Sampling is a popular Spark RDD operation. Sampling With Replacement and sampling without replacement are different ways of doing sampling. This article explains the difference between them.
Sampling with replacement:
Consider a population of potato sacks, each of which has either 12, 13, 14, 15, 16, 17, or 18 potatoes, and all the values are equally likely. Suppose that, in this population, there is exactly one sack with each number. So the whole population has seven sacks. If I sample two with replacement, then I first pick one (say 14). I had a 1/7 probability of choosing that one. Then I replace it. Then I pick another. Every one of them still has 1/7 probability of being chosen. And there are exactly 49 different possibilities here (assuming we distinguish between the first and second.) They are: (12,12), (12,13), (12, 14), (12,15), (12,16), (12,17), (12,18), (13,12), (13,13), (13,14), etc.
Sampling without replacement:
Consider the same population of potato sacks, each of which has either 12, 13, 14, 15, 16, 17, or 18 potatoes, and all the values are equally likely. Suppose that, in this population, there is exactly one sack with each number. So the whole population has seven sacks. If I sample two without replacement, then I first pick one (say 14). I had a 1/7 probability of choosing that one. Then I pick another. At this point, there are only six possibilities: 12, 13, 15, 16, 17, and 18. So there are only 42 different possibilities here (again assuming that we distinguish between the first and the second.) They are: (12,13), (12,14), (12,15), (12,16), (12,17), (12,18), (13,12), (13,14), (13,15), etc.
What’s the Difference?
When we sample with replacement, the two sample values are independent. Practically, this means that what we get on the first one doesn’t affect what we get on the second. Mathematically, this means that the covariance between the two is zero.
In sampling without replacement, the two sample values aren’t independent. Practically, this means that what we got on the for the first one affects what we can get for the second one. Mathematically, this means that the covariance between the two isn’t zero. That complicates the computations. In particular, if we have a SRS (simple random sample) without replacement, from a population with variance 
, then the covariance of two of the different sample values is 
, where N is the population size. (A brief summary of some formulas is provided here. For a discussion of this in a textbook for a course at the level of M378K, see the chapter on Survey Sampling in Mathematical Statistics and Data Analysis by John A. Rice, published by Wadsworth & Brooks/Cole Publishers. There is an outline of an slick, simple, interesting, but indirect, proof in the problems at the end of the chapter.)
, then the covariance of two of the different sample values is , where N is the population size. (A brief summary of some formulas is provided here. For a discussion of this in a textbook for a course at the level of M378K, see the chapter on Survey Sampling in Mathematical Statistics and Data Analysis by John A. Rice, published by Wadsworth & Brooks/Cole Publishers. There is an outline of an slick, simple, interesting, but indirect, proof in the problems at the end of the chapter.)
Population size — Leading to a discussion of “infinite” populations.
When we sample without replacement, and get a non-zero covariance, the covariance depends on the population size. If the population is very large, this covariance is very close to zero. In that case, sampling with replacement isn’t much different from sampling without replacement. In some discussions, people describe this difference as sampling from an infinite population (sampling with replacement) versus sampling from a finite population (without replacement).
Comments
Post a Comment