The beauty of trial and error ... if I leave the non x, y parameters i.e. h as
global parameters rather than formal parameters for gaussiankernel it works
fine basically I don't pass anymore h=0.5 to gaussiankernel but consume it from
a global variable. Ugly but works ...
Best regards,
Giovanni
On Apr 26, 2010, at 1:38 AM, Giovanni Azua wrote:
> Hello,
>
> I have the following function that receives a "function pointer" formal
> parameter name "fnc":
>
> loocv <- function(data, fnc) {
> n <- length(data.x)
> score <- 0
> for (i in 1:n) {
> x_i <- data.x[-i]
> y_i <- data.y[-i]
> yhat <- fnc(x=x_i,y=y_i)
> score <- score + (y_i - yhat)^2
> }
> score <- score/n
> return(score)
> }
>
> I would like to use it like this:
>
> ##
> ## Estimator function using Gaussian Kernel
> ##
> gaussiankernel <- function(x,y,h) {
> modelks <- ksmooth(x,y,kernel="normal",bandwidth=h,x.points=x)
> yhat <- modelks$y
> return(yhat)
> }
>
> scoreks <- loocv(data,gaussiankernel(h=0.5))
>
> I expected this to work but it doesn't :( basically I wanted to take
> advantage of the named parameters so I could pass the partially specified
> function parameter "gaussiankernel" to loocv specifying only the h parameter
> and then let loocv specify the remaining parameters as needed ... can this be
> tweaked to work? The idea is to have loocv generic so it can work for any
> estimator implementation ...
>
> I have more than 6 books now in R and none explains this important concept.
>
> Thanks in advance,
> Best regards,
> Giovanni
>
[[alternative HTML version deleted]]
______________________________________________
R-help@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.
