ray.data.Datasink.write#

Datasink.write(blocks: Iterable[pyarrow.Table | pandas.DataFrame], ctx: TaskContext) → Any[source]#

Write blocks. This is used by a single write task.

Parameters:

  • blocks – Generator of data blocks.

  • ctx – TaskContext for the write task.

Returns:

A user-defined output. Can be anything, and the returned value is passed to on_write_complete().