So I was trying to figure out a fast way to make matrices with randomly allocated 0 or 1 in each cell of the matrix. I reached out on Twitter, and got many responses (thanks tweeps!).

Here is the solution I came up with.

See if you can tell why it would be slow.

1
2
mm <- matrix(0, 10, 5)
apply(mm, c(1, 2), function(x) sample(c(0, 1), 1))
      [,1] [,2] [,3] [,4] [,5]
 [1,]    1    0    1    0    1
 [2,]    0    0    1    1    1
 [3,]    0    0    0    0    1
 [4,]    0    1    1    0    1
 [5,]    0    1    1    1    1
 [6,]    1    0    1    1    1
 [7,]    0    1    0    1    0
 [8,]    0    0    1    0    1
 [9,]    1    0    1    1    1
[10,]    1    0    0    1    1

Ted Hart (@distribecology) replied first with:

1
matrix(rbinom(10 * 5, 1, 0.5), ncol = 5, nrow = 10)
      [,1] [,2] [,3] [,4] [,5]
 [1,]    1    1    0    1    1
 [2,]    1    0    0    1    0
 [3,]    0    1    0    0    0
 [4,]    0    0    1    0    0
 [5,]    1    0    1    0    0
 [6,]    0    0    0    0    1
 [7,]    1    0    0    0    0
 [8,]    0    1    0    1    0
 [9,]    1    1    1    1    0
[10,]    0    1    1    0    0

Next, David Smith (@revodavid) and Rafael Maia (@hylospar) came up with about the same solution.

1
2
3
m <- 10
n <- 5
matrix(sample(0:1, m * n, replace = TRUE), m, n)
      [,1] [,2] [,3] [,4] [,5]
 [1,]    0    0    0    0    1
 [2,]    0    0    0    0    0
 [3,]    0    1    1    0    1
 [4,]    1    0    0    1    0
 [5,]    0    0    0    0    1
 [6,]    1    0    1    1    1
 [7,]    1    1    1    1    0
 [8,]    0    0    0    1    1
 [9,]    1    0    0    0    1
[10,]    0    1    0    1    1

Then there was the solution by Luis Apiolaza (@zentree).

1
2
3
m <- 10
n <- 5
round(matrix(runif(m * n), m, n))
      [,1] [,2] [,3] [,4] [,5]
 [1,]    0    1    1    0    0
 [2,]    1    0    1    1    0
 [3,]    1    0    1    0    0
 [4,]    1    0    0    0    1
 [5,]    1    0    1    1    0
 [6,]    1    0    0    0    0
 [7,]    1    0    0    0    0
 [8,]    1    1    1    0    0
 [9,]    0    0    0    0    1
[10,]    1    0    0    1    1

Last, a solution was proposed using RcppArmadillo, but I couldn’t get it to work on my machine, but here is the function anyway if someone can.

1
2
3
library(inline)
library(RcppArmadillo)
f <- cxxfunction(body = "return wrap(arma::randu(5,10));", plugin = "RcppArmadillo")

And here is the comparison of system.time for each solution.

1
2
3
4
5
mm <- matrix(0, 10, 5)
m <- 10
n <- 5

system.time(replicate(1000, apply(mm, c(1, 2), function(x) sample(c(0, 1), 1))))  # @recology_
   user  system elapsed
  0.470   0.002   0.471
1
system.time(replicate(1000, matrix(rbinom(10 * 5, 1, 0.5), ncol = 5, nrow = 10)))  # @distribecology
   user  system elapsed
  0.014   0.000   0.015
1
system.time(replicate(1000, matrix(sample(0:1, m * n, replace = TRUE), m, n)))  # @revodavid & @hylospar
   user  system elapsed
  0.015   0.000   0.014
1
system.time(replicate(1000, round(matrix(runif(m * n), m, n)), ))  # @zentree
   user  system elapsed
  0.014   0.000   0.014

If you want to take the time to learn C++ or already know it, the RcppArmadillo option would likely be the fastest, but I think (IMO) for many scientists, especially ecologists, we probably don’t already know C++, so will stick to the next fastest options.

Get the .Rmd file used to create this post at my github account.

Written in Markdown, with help from knitr, and nice knitr highlighting/etc. in in RStudio.