customRSF: Random Survival Forests using the 'randomSurvivalForest' R...
In bernau/survHDExtra: Addtitional Features for package survHD

Description Usage Arguments Value Author(s) References See Also Examples

Constructs Random Survival Forests using the function rsf. Usually, this function is not called directly but embedded in a call to customSurv from the main package survHD (see examples).

1	customRSF(Xlearn, Ylearn, learnind, ...)

`Xlearn`	Gene expression data (a `data.frame`)
`Ylearn`	Survival Response, an object of class `Surv`.
`learnind`	An index vector specifying the observations that belong to the learning set. May be `missing`; in that case, the learning set consists of all observations and predictions are made on the learning set.
`...`	Further arguments that shall be passed to function `rsf` from packagerandomSurvivalForest.

An object of class ModelLearned.

Christoph Bernau bernau@ibe.med.uni-muenchen.de

Ishwaran, H., et al. (2008) Random Survival Forests, The Annals of Applied Statistics, 2, 841-860

learnSurvival

##load example data
if(require(pensim)){
require(survHD)
require(survHDExtra)
data(beer.exprs,package="pensim")
data(beer.survival,package="pensim")
##create Surv object (y)
beerY <- with(beer.survival,Surv(os,status))
##create mini X matrix (only using first 200 genes)
beerX <- t(as.matrix(beer.exprs))
beerX <- beerX[,1:200]
colnames(beerX) <- make.names(colnames(beerX),unique=TRUE)
##scale everything for testing.
beerX <- scale(beerX)
##define training and test sets:
set.seed(4)
allind <- 1:nrow(beerX)
learnind <- sample(allind,size=round(length(allind)*9/10))
testind <- allind[-learnind]
##fit a model with ridge regression, and with L2 penalty equal to 100:
fit.rsf <- customSurv(X=beerX,y=beerY, learnind=learnind, ntree=50,customSurvModel=customRSF)
##linear risk score
##survival probabilities at each uncensored time
mytimegrid <- sort(c(0.0,beerY[beerY[,2]==1,1]))
SurvivalProbs <- predict(fit.rsf,beerX[testind,],type='SurvivalProbs',timegrid=mytimegrid,gbm=FALSE)@SurvivalProbs
plot(SurvivalProbs,xlab="Time", ylab="Probability of Survival")
}