[SPARK-37678] Incorrect annotations in SeriesGroupBy._cleanup_and_return

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 3.3.0
• Fix Version/s: 3.2.1, 3.3.0
• Component/s: PySpark
• Assignee: Maciej Szymkiewicz
• Reporter: Maciej Szymkiewicz
• Created: 17/Dec/21 18:33
• Updated: 20/Dec/21 22:18
• Resolved: 20/Dec/21 22:18

Description

SeriesGroupBy._cleanup_and_return annotations

    def _cleanup_and_return(self, pdf: pd.DataFrame) -> Series:
        return first_series(pdf).rename().rename(self._psser.name)

are inconsistent:

• If pdf is pd.DataFrame then output should be pd.Series.
• If output is ps.Series then pdf should be ps.DataFrame.

Doesn't seem like the method is used (it is possible that my search skills and PyCharm inspection failed), so I am not sure which of these options was intended.

Issue Links

• links to: [Github] Pull Request #34950 (zero323)
• links to: [Github] Pull Request #34950 (zero323)