|
|
|
|
|
|
by cscheid
705 days ago
|
|
|
It’s a well known exercise in prob textbooks (edit: it’s the algo referenced in the other reply) to convert one distribution to another. If you can generate gaussians (or any other distribution) you can generate uniform variates. It’s a very simple application of rejection sampling that involves some efficiency loss, but that’s irrelevant at the time you’re getting your OTPs. |
|
|