Hello,
I use the following function "bootstrapge" to calculate (and compare) the
generalization error of several bootstrap implementations:
##
## Calculates and returns a coefficient corresponding to the generalization
## error. The formula for the bootstrap generalization error is:
## $N^{-1}\sum_{i=1}^n B^{-1}\sum_{j=1}^B |y_i - (\beta_n^{*j})^T x|$
##
## x - mxn matrix where m is the number of model parameters and n is the
## number of observations
## y - n column-vector containing true values
## theta_star - mxn matrix where m is the number of bootstrapped samples
## and n is the number of model parameters
##
bootstrapge <- function(x,y,theta_star) {
B <- nrow(theta_star)
P <- ncol(theta_star)
t <- 0
for (b in 1:B) {
t <- t + abs(y - rbind(theta_star[b,])%*%x)
}
return(mean(t/B))
}
Is there a nicer/faster way to accomplish the same using implicit loop
functions e.g. apply, sapply etc I could not figure it out ...
Is there a way to get a similar coefficient using the boot library? I could not
find any way to get such a "generalization error" so I can compare my
implementation with that one ...
TIA,
Best regards,
Giovanni
[[alternative HTML version deleted]]
______________________________________________
[email protected] 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.
