Skip to content

[FEAT] Add CPU pandas casts to PDSH queries #23959

Description

@mroeschke

#23871 enabled the pdsh.py benchmarks to run with data generated from tcphgen-cli, namely decimal and date data, (without fallback) using cudf.pandas.

pdsh.py still has options to run these benchmarks with just pandas which currently fails because pandas doesn't natively support decimal or date types. The queries should contain a mapping of casts (e.g. decimal -> float64, date32 -> datetime64[ns]) to apply when running these benchmarks with only pandas

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    cudf.pandasIssues specific to cudf.pandasimprovementImprovement / enhancement to an existing function

    Type

    No type

    Projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions