Reputation: 567
I wrote this function which returns the probability that the value 1 does not appear in a random sample when iterated 1000 times for varying sample sizes.
bday.function <- function(sample.size){
x <- vector()
for (i in 1:1000){
x[i] <- !any(data.frame(table(sample(1:365, sample.size, replace=TRUE)))$Var1 == 1)
}
return(mean(x))
}
Now I want to use this function and another for-loop to calculate the probability for every sample size between 500 and 1500 and make a simple scatter plot of my results. Here is what I tried:
z <- vector()
for (i in 500:1500) {
z[i] <- bday.function(i)
return(plot(z))
}
Edit: when I run bday.function the output is number of TRUE values divided by the total (1000) TRUE/FALSE outcomes:
bday.function(750)
[1] 0.122
I would like to replicate this for sample sizes between 500 and 1500 to generate a simple scatter plot
Edit 2: Thanks to everybody for the help! Here's my final solution:
x <- vector(length = 1000)
for (i in 1:1000){
x[i] <- !any(sample(1:365, 500, replace=TRUE) == 1)
}
x
bday.function <- function(sample.size){
x <- vector(length= 1000)
for (i in 1:1000){
x[i] <- !any(sample(1:365, sample.size, replace=TRUE) == 1)
}
return(mean(x))
}
bday.function(750)
z <- vector(length = 1000)
tmp.index <- 500:1500
for (i in seq_along(tmp.index)) {
z[i] <- bday.function(tmp.index[i])
}
#Plot
plot(tmp.index, z, xlab = "sample size", ylab = "Probability of no birthdays")
Upvotes: 0
Views: 1162
Reputation: 174586
Are you looking for something like this?
bday.function <- function(sample.size) {
mean(sapply(seq(1000), function(x)
+!any(sample(365, sample.size, replace = TRUE) == 1)))
}
x <- 500:1500
y <- sapply(x, bday.function)
plot(x, y, xlab = "sample size", ylab = "Probability of no birthdays")
Upvotes: 1
Reputation: 39613
As @JohnColeman pointed in his sage comment, your function can be slow. Try these changes on your code for the printing output. I have run only 60 sims as I need to complete other things:
#Function
bday.function <- function(sample.size){
x <- vector()
for (i in 1:1000){
x[i] <- !any(data.frame(table(sample(1:365, sample.size, replace=TRUE)))$Var1 == 1)
}
return(mean(x))
}
#Loop
z <- vector()
vec <- 500:1500
for (i in seq_along(vec)) {
z[i] <- bday.function(vec[i])
}
#Plot
plot(z)
Output:
Upvotes: 1