Dask

class hamilton.plugins.h_dask.DaskDataFrameResult[source]

static build_result(**outputs: dict[str, Any]) → Any[source]

Builds a dask dataframe from the outputs.

This has some assumptions:

1. the order specified in the output will mirror the order of “joins” here.
2. it tries to massage types into dask types where it can
3. otherwise it duplicates any “scalars/objects” using the first valid input with an index as the template. It assumes a single partition.