aesara
aesara copied to clipboard
Aesara is a Python library for defining, optimizing, and efficiently evaluating mathematical expressions involving multi-dimensional arrays.
We should provide helper functions for `aesara.tensor.random.[default_rng|RandomState]` in `aesara.tensor.random.__init__` that construct `RandomStream` objects. _Originally posted by @brandonwillard in https://github.com/aesara-devs/aesara/pull/1084#discussion_r934870363_
Quite a few "docstrings" (things that are formatted to look like docstrings for classes, functions, etc.) in the documentation aren't actual docstrings, and their content isn't in the actual codebase....
This PR introduces a lot of refactoring that changes object names and comments to use the term "rewrite" instead of "optimize" when appropriate. Also, modules with the name `opt`—or a...
While we're still using the C backend as the default, we should be aware that there are no real C implementations for LAPACK function-based `Op`s. We should see if there's...
As in https://numpy.org/doc/stable/reference/generated/numpy.matmul.html
Here are a few important guidelines and requirements to check before your PR can be merged: + [x] There is an informative high-level description of the changes. + [ ]...
This PR adds the capability to effectively replace nodes affected due to variable replacement by cloning them whenever one of it's input is changed in the FunctionGraph. (Currently this is...
The way that `aesara.tensor.basic_opt.constant_folding` works does not obey the evaluation expectations of some `Op`s like `IfElse` and the stability expectations provided by our stabilizing rewrites. In general, `constant_folding` is a...
This PR enables type checking for a few of the `aesara.tensor` modules. There are still some inherent class/interface design issues that prevent type checking from working without manual intervention. One...