several times

1
2
3

  lognetProbabilityReusable(lognet, ds,
    imputeDs2FitDsProperties = normalImputationConversion(),
    usecol, verbosity = 0)

`lognet`	binomial `glmnet` fit to use for prediction
`ds`	`numdfr` or `data.frame` that holds the structure from which dataset predictions will need to happen.
`imputeDs2FitDsProperties`	see `imputeDs2FitDs` object that will provide the conversion from imputed dataset to one that is ready for fitting the predictor model
`usecol`	which of the column of beta must be used. Must be provided if `lognet` is not the result of a 1-lambda `glmnet` call.
`verbosity`	The higher this value, the more levels of progress and debug information is displayed (note: in R for Windows, turn off buffered output)

object of class "lognetProbabilityReusable": list with items:

`conversionProps`	`data.frame` with columns `fromCols` (original column number), `newNames` (new column name), `mustEqual` (which value must it equal for this level of the factor), `useBeta` (coefficient)
`originalLognet`	`lognet` that was passed in
`usedcol`	usecol that was passed in (or 1 if only it was missing)

Nick Sabbe nick.sabbe@ugent.be

dfrConversionProps

data(iris)
	iris.nd2<-numdfr(iris)
	y2<-rbinom(nrow(iris), 1, 0.5)
	iris.nic2<-normalImputationConversion(
		scalingParams=typicalScaleAndCenter(),
		transformParams=typicalTransformations())
	iris.cp2<-imputeDs2FitDsProps(iris.nic2,iris.nd2,verbosity=1)

	iris.cvtd2<-imputeDs2FitDs(iris.cp2,ds=iris.nd2,verbosity=3)

	lnet<-glmnet(iris.cvtd2, y2, family="binomial")
	lognetProbabilityReusable(lnet, iris.nd2, imputeDs2FitDsProperties=iris.cp2, usecol=5, verbosity=1)