[R] generate 3 distinct random samples without replacement

[Sarah Goslee]

Cesar, your indexing is wrong:

On Mon, Mar 7, 2011 at 2:17 PM, Cesar Hincapié
<cesar.hincapie at utoronto.ca> wrote:
> Hello:
>
> I wonder if I could get a little help with random sampling in R.
>
> I have a vector of length 7375.  I would like to draw 3 distinct random samples, each of length 100 without replacement.  I have tried the following:
>
> d1 <- 1:7375
>
> set.seed(7)
> i <- sample(d1, 100, replace=F)
> s1 <- sort(d1[i])
> s1

d1 is a continuous vector of integers, 1 thru 7375 and of length 7375

> d2 <- d1[-i]

but you've taken out 100 of those numbers, so d2 is now of length
7275 and has gaps in the sequence.

> set.seed(77)
> j <- sample(d2, 100, replace=F)
> s2 <- sort(d2[j])
> s2

j is a sample *of the values* and those values are no longer the
indices of the vector d2

You need instead
j <- sample(1:length(d2), 100, replace=FALSE)
s2 <- sort(d2[j])

Some of the value in j no longer exist in d2 as indices. 7375 could be
selected, but since d2 only has 7275 elements d2[7375] doesn't
return anything (actually NA).

Same for your third sample, only the indices are even less like the
elements of the vector because you've removed another random
set of values.

Sarah


-- 
Sarah Goslee
http://www.functionaldiversity.org