Sep-19-2019, 03:39 PM
Let me make a wild guess that the issue is in combinatorics. I think.
The objection may be that the probabilities of the numbers being selected varies as the selections are made. Example - select 10 numbers in the range 1 to 20. The first number selected had a probability of being selected of 1 in 20. However, when you get to the last number selected, that number had a probability of being selected of 1 in 11. So, the probabilities vary.
For this problem where it appears you want 25 unique out of [1:100) you could grab 25 numbers, check to see if duplicates, if so then reject that panel and pull another 25 numbers. That would keep the probabilities equal.
The objection may be that the probabilities of the numbers being selected varies as the selections are made. Example - select 10 numbers in the range 1 to 20. The first number selected had a probability of being selected of 1 in 20. However, when you get to the last number selected, that number had a probability of being selected of 1 in 11. So, the probabilities vary.
For this problem where it appears you want 25 unique out of [1:100) you could grab 25 numbers, check to see if duplicates, if so then reject that panel and pull another 25 numbers. That would keep the probabilities equal.