# Talk:Statistics/Normal distribution

## Dupe with extra work[edit]

Random numbers turns a uniform distribution into a normal one, but it doesn't have all the crazy extra stats. This seems like a duplicate. I think the extra work is just noise to prove the formula, but it's already been proven, so it's not really necessary. --Mwn3d 01:33, 27 July 2011 (UTC)

- I know. There's a whole school of techniques to produce different distributions from a uniform one, I was thinking about how to organize them into a topic, and normal dist seemed to be the most obvious one. It's difficult to keep it both simple and useful. Maybe should scrap it for now, I'll work with the sandbox first. --Ledrug 02:44, 27 July 2011 (UTC)
- Perhaps a Poisson distribution would be reasonable as an alternative? It would also be reasonable to have the task permit solution via a Generator (i.e., a function/object/method that yields a single new value of the distribution each time called) as well as by generating an actual collection; after all, the latter is trivial to derive from the former. –Donal Fellows 07:59, 27 July 2011 (UTC)