DataFrame.dropna DataFrame.dropna( how: str = typing.Literal['any', 'all'], thresh: int | None = None, subset: str | Sequence[str] | None = None, ) PySpark API Reference Incompatibility Note The thresh parameter is not honored.