raulcd commented on issue #46138: URL: https://github.com/apache/arrow/issues/46138#issuecomment-2804288078
I think there is an issue with how `CastOptions` interact between table/compute on the `cast` function. I don't have much experience with how this should work and can't spend much to investigate internally, it has to do with how Table/Column(aka RecordBatch)/compute interact with the cast function, basically table uses: https://github.com/apache/arrow/blob/b603ea5b27092dc870701baade7cdd50462a3e09/python/pyarrow/table.pxi#L4721-L4725 `field.type` will always have a value and options (on this case) too but then we check that either field and safe or options have to be populated: https://github.com/apache/arrow/blob/b603ea5b27092dc870701baade7cdd50462a3e09/python/pyarrow/_compute.pyx#L2775-L2779 so from what I can see, for table.cast if we add `CastOptions` this will fail. If someone can investigate a little more that would be awesome. As per your specific example you can do (without using `CastOptions` but that's still not solving the issue with the interaction of `CastOptions` on `table.cast()`: ``` >>> table.cast(new_schema, safe=False) pyarrow.Table timestamp: timestamp[us, tz=UTC] ---- timestamp: [[2023-03-15 15:15:00.123456Z]] ``` -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
