如何在R中创建子矩阵？

Question

I have a matrix "Mat.return" with 390 rows and 2749 columns and I want to create 2499 sub-matrices from it, each with 250 columns and 80 rows.

The first sub-matrix would be:

B1=(Mat.return)[sample(nrow((Mat.return)),size=80,replace=TRUE),][,c(1:250)]

The second one, would start from the second column of "Mat.return" and would select 250 following columns. It would thus be:

B2=(Mat.return)[sample(nrow((Mat.return)),size=80,replace=TRUE),][,c(2:251)]

The third one would start from the third column and would select the 250 following column, and so on [until matrix n°2499]

Is there a function or a code that could do this, instead of computing it manually?

Thank you!

Answer 1

Just make a loop from 1 to 2499 around your function. this code will give you a list of 2499 matrix on 80 rows and 251 columns

Mat.return <- matrix(rnorm(390*2749), nrow = 390, ncol = 2749)

lmat <- lapply(1:2499, function(i){
  (Mat.return)[sample(nrow((Mat.return)),size=80,replace=TRUE),][,c(i:(250 + i))]
})

str(lmat, list.len = 10)
#> List of 2499
#>  $ : num [1:80, 1:251] 0.493 -0.295 2.299 -1.427 -0.174 ...
#>  $ : num [1:80, 1:251] -0.4 -1.632 1.21 0.529 -1.045 ...
#>  $ : num [1:80, 1:251] -1.71 -1.458 0.186 0.808 -1.179 ...
#>  $ : num [1:80, 1:251] 0.237 -0.952 -0.632 -0.204 -1.702 ...
#>  $ : num [1:80, 1:251] -1.828 -0.895 -1.31 1.009 -0.451 ...
#>  $ : num [1:80, 1:251] 0.128 0.461 -0.393 0.358 1.549 ...
#>  $ : num [1:80, 1:251] -0.44814 0.52248 0.28651 0.39365 -0.00774 ...
#>  $ : num [1:80, 1:251] 0.136 0.615 -0.435 -0.846 0.788 ...
#>  $ : num [1:80, 1:251] 0.761 0.11 -1.486 -0.488 0.118 ...
#>  $ : num [1:80, 1:251] -0.9064 -1.3382 -0.9678 0.0654 -0.5952 ...
#>   [list output truncated]