BUG: fixes Arrow Dataframes/Series producing a Numpy object result

[mcgeestocks]

• closes BUG: dot on Arrow Dataframes/Series produces a Numpy object result #53979
• Tests added and passed if fixing a bug or adding a new feature
• All code checks passed.
• Added type annotations to new arguments/methods/functions.
• Added an entry in the latest doc/source/whatsnew/vX.X.X.rst file if fixing a bug or adding a new feature.

Notes:

.dot() 's returned object dtype is now the result of dtypes.cast.find_common_type().

It's worth noting the new object's dtype's is of a higher precision but this seems line up what's expected: https://arrow.apache.org/docs/python/pandas.html#pandas-arrow-conversion

[mcgeestocks]

Hey @phofl do you know who I should tag for a review on this?

[mroeschke]

Could you add a test for masked types as well? I suspect they failed before as well

[mcgeestocks]

Could you add a test for masked types as well? I suspect they failed before as well

Sure! I made an assumption about what you meant by the masked types Int8, Int16, Int32....

After I updated the tests I noticed another potential edge case that when calling dot with two different masked dtypes, for example Float64 and float[pyarrow]. I get return type Object

How would we want to handle this case?

[mroeschke]

After I updated the tests I noticed another potential edge case that when calling dot with two different masked dtypes, for example Float64 and float[pyarrow]. I get return type Object

Let's ignore this case for now

[mroeschke]

Thanks @mcgeestocks