Mark Mark - 1 month ago 17
R Question

R: probability based true/false statements

I have a large number of data points, each with a different associated probability of being 0 or 1. Is there a way in R to create a vector of 0s and 1s sampled from these probabilities? The following code shows how to do what I want, but it's in a loop.

n <- 10
x <- letters[1:n] # my data
p <- runif(n) # my probabilities of the result being 1
result <- numeric()
for(ii in 1:n){
result[ii] <- sample(x = c(0,1), 1, prob = c(1-p[ii], p[ii]))
}
print(cbind(x, p, result))


The result is something like:

x p result
[1,] "a" "0.0407166101504117" "0"
[2,] "b" "0.632280522491783" "1"
[3,] "c" "0.754841333255172" "0"
[4,] "d" "0.0509465073700994" "0"
[5,] "e" "0.629663853673264" "0"
[6,] "f" "0.357108945958316" "0"
[7,] "g" "0.808141406625509" "1"
[8,] "h" "0.671664241468534" "1"
[9,] "i" "0.0218871515244246" "1"
[10,] "j" "0.689538966864347" "1"

Answer

I'm not sure if I understand correctly, because I get confused by the letters... How about this?

set.seed(1) # reproducibility
n <- 10
p <- runif(n) # probabilities
result <- rbinom(n,1,p)

cbind(p,result)
               p result
 [1,] 0.26550866      0
 [2,] 0.37212390      0
 [3,] 0.57285336      0
 [4,] 0.90820779      1
 [5,] 0.20168193      0
 [6,] 0.89838968      1
 [7,] 0.94467527      1
 [8,] 0.66079779      0
 [9,] 0.62911404      1
[10,] 0.06178627      0
Comments