Analysis of a large sample, first of all, I generate a random sequence, and then the overall ranking data to generate the random Numbers are ascending, and then I draw in the top 200 data, the reasonable sampling like this?
PS: every time the sample size is different, but the fixed is from 200