Skip to content

G-test raises ValueError when reference and current datasets have different sample sizes #1914

Description

@devendrakushwah80

Bug

The g_test statistical drift detector raises a ValueError when the
reference and current datasets have different sample sizes.

Reproduction

import pandas as pd

from evidently.legacy.calculations.stattests.g_stattest import _g_stat_test
from evidently.legacy.core import ColumnType

reference = pd.Series(["a"] * 500 + ["b"] * 500)
current = pd.Series(["a"] * 50 + ["b"] * 50)

_g_stat_test(reference, current, ColumnType.Categorical, 0.05)

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions