R better way to replace matrix elements with zeroes in symetric matrix

Question

I have a 0-diagonal symmetric matrix X, such as

          [1]        [2]       [3]        [4]        [5]    
[1]  0.00000000 0.07065048 0.1019865 0.23641082 0.23641082 
[2]  0.07065048 0.00000000 0.0000000 0.23641082 0.23641082 
[3]  0.10198654 0.00000000 0.0000000 0.00000000 0.23641082 
[4]  0.23641082 0.23641082 0.0000000 0.00000000 0.08870087 
[5]  0.23641082 0.23641082 0.2364108 0.08870087 0.00000000 

and I need to replace some of its elements with 0's according to a second matrix Y:

      [1]    [2]
[1]    3      2
[2]    4      3
[3]    1      2

So X[3,2], X[4,3], X[1,2] and, for symmetry's sake, X[2,3], X[3,4], X[2,1] should be replaced with zero's.

I came up with this solution:

for (i in 1:nrow(Y)) {
  X[Y[i,1], Y[i,2]] <- 0
  X[Y[i,2], Y[i,1]] <- 0 
}

which is doing the trick, but I was wondering whether there is a better, more efficient way to code this using apply-like functions.

thank you very much for your attention.

Answer 1

This should work:

X[Y] <- 0
X[Y[, 2:1]] <- 0

or in one (balanced) statement:

X[rbind(Y[, 1:2],
        Y[, 2:1])] <- 0

This form of indexing is commonly referred to as matrix indexing. Here is the relevant part from the ?"[" doc:

A third form of indexing is via a numeric matrix with the one column for each dimension: each row of the index matrix then selects a single element of the array, and the result is a vector. [...]

This will be faster than any type of loop, including *apply functions.