Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Array API adoption #12

Open
matthewfeickert opened this issue May 3, 2024 · 8 comments
Open

Array API adoption #12

matthewfeickert opened this issue May 3, 2024 · 8 comments

Comments

@matthewfeickert
Copy link
Member

No description provided.

@betatim
Copy link

betatim commented May 6, 2024

I am interested in this topic. I've been working on Array API support in scikit-learn and would be interested to talk with other maintainers who are doing similar work in other array consuming libraries. I am also part of the committee working on the standard itself. I am up for comparing notes, discussing warts, working on scikit-learn itself, mentoring others who want to work on Array API support in scikit-learn

@betatim betatim self-assigned this May 6, 2024
@matthewfeickert matthewfeickert self-assigned this May 6, 2024
@matthewfeickert
Copy link
Member Author

matthewfeickert commented May 6, 2024

For the particle physics people, in addition to engaging in broader community adoption, we (myself and @nsmith-) should also follow up wtih @jpivarski in advance of the Summit on the state and ideas of https://github.com/scikit-hep/ragged for Awkward Array / Array API integration.

@nsmith- nsmith- self-assigned this May 6, 2024
@QuLogic QuLogic self-assigned this May 15, 2024
@steppi
Copy link

steppi commented May 22, 2024

Hi everyone. I've been working on laying the groundwork for a special function array API extension and would like to participate in this topic. For some reason I'm not able to assign myself to this issue.

@mdhaber
Copy link

mdhaber commented May 22, 2024

I've been converting scipy.stats to the array API (scipy/scipy#20544) and dispatching scipy.special functions to the appropriate backend (scipy/scipy#19023). Supporting projects include arbitrary precision array API arrays (mparray) and array-API-backend agnostic masked arrays (scipy/scipy#20363); it would be nice to get feedback on whether/how these might be used in the rest of the ecosystem.

@steppi
Copy link

steppi commented May 22, 2024

Thanks for assigning me @mdhaber

@ksunden ksunden self-assigned this May 22, 2024
@j-bowhay j-bowhay self-assigned this May 22, 2024
@seberg seberg self-assigned this May 23, 2024
@thomasjpfan thomasjpfan self-assigned this May 27, 2024
@thomasjpfan
Copy link

From working on Array API adopting and seeing the progress in Scikit-learn and SciPy, I have noticed a few areas of friction for adoption.

For the summit, I want to write down all the things we learned from adopting the standard. This way, libraries that are interested in consuming the Array API standard can have an easier time.

@betatim
Copy link

betatim commented Jun 3, 2024

Some topics to talk about on Monday morning:

  • missing functions in the array API standard
  • casting/conversion for inputs?
    • e.g. if user inputs ints, should we convert to floating point? should we require the user converts?
  • array-api-compat versioning? what to do if users have an old version?
  • GitHub GPU CI/testing of array API code
    • eScience has experience running the GitHub GPU runners
  • should there be a scipy built on top of array API?
  • dispatching

@thomasjpfan
Copy link

Shared hackmd: https://hackmd.io/@thomasjpfan/BJ0bF_oNC/edit

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
Status: Todo
Development

No branches or pull requests

10 participants