Sampling and replacing random elements of a vector, conditionally - r

Question

Suppose I have a vector containing data:
c <- c(1:100)
c[1:75] <- 0
c[76:100] <- 1
What I need to do is select a number of the 0's and turn them into 1's. There are potentially many ways to do this - like if I'm switching 25 of the 0's, it'd be 75 choose 25, so 5.26x10^19 - so I need do it, say, 1000 times randomly. (this is part of a larger model. I'll be using the mean of the results.)
I know (think), that I need to use sample() and a for loop - but how do I select n values randomly among the 0's, then change them to 1's?

Answer 1

vec <- c(rep(0, 75), rep(1, 25))
n <- 25
to_change <- sample(which(vec == 0), n)
modified_vec <- vec
modified_vec[to_change] <- 1
Something like this. You could wrap it up in a function.
And you should really do it in a matrix with apply, rather than a for loop.
This small example is easy to see it work:
n_vecs <- 5
vec_length <- 10
n_0 <- 7 # Number of 0's at the start of each vector
vec_mat <- matrix(c(rep(0, n_vecs * n_0), rep(1, n_vecs * (vec_length - n_0))),
nrow = vec_length, ncol = n_vecs, byrow = T)
> vec_mat
[,1] [,2] [,3] [,4] [,5]
[1,] 0 0 0 0 0
[2,] 0 0 0 0 0
[3,] 0 0 0 0 0
[4,] 0 0 0 0 0
[5,] 0 0 0 0 0
[6,] 0 0 0 0 0
[7,] 0 0 0 0 0
[8,] 1 1 1 1 1
[9,] 1 1 1 1 1
[10,] 1 1 1 1 1
change_n_0 <- function(x, n) {
x_change <- sample(which(x == 0), n)
x[x_change] <- 1
return(x)
}
vec_mat <- apply(vec_mat, MARGIN = 2, FUN = change_n_0, n = 2)
> vec_mat
[,1] [,2] [,3] [,4] [,5]
[1,] 1 1 0 0 1
[2,] 0 0 0 1 0
[3,] 0 0 0 0 0
[4,] 0 0 0 0 0
[5,] 0 0 1 0 1
[6,] 0 1 0 1 0
[7,] 1 0 1 0 0
[8,] 1 1 1 1 1
[9,] 1 1 1 1 1
[10,] 1 1 1 1 1
You can scale up the constants at the beginning as big as you'd like.