R help - Jan 2008 - q about efficient cross-classification/creating tables

Dear R Users,
I am posting a smaller portion of a question I posted last week. Code I have
written
takes weeks to run, and I am trying to make all the lines as efficient as
possible. Here is
one part of the code.

I have some data that looks like

     [,1] [,2] [,3] [,4]
[1,]    1    3    1    3
[2,]    3    3    2    3
[3,]    1    3    3    3
and a binary response. I am trying to obtain a vector of counts for all
cross classifications of the response
and each pair of columns from the data.

Have I coded this in an efficient manner?

library(combinat)
# some example data
myData <- matrix(sample(c(1:3),500,replace=TRUE),nrow=100,ncol=4)
response <- c(rep(1,50),rep(0,50))
#initialize result matrix
count.matrix <- matrix(-1,ncol=6,nrow=18);
# indices of all pairs of columns
all.pairs <- combn2(1:ncol(myData))

getCounts <- function(index)
{
  counts <- as.vector(table(myData[,index[1]],myData[,index[2]],response));

}

count.matrix <- apply(all.pairs, 1, getCounts)


Thanks,

Juliet

	[[alternative HTML version deleted]]