Dear Friends,
The smooth.spline implemented in R is to minimize the sum:
sum[ y(i) -g(i) ]^2/n + lamda * integral [g''(x)]^2
and the lamda is a funtion of spar.
But in the text book "Nonparametric Regression and Generalized Linear
Models, A roughness penalty approach", the penalized sum is :
sum[y(i) - g(i)]^2 + alpha *integral[g''(x)]^2 . From
the equations alone, one may think that the alpha should be equal to
lamda* n. But , it is not the case.
Is there anybody know the relationship among the alpha, lamda and the
spar ?
Also, in the subroutine sgram(), only the upper four diagonals are
computed for sgm[0-3](nb). Why don't compute all of them ? Will there
any errors be incurred due to this?
Thanks a lot.
Chuanjun
Alec Stephenson wrote:
>See Programmer's Niche in Volume 1/1 of R News
>
>
>Alec Stephenson
>Department of Statistics
>Macquarie University
>NSW 2109, Australia
>
>
>
>>>><[EMAIL PROTECTED]> 08/01/04 09:40am >>>
>>>>
>>>>
>How best may I form a list containing all choose(n,k) k-element subsets
>of
>a vector x of length n? (and for subsequences?)
>
>Murray Jorgensen
>
>______________________________________________
>[EMAIL PROTECTED] mailing list
>https://www.stat.math.ethz.ch/mailman/listinfo/r-help
>PLEASE do read the posting guide!
>http://www.R-project.org/posting-guide.html
>
>______________________________________________
>[EMAIL PROTECTED] mailing list
>https://www.stat.math.ethz.ch/mailman/listinfo/r-help
>PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
>
>
[[alternative HTML version deleted]]
______________________________________________
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
