jorisvandenbossche commented on issue #43298: URL: https://github.com/apache/arrow/issues/43298#issuecomment-2441268379
In general Arrow does not support a mixed type object column from pandas. PyArrow will try to infer the type for an object-dtype column (e.g. it it contains all strings, the conversion will work fine), but in the example case in the top post, it is a mix of strings and integer. Such a mix will fail by default, and possible workarounds are first casting everything to strings (to ensure a consistent type), or in theory one could also use the Arrow `union` type (which does allow storing mixed types in a single column, https://arrow.apache.org/docs/dev/python/data.html#union-arrays. However, conversion from pandas to pyarrow is not supported for union, so one would have to do the conversion more manually as shown in the link) -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
