Displaying 1 result from an estimated 1 matches for "confidenceset".
Did you mean:
confidences
2012 Dec 06
1
bootstrap based confidence band
...,Y))
B=1000
Boot.Result=matrix(nrow=B,ncol=n)
for ( b in 1:B){
Data.Orig.Boot=PairedBootstrap(Data.Orig)
fit.Boot=predict(fit,newdata=Data.Orig.Boot,type="response")
Boot.Result[b,]=fit.Boot
}
And try to find 95% confidence interval for 1000 copies of y corresponding to each x.
ConfidenceSet.Pointwise=function(Boot.Result,alpha){
n=ncol(Boot.Result)
B=nrow(Boot.Result)
SetBounds=matrix(ncol=2,nrow=n)
for(j in 1:n){
Result.Sort=sort(Boot.Result[,j])
SetBounds[j,1]=Result.Sort[floor(0.5*B*alpha)]
SetBounds[j,2]=Result.Sort[ceiling(B*(1-0.5*alpha))]
}
retu...