Series apply method backed by masked UDFs
#9217
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
brandon-b-miller
added
feature request
brandon-b-miller
removed request for
a team,
hyperbolic2346 and
robertmaynard
brandon-b-miller
removed
gpuCI
libcudf
|
rerun tests |
brandon-b-miller
added
3 - Ready for Review
|
This is now ready for review. |
thomcom
approved these changes
Sep 30, 2021
|
rerun tests |
galipremsagar
approved these changes
Oct 1, 2021
brandon-b-miller
added
5 - Ready to Merge
|
@gpucibot merge |
rapids-bot bot
pushed a commit
that referenced
this pull request
Oct 12, 2021
DEPENDS on #9217 Introduces a row-like abstraction to the numba UDF pipeline which enables functions of the following form: ```python def f(row): return row['a'] + row['b'] ``` To be applied to dataframes with the corresponding column labels using ```python df.apply(f, axis=1) ``` Removes the `nulludf` decorator and as such is a breaking change. However since it was just introduced anyways as somewhat of a stopgap, the impact is hopefully low. Users will still be able to write functions the old way, but will require the `numba.cuda.jit(device=True)` decorator for the function to work when wrapped in a lambda ```python @cuda.jit(device=True) def f(x, y): return x +y df.apply(lambda row: f(row['a'], row['b']) ``` Makes it so that pandas and cudf can consume the exact same UDF in the same way. Authors: - https://github.com/brandon-b-miller Approvers: - Graham Markall (https://github.com/gmarkall) - Ashwin Srinath (https://github.com/shwina) URL: #9343
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Depends on #9174
Adds
Series.applywhich applies a scalar UDF elementwise to the series data returning a new series. Null sensitive. Works in terms of our numbaMaskedTypeextension type. Similar topd.Series.apply.