R help - Jul 2011 - apply family functions (tapply, sapply, mapply etc)

Dear all,

I am trying to use apply or similar functions in order to improve the
efficiency of my code but I guess I am misunderstanding the function of
these commands. What I want to do is possible to see with the following code
that I use a for loop command. In words what I want to do are two things 1)
to multiply each row of a matrix with its transpose (the same row in a
column) for each row the matrix. 

N<-250
x.1<-as.matrix(rnorm(N,0,1))												
x.2<-as.matrix(rnorm(N,1,1))												
x.3<-as.matrix(rnorm(N,1,1))												
x.4<-as.matrix(rnorm(N,0,1))												
X<-cbind(x.1, x.2,x.3,x.4)
q<-ncol(X)-1	
h0<-0.3														
Xu<-cbind(x.1,x.2,x.3)																		
beta.in<-as.matrix(c(1,1,1))

XX<-matrix(0,N*q,q)
Qn<-matrix(0,N*q,q)
DDS<-function(Betas, h, Xs){
	v<-(Xs[,1:q]%*%Betas+Xs[,q+1])/h
	((105/64)*(matrix(1,N,1)-10*v+28*v^3-18*v^5)*(v>-1)*(v<1))}

for (j in 1:N){
	nq<-(j-1)*q+1
	nq1<-j*q
	Qn[nq:nq1,]<-Xu[j,]%*%t(Xu[j,])*(DDS(beta.in, h=h0, Xs=X)[j,])
	XX<-Xu[j,]%*%t(Xu[j,])
}

for (j in 1:N){
		nq<-(j-1)*q+1
		nq1<-j*q
		Q<-0+Qn[nq:nq1,]
}


Another issue that I try to solve is that I want to calculate the result of
let say a function "S" for every different row of the matrix, lets
call it
B, where the rows of B I want to be the argument Betas (that is a vector) of
function "S"


Sn  <-function(Betas, h, Xs){
	v<-(Xs[,1:q]%*%Betas+Xs[,q+1])/h

(2*Y-1)*(0.5+(105/64)*(v-(5/3)*v^3+(7/5)*v^5-(3/7)*v^7))*(v>-1)*(v<1)+matrix(1,N,1)*(v>=1)}
	
S  <-function(Betas, h, Xs){colMeans(Sn(Betas, h, Xs))}

Any help would be highly appreciated

Thanks

Dimitris


--
View this message in context:
http://r.789695.n4.nabble.com/apply-family-functions-tapply-sapply-mapply-etc-tp3652276p3652276.html
Sent from the R help mailing list archive at Nabble.com.

On 07.07.2011 20:30, Dimitris.Kapetanakis wrote:
> Dear all,
>
> I am trying to use apply or similar functions in order to improve the
> efficiency of my code but I guess I am misunderstanding the function of
> these commands. What I want to do is possible to see with the following
code
> that I use a for loop command. In words what I want to do are two things 1)
> to multiply each row of a matrix with its transpose (the same row in a
> column) for each row the matrix.
This should be

diag(crossprod(t(X)))

>
> N<-250
> x.1<-as.matrix(rnorm(N,0,1))												
> x.2<-as.matrix(rnorm(N,1,1))												
> x.3<-as.matrix(rnorm(N,1,1))												
> x.4<-as.matrix(rnorm(N,0,1))												
> X<-cbind(x.1, x.2,x.3,x.4)
> q<-ncol(X)-1	
> h0<-0.3														
> Xu<-cbind(x.1,x.2,x.3)																		
> beta.in<-as.matrix(c(1,1,1))
>
> XX<-matrix(0,N*q,q)
> Qn<-matrix(0,N*q,q)
> DDS<-function(Betas, h, Xs){
> 	v<-(Xs[,1:q]%*%Betas+Xs[,q+1])/h
> 	((105/64)*(matrix(1,N,1)-10*v+28*v^3-18*v^5)*(v>-1)*(v<1))}
>
> for (j in 1:N){
> 	nq<-(j-1)*q+1
> 	nq1<-j*q
> 	Qn[nq:nq1,]<-Xu[j,]%*%t(Xu[j,])*(DDS(beta.in, h=h0, Xs=X)[j,])
> 	XX<-Xu[j,]%*%t(Xu[j,])
> }
>
> for (j in 1:N){
> 		nq<-(j-1)*q+1
> 		nq1<-j*q
> 		Q<-0+Qn[nq:nq1,]
> }
>
>
> Another issue that I try to solve is that I want to calculate the result of
> let say a function "S" for every different row of the matrix,
lets call it
> B, where the rows of B I want to be the argument Betas (that is a vector)
of
> function "S"
>
>
> Sn<-function(Betas, h, Xs){
> 	v<-(Xs[,1:q]%*%Betas+Xs[,q+1])/h
>
>
(2*Y-1)*(0.5+(105/64)*(v-(5/3)*v^3+(7/5)*v^5-(3/7)*v^7))*(v>-1)*(v<1)+matrix(1,N,1)*(v>=1)}
> 	
> S<-function(Betas, h, Xs){colMeans(Sn(Betas, h, Xs))}

apply(B, 1, S, h=h, Xs=Xs)

Uwe Ligges



> Any help would be highly appreciated
>
> Thanks
>
> Dimitris
>
>
> --
> View this message in context:
http://r.789695.n4.nabble.com/apply-family-functions-tapply-sapply-mapply-etc-tp3652276p3652276.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> 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.