Apply sample() across columns of a multidimensional array

Question

Given the following array in R:

x <- 1:24
dim(x) <- c(2, 4, 3)

Which gives:

> x
, , 1

     [,1] [,2] [,3] [,4]
[1,]    1    3    5    7
[2,]    2    4    6    8

, , 2

     [,1] [,2] [,3] [,4]
[1,]    9   11   13   15
[2,]   10   12   14   16

, , 3

     [,1] [,2] [,3] [,4]
[1,]   17   19   21   23
[2,]   18   20   22   24

How to apply the sample() function to permute the columns across the matrices ? For instance in column 1, the value 1, 2, 9, 10 and 17, 18 would be permuted across the 3 matrices while staying in column 1.

One sampling could be:

> x
, , 1

     [,1] [,2] [,3] [,4]
[1,]    10    11    5    23
[2,]    2    19    22    8

, , 2

     [,1] [,2] [,3] [,4]
[1,]    1   3   6   24
[2,]   17   12   14   16

, , 3

     [,1] [,2] [,3] [,4]
[1,]   9   4   21   7
[2,]   18   20   13   15

jay.sf · Accepted Answer · 2024-06-28 09:40:25Z

2

It's easy when we transpose.

> set.seed(666)
> aperm(aperm(x, c(1, 3, 2))[sample.int(dim(x)[1]), sample.int(dim(x)[3]), ], c(1, 3, 2))
, , 1

     [,1] [,2] [,3] [,4]
[1,]   18   20   22   24
[2,]   17   19   21   23

, , 2

     [,1] [,2] [,3] [,4]
[1,]    2    4    6    8
[2,]    1    3    5    7

, , 3

     [,1] [,2] [,3] [,4]
[1,]   10   12   14   16
[2,]    9   11   13   15

edited Jun 28, 2024 at 9:40

answered Jun 28, 2024 at 9:09

jay.sf

76.3k8 gold badges66 silver badges132 bronze badges

Sign up to request clarification or add additional context in comments.

6 Comments

dputhier Over a year ago

Sorry if it was unclear but here 7 and 8 values are now in column 2 while they were initially in column 4. They should stay in column 4 but be dispatched in x[,,1], x[,,2] or x[,,3].

jay.sf Over a year ago

@dputhier See update.

dputhier Over a year ago

Not that readable but thanks for posting this solution.

jay.sf Over a year ago

@dputhier You wanted an efficient solution.

dputhier Over a year ago

That's it. Thanks

|

dputhier · Accepted Answer · 2024-06-28 09:12:39Z

1

I found this solution with a for loop but there is probably better and more efficient with large arrays (?)

     for(i in 1:ncol(x)){x[,i,] <- sample(x[,i,])}

edited Jun 28, 2024 at 9:12

answered Jun 28, 2024 at 9:04

dputhier

8041 gold badge9 silver badges23 bronze badges

Collectives™ on Stack Overflow

Apply sample() across columns of a multidimensional array

2 Answers 2

6 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

6 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related