Add read-only functions on string dtypes to DataFrame.apply and Series.apply
#11319
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
Co-authored-by: Vyas Ramasubramani <[email protected]>
raydouglass
reviewed
Sep 14, 2022
|
rerun tests |
|
Can you remerge with 22.10? There are bunch of extra files in the PR. |
brandon-b-miller
changed the base branch from
feature/string_udfs
to
branch-22.10
shwina
approved these changes
Sep 20, 2022
This should be resolved with the change of base. |
davidwendt
approved these changes
Sep 20, 2022
|
@gpucibot merge |
raydouglass
pushed a commit
that referenced
this pull request
Sep 29, 2022
## Description This switches to using CubinLinker (from PTXCompiler, but CubinLinker uses PTXCompiler internally) for Minor Version Compatibility. This enables support for all Numba features except linking archives with MVC, in support of use cases such as String UDFs (#11319) with MVC. ## Checklist - [X] I am familiar with the [Contributing Guidelines](https://github.com/rapidsai/cudf/blob/HEAD/CONTRIBUTING.md). - [X] New or existing tests cover these changes. - [X] The documentation is up to date with these changes. Authors: - Graham Markall (https://github.com/gmarkall) - https://github.com/brandon-b-miller - Ashwin Srinath (https://github.com/shwina) Approvers: - Ray Douglass (https://github.com/raydouglass)
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Outdated suggestions cannot be applied.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.
This PR provides initial support for string data inside UDFs passed to
DataFrame.applyandSeries.apply. The allowed APIs are based on python'sstrclass. It aims to implement python string semantics as closely as possible starting with APIs that return numeric data only. These are the following 21 functions:str.countstr.startswithstr.endswithstr.findstr.rfindstr.isalnumstr.isdecimalstr.isdigitstr.islowerstr.isupperstr.isalphastr.istitlestr.isspace==,!=,>=,<=,>,<(between two strings)len__contains__The following 3 functions are not included due to having no libcudf equivalent code available to back them (due to them referring to python concepts)
str.isasciistr.isidentifierstr.isprintableThis works by creating a library of
__device__functions based on libcudf which perform the above functions for one single string. The rest of the code is a library of numba extensions that replace a python UDF with a chain of those__device__functions and creates a kernel that calls the result across a grid of threads, taking a full column of strings as input.cc @davidwendt @gmarkall