AlenkaF commented on issue #35235: URL: https://github.com/apache/arrow/issues/35235#issuecomment-1515995973
Here is an update after going through the issue with Joris: * the first failing test where we get the casting error is happening due to the change on the Pandas side with https://github.com/pandas-dev/pandas/pull/52201. Due to `__from_pyarrow__` being added for pandas `DatetimeTZDtype` the conversion of the column in a table on the arrow side is now happening through this method and not through `_array_like_to_pandas` as it did before. Because in pyarrow we still cast datetime types to `ns` resolution (https://github.com/apache/arrow/issues/34462), the `__from_pyarrow__` on the pandas side receives the casted data type (that is in `ns` resolution) and does the casting there but fails as the value is out of bounds. For now I will skip this test for dev version of pandas as this is a bigger issue that needs to be solved in pyarrow for 13.0.0 (https://github.com/apache/arrow/issues/33321). * the second failing test has the same root cause but can be fixed easily by passing name to `pandas_api.series(arr)` in `_to_pandas()` (table.pxi) I will create a PR asap to skip/fix the tests. -- 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]
