[ARROW-8442] [Python] NullType.to_pandas_dtype inconsisent with dtype returned in to_pandas/to_numpy

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.17.0
• Component/s: Python
• Labels: pull-request-available
• External issue URL: https://github.com/apache/arrow/issues/24621
• Assignee: Joris Van den Bossche
• Reporter: Joris Van den Bossche
• Created: 14/Apr/20 09:22
• Updated: 11/Jan/23 08:00
• Resolved: 14/Apr/20 11:08

Description

There is this behaviour of to_pandas_dtype returning float, while all actual conversions to numpy or pandas use object dtype:

In [23]: pa.null().to_pandas_dtype()                                                                                                                                                                               
Out[23]: numpy.float64

In [24]: pa.array([], pa.null()).to_pandas()                                                                                                                                                                       
Out[24]: Series([], dtype: object)

In [25]: pa.array([], pa.null()).to_numpy(zero_copy_only=False)                                                                                                                                                    
Out[25]: array([], dtype=object)

So we should probably fix NullType.to_pandas_dtype to return object, which is used in practice.

Issue Links

• links to: GitHub Pull Request #6930