jorisvandenbossche commented on issue #38644: URL: https://github.com/apache/arrow/issues/38644#issuecomment-1810016987
> I was weired since I expected that pa.table.to_pandas() with default params does the best to do not convert data to pandas 2.x types so I expected: Note that it actually _did_ convert the data to the pandas 2.x data types, because "int64" is the default integer data type in pandas 2.0. The ArrowDtype you were expecting is only an experimental, opt-in data type. > At the end I decided to double check that zero-copy is used by: Yeah, and so at that point using `types_mapper` with `zero_copy_only=True` doesn't really work well together, as explained in my previous post. Short term you can of course just leave out the `zero_copy_only=True`. On the pyarrow side, we should at least improve the error message, and _maybe_ we could ignore the zero-copy keyword for columns that go through types_mapper (didn't look in detail how complex that would be). -- 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]
