CODEWITHSUNDEEP
rmatrixsum
Chris T.
Chris T.

Reputation: 1801

Summing and collapsing multiple column values in a matrix into a new matrix with fewer columns (with R)

I want to sum multiple column values (in a matrix) and collapse them into a new matrix with fewer columns but maintain the same number of rows as in the original matrix. Below is my example:

I want to sum the values of column 1 and 3 and collapse them into a new column (say column n1), sum the values of column 2 and 4 and collapse them into a new column (say column n2), sum the values of column 5 and 6 and collapse them into a new column (say column n3):

mat1 <- matrix(c(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18), nrow = 3, ncol = 6, byrow = TRUE)

 [,1] [,2] [,3] [,4] [,5] [,6]
[1,]    1    2    3    4    5    6
[2,]    7    8    9   10   11   12
[3,]   13   14   15   16   17   18

So the desired output will be

     [,1] [,2] [,3]
[1,]    4   16   28
[2,]    6   18   30
[3,]   11   23   34

Any way I can efficiently achieve this? Thanks.

Upvotes: 1

Views: 151

Answers (2)

GKi
GKi

Reputation: 39667

You can store the paired rows in an index and use this for subseting when calculating rowSums:

idx <- matrix(c(1,3,2,4,5,6), 2)
t(apply(idx, 2, function(i) rowSums(mat1[,i])))
#     [,1] [,2] [,3]
#[1,]    4   16   28
#[2,]    6   18   30
#[3,]   11   23   35

In case you want to select not only 2 columns you can store the indices in a list.

idx <- list(c(1,3), c(2,4), c(5,6))
t(sapply(idx, function(i) rowSums(mat1[,i])))

Upvotes: 4

Ronak Shah
Ronak Shah

Reputation: 389047

If there is no sequence or logic involved in selection of columns you need to add them manually and cbind :

cbind(mat1[, 1] + mat1[, 3], mat1[, 2] + mat1[, 4], mat1[, 5] + mat1[, 6])

Upvotes: 2

Related Questions