stats significance tests: why no functions for proportions?
Florin Andrei <[email protected]> Tue, 28 Jul 2020 21:21:10 -0700
| Newsgroups | gmane.comp.python.scientific.user |
|---|---|
| Message-ID | <[email protected]> |
See this cheatsheet I made for myself for a few basic stats problems: https://github.com/FlorinAndrei/misc/blob/master/ch24_review.ipynb These are all significance tests - I'm testing claims. Whenever I test a mean (one sample), or I compare two means (two samples), there's a convenient scipy.stats function that does everything in one line. But when I test a proportion (one sample), or I compare two proportions (two samples), I need to compute the z-score first, and then feed that into stats.norm.cdf(). Why aren't there any functions for proportions, where I could just feed the sample(s) and the functions would return the P-value in each case? -- Florin Andrei https://florin.myip.org/