rapidsai/cudf

[FEA] pandas DatetimeIndex.indexer_between_time

Open

#15,380 opened on Mar 23, 2024

View on GitHub
 (6 comments) (0 reactions) (0 assignees)C++ (6,000 stars) (735 forks)batch import
Pythonfeature requestgood first issue

Description

I watched @shwina's GTC talk (https://register.nvidia.com/flow/nvidia/gtcs24/attendeeportaldigital/page/sessioncatalog/session/1695219773174001AmnA thanks Ashwin I really enjoyed it!)

Is your feature request related to a problem? Please describe. No. I just noticed there wasn't an issue for DatetimeIndex.indexer_between_time. I also enjoyed the user experience that %%cudf.pandas.profile points users to raise issues to highlight pandas API that falls back to CPU (https://github.com/rapidsai/cudf/blob/branch-24.06/python/cudf/cudf/pandas/profiler.py#L300).

Describe the solution you'd like

import pandas as pd
pd.date_range("2023-01-01", "2023-01-02", freq="1h").indexer_between_time("09:00", "16:00")
import cudf
cudf.date_range("2023-01-01", "2023-01-02", freq="1h").indexer_between_time("09:00", "16:00")

Describe alternatives you've considered There may be a cudf work around in the meantime for a user who needs indexer_between_time that could be captured at https://docs.rapids.ai/api/cudf/stable/cudf_pandas/

Additional context Could create a new issue template with the "pandas" label (https://github.com/rapidsai/cudf/issues?q=is%3Aopen+is%3Aissue+label%3Apandas) to be used at https://github.com/rapidsai/cudf/blob/branch-24.06/python/cudf/cudf/pandas/profiler.py#L297 fixed a bug at https://github.com/rapidsai/cudf/pull/15381

Contributor guide

Tech stack
python
Domain
data
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
half day
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
clear
Prerequisites
Basic pandas datetime indexingFamiliarity with cudf
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
45
Research direction
The issue requests implementing pandas' DatetimeIndex.indexer between time in cudf. Start by examining the existing DatetimeIndex methods in cudf's source (e.g., python/cudf/cudf/core/indexes/datetime.py) and the reference pandas implementation (pandas/core/indexes/datetimes.py). The solution should leverage GPU accelerated time comparisons and avoid CPU fallback. Review linked PR #15381 for related profiling work. Consider using cudf's date range and comparison operators to build the indexer.
[FEA] pandas DatetimeIndex.indexer_between_time · rapidsai/cudf#15380 | Good First Issue