R - Sum list of matrix with different columns - r

Question

I have a large list of matrix with different columns and I would like to sum these matrix counting 0 if column X does not exist in one matrix.
If you have used the function rbind.fill from plyr I would like something similar but with sum function. Of course I could build a function to do that, but I'm thinking about a native function efficiently programmed in Frotrain or C due to my large data.
Here an example:
This is the easy example where I have the same columns:
aa <- list(
m1 = matrix(c(1,2,3,4,5,6,7,8,9), nrow = 3, dimnames = list(c(1,2,3),c('a','b','c'))),
m2 = matrix(c(1,2,3,4,5,6,7,8,9), nrow = 3, dimnames = list(c(1,2,3),c('a','b','c')))
)
aa
Reduce('+',aa)
Giving the results:
> aa
$m1
a b c
1 1 4 7
2 2 5 8
3 3 6 9
$m2
a b c
1 1 4 7
2 2 5 8
3 3 6 9
> Reduce('+',aa)
a b c
1 2 8 14
2 4 10 16
3 6 12 18
And with my data:
bb <- list(
m1 = matrix(c(1,2,3,7,8,9), nrow = 3, dimnames = list(c(1,2,3),c('a','c'))),
m2 = matrix(c(1,2,3,4,5,6,7,8,9), nrow = 3, dimnames = list(c(1,2,3),c('a','b','c')))
)
bb
Reduce('+',bb)
Here I would like to have b = c(0,0,0) in the first matrix to sum them.
> bb
$m1
a c
1 1 7
2 2 8
3 3 9
$m2
a b c
1 1 4 7
2 2 5 8
3 3 6 9
Many thanks!
Xevi

Answer 1

One option would be
un1 <- sort(unique(unlist(lapply(bb, colnames))))
bb1 <- lapply(bb, function(x) {
nm1 <- setdiff(un1, colnames(x))
m1 <- matrix(0, nrow = nrow(x), ncol = length(nm1), dimnames = list(NULL, nm1))
cbind(x, m1)[, un1]})
and use the Reduce
Reduce(`+`, bb1)
# a b c
# 1 2 4 14
# 2 4 5 16
# 3 6 6 18