pymc-devs / pymc-devs/pytensor

Consider specialized DimShuffle C-implementation

Open
#1,246 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

C-backend performance
Dominant language
Python
Stars
644
Forks
208
Avg merge
2d 14h
Merged PRs (30d)
16

Description

Description

Right now DimShuffle uses an external C-op that is generic over variations of DimShuffle. In the older time it used to be one of the regular unreadable inplice c_code methods, which used the input/output ndim information directly, and could just define shape/strides as a constant size array and allocate it without a loop, like we have now.

This is probably a minor optimization, may not even buy us anything. This issue is a low-priority thing to just keep in mind.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by locating DimShuffle's external generic C-op and compare it with the older inline c_code implementation described in the issue. Benchmark the current and specialized approaches, including constant-size shape and stride arrays; done means either a measured optimization or evidence that no change is worthwhile.

Written by the indexing model from the issue text.

Assessment

Tech stack
c, python
Domain
performance
Issue type
Refactor
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.