Aggregate Matrix Columns by Row Names in R

To aggregate matrix columns by row names, we can use colSums with sapply and transpose the output. For example, if we have a matrix called M then the aggregate matrix columns by row names can be done using t(sapply(by(M,rownames(M),colSums),identity)).

Example1

Consider the below matrix −

Live Demo

> M1<-matrix(rpois(40,2),ncol=2)
> rownames(M1)<-sample(c("A","B","C","D"),20,replace=TRUE)
> M1

Output

  [,1] [,2]
B    4    6
D    2    1
B    1    5
C    0    0
A    2    3
B    1    0
B    5    3
D    1    3
C    0    1
C    0    2
D    1    1
B    7    2
C    2    3
C    2    3
C    2    2
C    1    2
B    1    1
C    5    3
D    2    2
D    5    2

Finding the aggregate of M1 by row names −

> t(sapply(by(M1,rownames(M1),colSums),identity))

Output

  V1 V2
A  2  3
B 19 17
C 12 16
D 11  9

Example2

Live Demo

> M2<-matrix(rpois(40,10),ncol=2)
> rownames(M2)<-sample(c("S1","S2","S3"),20,replace=TRUE)
> M2

Output

   [,1] [,2]
S1    6   10
S3    7   11
S3   13    6
S1    9    9
S2   11    9
S3    6    8
S1    4    5
S3    8   12
S2   14    8
S3   11    7
S1   15   12
S3    8   13
S3    9   10
S1   10    8
S2   14   18
S1   13   10
S1   10    4
S3   12    7
S2    5    9
S3   13    8

Finding the aggregate of M2 by row names −

> t(sapply(by(M2,rownames(M2),colSums),identity))

Output

   V1 V2
S1 67 58
S2 44 44
S3 87 82