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

Expose step size parameters for AutoFiniteDiff #688

Open
1-Bart-1 opened this issue Jan 21, 2025 · 6 comments · May be fixed by #693
Open

Expose step size parameters for AutoFiniteDiff #688

1-Bart-1 opened this issue Jan 21, 2025 · 6 comments · May be fixed by #693

Comments

@1-Bart-1
Copy link

When using AutoFiniteDiff, one should be able to change the absstep and relstep parameters as described in the FiniteDiff.jl documentation. This seems to be impossible with the current DifferentiationInterface.jl API.

From this question on the Julia discourse: https://discourse.julialang.org/t/changing-the-relstep-of-autofinitediff-in-an-optimizationfunction/125004

@gdalle
Copy link
Member

gdalle commented Jan 21, 2025

This will first require SciML/ADTypes.jl#101 to be implemented on the ADTypes side, and then we can take it into account here! Do you want to give it a try?

@1-Bart-1
Copy link
Author

Try as in forking ADTypes.jl, adding the changes and creating a pull request there? I can try doing that!

@gdalle
Copy link
Member

gdalle commented Jan 21, 2025

Yes exactly!

@1-Bart-1
Copy link
Author

SciML/ADTypes.jl#102

@gdalle
Copy link
Member

gdalle commented Jan 21, 2025

@1-Bart-1 wanna try to make the modifications on this side now that the new version of ADTypes has been released?

@1-Bart-1 1-Bart-1 linked a pull request Jan 21, 2025 that will close this issue
@1-Bart-1
Copy link
Author

#693

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

Successfully merging a pull request may close this issue.

2 participants