[BEAM-14535] Add support for Pandas Dataframes to sklearn RunInference Implementation

• Type: Sub-task
• Status: Open
• Priority: P2
• Resolution: Unresolved
• Component/s: sdk-py-core
• Assignee: Ryan Thompson
• Reporter: Ryan Thompson
• Created: 31/May/22 15:50
• Updated: 04/Jun/22 23:33

Description

Sklearn pipelines are often set up to take pandas dataframes.

 

Our current implementation only supports numpy arrays.

 

This FR allows the sklearn implementation to autodetect pandas dataframes or numpy arrays and then combine them (via concat).

 

In the case of a pandas dataframe that value will be passed through to the pipeline.

Issue Links

• links to: GitHub Pull Request #17800