What is the canonical way to sum a 3D array along dimension 3 (thereby yielding a matrix)?
I know I can apply(A,c(1,2),sum) but (wrongly or rightly) I got the impression from somewhere that using apply is no better than using for loops.
I could probably aperm the array, colSum it, then unaperm it again, but that wouldn't be very readable.
Is there a better way?
Use rowSums. It offers a dims parameter which specifies "[w]hich dimensions are regarded as ‘rows’ or ‘columns’ to sum over. For row*, the sum or mean is over dimensions dims+1".
a <- array(1:16, c(2, 2, 2))
#, , 1
#
# [,1] [,2]
#[1,] 1 3
#[2,] 2 4
#
#, , 2
#
# [,1] [,2]
#[1,] 5 7
#[2,] 6 8
rowSums(a, dims = 2)
# [,1] [,2]
#[1,] 6 10
#[2,] 8 12
