FIX: Solving Int64 precision loss when read_csv(StringIO(data), dtype…

[vitorcf10]

…={a:Int64}, engine=pyarrow)

• [] closes BUG: read_csv losing precision when reading Int64[pyarrow] data with N/A values #56135 and BUG: read_csv loses precision when engine='pyarrow' and dtype Int64 #56136
• 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.

Enable the integer_object_nulls option by setting it to True in the following line of code in the "arrow_parser_wrapper.py" file:
table.to_pandas(types_mapper=self.kwds["dtype"].get, integer_object_nulls=True)
This adjustment is crucial when using the "pyarrow" engine as a parameter for the 'read_csv' function. It ensures that columns containing integers (Int64) with null elements within a dataframe maintain their precision. Without this setting, numpy may treat columns with null elements by converting the entire column data to Float, leading to precision loss. Enabling 'integer_object_nulls' prevents this issue and preserves the precision of integer data.

[phofl]

Doing what you are describing here has huge performance implications, I don't want to do this here

[mroeschke]

Thanks for the pull request, but it appears to have gone stale. If interested in continuing, please merge in the main branch, address any review comments and/or failing tests, and we can reopen.