[R] Bootstrap

[Weidong Gu]

I didn't see bootstrap steps in your code. One way to modify your codes

for (Ncount in 1:100)
{
b.data<-data[sample(1:nrow(data),replace=T),]
y <-b.data[,1]
x <-b.data[,2]
n <- length(x)
... ### make appropriate changes if needed
}

Weidong Gu

On Wed, Jul 20, 2011 at 6:09 PM, Val <valkremk at gmail.com> wrote:
> Hi all,
>
> I am facing difficulty on  how to use bootstrap sampling and
> below is my example of function.
>
> Read a data , use some functions and  use iteration to find the solution(
> ie, convergence is reached).  I want to use bootstrap approach to do it
> several times (200 or 300 times) this whole process  and see the
> distribution of parameter of interest.
>
> Below is a small example that resembles my problem. However,  I  found out
> all samples are the same. So I would appreciate your help on this case.
>
> #**************************************
> rm(list=ls())
>  xx <- read.table(textConnection(" y x
>    11 5.16
>    11 4.04
>    14 3.85
>    19 5.68
>    4 1.26
>    23  7.89
>    15 4.25
>    17 3.94
>    7 2.35
>    17 4.74
>    14 5.49
>    11 4.12
>    17 5.92"), header=TRUE)
>    data <- as.matrix(xx)
>    closeAllconnections()
>
> Nt <- NULL
> for (Ncount in 1:100)
>  {
>    y <- data[,1]
>    x <- data[,2]
>    n <- length(x)
>
>    X <- cbind(rep(1,n),x)                 #covariate/design matrix
>    obeta<- c(1,1)                         #previous/starting values of beta
>
>    nbeta <- c(0,0)                        #new beta
>    iter=0
>
>  while(crossprod(obeta-nbeta)>10^(-12))
>   {
>    nbeta <- obeta
>    eta   <- X%*%nbeta
>    mu    <- eta
>    mu1   <- 1/eta
>    W     <- diag(as.vector(mu1))
>    Z     <- X%*%nbeta+(y-mu)
>    XWX   <- t(X)%*%W%*%X
>    XWZ   <- t(X)%*%W%*%Z
>    Cov   <- solve(XWX)
>    obeta <- Cov%*%XWZ
>    iter  <- iter+1
>
>    cat("Iteration #  and beta1= ",iter, nbeta, "\n")
>    }
>
>  Nt[Ncount] <- nbeta[1,1]
> }
> Nt
> summary(Nt)
> #**************e*****************************************
>
>        [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>