R help - Mar 2011 - In ppls package kernel method is unsupported?

require(ppls)
data(BOD)
X<-BOD[,1]
y<-BOD[,2]
Xtest=seq(min(X),max(X),length=200)
dummy<-X2s(X,Xtest,deg=3,nknot=20)
Z<-dummy$Z
Ztest<-dummy$Ztest
size<-dummy$sizeZ
P<-Penalty.matrix(size,order=2)
lambda<-200
number.comp<-3

penalized.pls(Z,y,P=lambda*P,ncomp=number.comp)$coefficients # By default
kernel=F
penalized.pls(Z,y,P=lambda*P,ncomp=number.comp,kernel=TRUE)$coefficients #
Same as above...?!
penalized.pls.kernel(Z,y,M=lambda*P,ncomp=number.comp) # But using directly,
coefficients are different.

-- 
Mi³ego dnia

	[[alternative HTML version deleted]]

W dniu 22 marca 2011 10:27 u¿ytkownik £ukasz Rêc³awowicz <
lukasz.reclawowicz@gmail.com> napisa³:
>
> penalized.pls.kernel(Z,y,M=lambda*P,ncomp=number.comp) # But using
> directly, coefficients are different.
>
I see me error P=!M, but still results are the same...

 p <- ncol(Z)
Minv <- diag(p) + P
M <- solve(Minv)
penalized.pls.default(Z,y,M=M,ncomp=number.comp)
penalized.pls.kernel(Z,y,M=M,ncomp=number.comp)



-- 
Mi³ego dnia

	[[alternative HTML version deleted]]