6 Mar
2023
6 Mar
'23
5:30 p.m.
Hi everyone, I propose the addition of Dunnett's test function. It is similar to Tukey’s test, but instead compares all sample groups to a control group. Hence the test performs less p-value calculation and is more powerful in that situation. It is pretty much the recommended, and widely used, method when comparing groups against a control. Note that the function is listed on our roadmap in the stats section under “post hoc tests": https://scipy.github.io/devdocs/dev/roadmap-detailed.html#stats We would gladly get your opinion on the PR: https://github.com/scipy/scipy/pull/18107 Cheers, Pamphile