##' Prediction with new data and return a saved forest brier score function
##' This function has some problems
##' @export
rsfes.bs_predict <-
function(rsfesfit,testdat,rii,trlength=500){
trees=rsfesfit$pectrees
colindexes=rsfesfit$colindexes
newindexes=rsfesfit$newindexes
newdata=testdat[,-c(rii)]
if (trlength>length(rsfesfit$pectrees))
stop("Number of Trees for prediction should not be more than Number of Trees Fitted")
# classify the test data
testpre<-NULL
for (i in 1:trlength) {
#if (oobacc[i]<=avroobacc)
{
# preparing for testing
if (ncol(newdata)<=100){
testdata=extspace_testdat(newdata,newindexes[[i]])
testdata=testdata[,colindexes[[i]]]
}else{
testdata=newdata[,colindexes[[i]]]
testdata=extspace_testdat(testdata,newindexes[[i]])
}
newtestdat=cbind.data.frame(testdat[,c(rii)],testdata)
pecerror <- pec(list("rsfse"=trees[[i]]),formula=Surv(time,status)~., data=newtestdat,cens.model = "marginal",reference = FALSE)
print((pecerror))
pecerror$AppErr$rsfse[is.na(pecerror$AppErr$rsfse)]=0
predicts=crps(pecerror)[1]
print(crps(pecerror))
testpre<-cbind(predicts,testpre)
}
}
ensemble_predictions<-rowMeans(testpre)
return(ensemble_predictions)
}
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.