How to extract auc, specificity and sensitivity

I am running my code in a loop and it does not work but when I run it
outside the loop I get the values I want.

n <- 1000; # Sample size

fitglm <- function(sigma,tau){
   x <- rnorm(n,0,sigma)
   intercept <- 0
   beta <- 0
   ystar <- intercept+beta*x
   z <- rbinom(n,1,plogis(ystar))
   xerr <- x + rnorm(n,0,tau)
   model<-glm(z ~ xerr, family=binomial(logit))
   int<-coef(model)[1]
   slope<-coef(model)[2]
   pred<-predict(model)

   result<-ifelse(pred>.5,1,0)

   accuracy<-length(which(result==z))/length(z)
   accuracy

   rocpreds<-prediction(result,z)
   auc<-performance(rocpreds,"auc")@y.values
   sentiv<-performance(rocpreds,"sens")@y.values
   sentiv<-slot(fp,"y.values")[[1]]
   sentiv<-sentiv[2]
   sentiv
   specs<-performance(rocpreds,"spec")@y.values
   specs<-slot(fp2,"y.values")[[1]]
   specs<-specs[2]
   specs
   output<-c(int,slope,.5,accuracy,auc,sentiv,specs)

names(output)<-c("Intercept","Slope","CutPoint","Accuracy","AUC","Sentivity",Specificity")

   return(output)

}

y<-fitglm(.05,1)
y

The code runs without the sentiv and specs but when I remove the loop i can
get the sensitivity and spec. values ???

	[[alternative HTML version deleted]]