-
|
Are there plans to support this interface? Would that work similar to GroupBy/Apply (with RelationalGroupedDataset)? Hopefully it would allow the remote UDF's to interact with the Apache Arrow recordbatch for inputs and outputs. I can use GroupBy/Apply in the meantime. But it seems like I'm always trying to find "artificial" ways of grouping the data just to be able to exchange recordbatch with the remote executors. I believe the "artificial" groupings would become unnecessary if we could run this mapInPandas. |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
|
@dbeavon Could you create a feature request? I think we should be able to support this. |
Beta Was this translation helpful? Give feedback.
-
|
@imback82 Done. I wasn't certain if this was already hidden somewhere in the existing code, or if it was a work in progress. |
Beta Was this translation helpful? Give feedback.
@dbeavon Could you create a feature request? I think we should be able to support this.