I have a matrix A[400x100] . How could I get the average of each 20 rows in each columns. Finally, I put all the values into matrix B[20,100]. For example :
B[1,1] = mean(A[1:20])
B[2,1] = mean(A[21:40]) .....
Asked
Active
Viewed 107 times
-1
-
What did you try? Can you give a [Minimal, Complete, and Verifiable Example](http://stackoverflow.com/help/mcve)? – symbolrush Aug 11 '16 at 06:42
-
1You could use `aggregate(A, by = list(rep(1:20, each = 20)), mean)` – talat Aug 11 '16 at 06:59
1 Answers
1
We can do
i1 <- as.numeric(gl(nrow(A), 20, nrow(A)))
B <- t(sapply(split(seq_len(nrow(A)), i1), function(i) colMeans(A[i,])))
dim(B)
#[1] 20 100
data
A <- matrix(1:40000, nrow = 400, ncol= 100)
akrun
- 874,273
- 37
- 540
- 662