-
Notifications
You must be signed in to change notification settings - Fork 901
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
Add dispatch for cudf.Dataframe
to/from pyarrow.Table
conversion
#13558
Add dispatch for cudf.Dataframe
to/from pyarrow.Table
conversion
#13558
Conversation
kwargs.pop("self_destruct", None) | ||
if kwargs: | ||
warnings.warn( | ||
f"Ignoring the following arguments to " | ||
f"`from_pyarrow_table_dispatch`: {list(kwargs)}" | ||
) |
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.
Note that self_destruct
is currently the only argument "needed" by the p2p algorithm (for pandas backed data). However, cudf.DataFrame.from_arrow
doesn't take this (or any other) key-word argument.
Since I expect a bit of iteration on exactly which key-word arguments these dispatch functions will need to support, I am temporarily using kwargs
and warning the user if an unexpected argument is encountered. I'd prefer for these functions to not raise an error until we have time to nail down the exact API we need.
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.
Minor, but looks good, thanks @rjzamora
/merge |
Description
Registers the necessary cudf-specific logic for the dispatch functions introduced in dask/dask#10312. These dispatch functions are necessary to enable "p2p" shuffling with a cudf backend (see: dask/distributed#7743)
Checklist