Populating a Matrix by doing Calculations along the Diagonal-CodePudding

I have a matrix n*n with values along its diagonal. I would like to create a new matrix by multiplying elements of the diagonal matrix b, corresponding to the position of the elements along the diagonal.

For example, say I have a 4*4 matrix:

set.seed(5)
b <- diag(sample(4))
colnames(b) <- rownames(b) <- c("AAA" , "BBB" , "CCC" , "DDD")

Giving matrix b:

    AAA BBB CCC DDD
AAA   2   0   0   0
BBB   0   3   0   0
CCC   0   0   1   0
DDD   0   0   0   4

I would like to compute matrix d by multiplying as follows:

b[AAA , AAA] = 2 * 2 = 4

and

b[AAA , BBB] = 2 * 3 = 6

The end result should give:

    AAA BBB CCC DDD
AAA   4   6   2   8
BBB   6   9   3  12
CCC   2   3   1   4
DDD   8  12   4  16

Is there an efficient code that can do this without having to do this by hand? This would not be possible if the matrix is very large.

CodePudding user response：

What you are describing is the outer product of the diagonal. You can get this by doing:

outer(diag(b), diag(b))
    AAA BBB CCC DDD
AAA   4   6   2   8
BBB   6   9   3  12
CCC   2   3   1   4
DDD   8  12   4  16

 diag(b) %o% diag(b)
    AAA BBB CCC DDD
AAA   4   6   2   8
BBB   6   9   3  12
CCC   2   3   1   4
DDD   8  12   4  16