ray.data.Datasource.write#

Datasource.write(blocks: Iterable[Union[pyarrow.Table, pandas.DataFrame]], ctx: ray.data._internal.execution.interfaces.task_context.TaskContext, **write_args) → Any[source]#

Write blocks out to the datasource. This is used by a single write task.

Parameters

  • blocks – List of data blocks.

  • ctx – TaskContext for the write task.

  • write_args – Additional kwargs to pass to the datasource impl.

Returns

The output of the write task.