-
Notifications
You must be signed in to change notification settings - Fork 46
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
Fix documented dtype inputs for rfft-like functions #696
Conversation
- Replace complex with real - Use must language instead of should
These changes should likely be backported, correct? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is incorrect. Rfft is for r2c transforms, and irfft for c2r.
I should've remembered to do that 😅 Backported to
Oh okay, did only skim over #189 so I might of missed relevant discussion? I was looking at |
Sorry for brevity @honno, I left the comment on cell in a rush...
No, that's a misnomer. Now, repeat after me 🙂
This is the only way that |
Missed this one. No, this is a behavior that we purposely forbid. Ignoring the imaginary part (either silently or raising a warning) was a poor behavior in NumPy, and I pushed during the discussions to ensure array API does not pick it up. |
@honno Have you had the chance to follow up here? |
Here's a summary of input-output behaviour
I think I see what @leofang means now, so disregard this PR! 😅 |
For the real fft family of functions:
rfft
is right now, although I could see maybe that it should all be "should"?