You use the sample function to take a random sample
from a vector. This procedure has two basic parameters: the vector
to sample from and the number of samples to take.
sample1 = sample(values, 30)
Traditionally, there is one other decision you make when doing random
sampling: In a sample without replacmeent , once
you've used a value from the vector in the sample, you cannot use
it again. Such techniques are appropriate for things like choosing
pennies from a jar. In a sample with replacement,
you can select the same value multiple times. You might use such a
technique if you are sampling from a vector that represents a larger
population. By default, R does sampling without replacement. To have
it ample with replacment, you add replace=T.
> values = seq(from=1,to=10,step=1)
> values
[1] 1 2 3 4 5 6 7 8 9 10
> sample(values,5)
[1] 1 8 9 4 2
> sample(values,5)
[1] 5 2 3 6 8
> sample(values,5,replace=T)
[1] 1 10 8 10 5
> sample(values,5,replace=T)
[1] 10 5 7 4 7
> sample(values,5,replace=T)
[1] 9 4 9 9 6
> sample(values,5,replace=T)
[1] 2 2 2 5 10
> sample(values,5,replace=T)
[1] 9 1 3 6 7
As the example suggests, sampling with replacement can, but need not,
duplicate values.
