predict.lspcr.glm: Predict method for LS-PCR fits.

In lsplsGlm: Classification using LS-PLS for Logistic Regression

Description

Obtains predictions and prediction probabilities from a fitted LS-PCR object.

Usage

## S3 method for class 'lspcr.glm'
predict(object,newX,newD,...)

Arguments

object	results from fit.lspcr.glm function.
newX	new matrix of genetic data.
newD	new matrix of clinical data.
...	further arguments. Currently not used.

Details

This function is used to obtained predicted values using a model fitting with fit.lspcr.glm. It returns predictions and prediction probabilites in case that the user wants to choose his own decision rule. By default (in the prediction vector) if probabilities are under 0.5 observations are in the 0 class and if probabilities are upper than 0.5 observations are in the 1 class.

Value

newY	the newY containing the n predicted values of the response variables for the observations from newX and newD.
newPi	the newPi containing the n probabilities of the response variables for the observations from newX and newD.
cvg	the 0-1 value indicating convergence of the algorithm (1 for convergence, 0 otherwise).

Author(s)

Caroline Bazzoli, Thomas Bouleau, Sophie Lambert-Lacroix

See Also

fit.lspcr.glm.

Examples

#Data
data(BreastCancer)
#Vector of response
Y<-BreastCancer$Y
#Genetic data
X<-BreastCancer$X
#Clinical data
D<-BreastCancer$D

#Learn dataset and test dataset (80/20)
index<-preselected.sample(Y,round(length(Y)*0.8))
XL<-X[index,]
XT<-X[-index,]
DL<-D[index,]
DT<-D[-index,]
YL<-Y[index]
#Apply fit.lspcr.glm with 5 components
fit<-fit.lspcr.glm(Y=YL,X=XL,D=DL,ncomp=5)
#predictions
pred<-predict.lspcr.glm(fit,newD=DT,newX=XT)

lsplsGlm documentation built on May 2, 2019, 12:36 p.m.