Open brandonwillard opened 1 year ago
Just a note that #184 originally introduced a rewrite for at.square
as well, but I think it would be better to only register at.pow
here and make sure they're equivalent Aesara-side: https://github.com/aesara-devs/aesara/issues/1213
It uses a positive support in cases of ambiguity.
Another arguments in favor of propagating domain information; we shouldn't have to make assumptions here.
This PR adds support for general power transforms and replaces #184. It uses a positive support in cases of ambiguity.
In its current form, it appears to work as a replacement for
at.reciprocal
, but the $Z = X^2$, where $X \sim \operatorname{N}\left(0, 1\right)$, test example fails.