🚀 go-pugleaf

In article <9203042028.AA05869@essss1.stat.ncsu.edu> nychka@stat.ncsu.edu
(Doug Nychka) writes: 
>
>   Is there an easy way to do Kronecker products in S?
>   Doug Nychka

Here's the slow, for-loop way (I needed it in a hurry, once :-)

"%x%" <-
function(a, b)
{
        if(!is.matrix(a) || !is.matrix(b))
                stop("A and B must be matrices")
        C <- NULL
        for(i in 1:dim(a)[1]) {
                D <- NULL
                for(j in 1:dim(a)[2]) {
                        D <- cbind(D, a[i, j] * b)
                }
                C <- rbind(C, D)
        }
        C
}

After which definition you can just say "A %x% B"

However all the entries you want are in the array "outer(A, B)"
more-or-less in a flash.  There is a then just the small problem of
arranging the four-way array as a two-way array afterwards.

It may even be preferable to think about your application a bit more and
actually use the four-way array, with Bill Dunlap's (?) recently posted
%tensor% function to do the subsequent work.  In a sense the tensor product
"flattening out" of the four-way array is often a bit of a mathematical
mistake anyway.
--
Bill Venables, Dept. of Statistics, | Email: venables@stats.adelaide.edu.au
Univ. of Adelaide, South Australia. | Tel: +61 8 228 5412  Fax: ...232 5670