BUG: infer_dtype result for float with embedded pd.NA

[MarkusZimmerDLR]

Pandas version checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas.

• I have confirmed this bug exists on the main branch of pandas.

Reproducible Example

from pandas.api.types import infer_dtype
assert infer_dtype(pd.Series([1.,2.,.3,pd.NA], dtype=object)) ==  infer_dtype(pd.Series([1.,2.,.3,np.nan], dtype=object))

Issue Description

Dear pandas-folks,

This was checked for pandas V 2.3.0 and 2.2.X

When using pandas' infer_dtype on an object array consisting out of floats with embedded pd.NA, the result will be mixed-integer-float tough skipna is True as a default.

The same test for embedded np.nan returns floating.

    >>> from pandas.api.types import infer_dtype
    >>> infer_dtype(pd.Series([1,2,3,pd.NA], dtype=object))
    'integer'
    >>> infer_dtype(pd.Series([1,2,3,np.nan], dtype=object))
    'integer'
    >>> infer_dtype(pd.Series([1.,2.,.3,pd.NA], dtype=object))
    'mixed-integer-float' v <<< should be `floating`
    >>> infer_dtype(pd.Series([1.,2.,.3,np.nan], dtype=object))
    'floating'
    >>> infer_dtype(pd.Series(['1.0', np.nan],dtype=object))
    'string'
    >>> infer_dtype(pd.Series(['1.0', pd.NA],dtype=object))
    'string'

In case of other types, like integer or strings, the function does not produce a false / different output w.r.t. the na-type.

Context, I am maintaining a small project which assures integers in columns to stay integers - a common known issue. I you know of a well established extension for this purpose, feel free to point me towards it.

Expected Behavior

>>> infer_dtype(pd.Series([1.,2.,.3,pd.NA], dtype=object)) should return floating

Installed Versions

INSTALLED VERSIONS

commit : 2cc3762
python : 3.13.3
python-bits : 64
OS : Linux
OS-release : 4.18.0-553.51.1.el8_10.x86_64
Version : #1 SMP Fri Apr 25 00:55:37 EDT 2025
machine : x86_64
processor : x86_64
byteorder : little
LC_ALL : None
LANG : en_US.UTF-8
LOCALE : en_US.UTF-8

pandas : 2.3.0
numpy : 2.2.6
pytz : 2025.2
dateutil : 2.9.0.post0
pip : 25.1.1
Cython : None
sphinx : None
IPython : 9.2.0
adbc-driver-postgresql: None
adbc-driver-sqlite : None
bs4 : None
blosc : None
bottleneck : None
dataframe-api-compat : None
fastparquet : None
fsspec : None
html5lib : None
hypothesis : None
gcsfs : None
jinja2 : None
lxml.etree : None
matplotlib : None
numba : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
psycopg2 : None
pymysql : None
pyarrow : 20.0.0
pyreadstat : None
pytest : None
python-calamine : None
pyxlsb : None
s3fs : None
scipy : None
sqlalchemy : None
tables : None
tabulate : 0.9.0
xarray : None
xlrd : None
xlsxwriter : None
zstandard : None
tzdata : 2025.2
qtpy : None
pyqt5 : None

[arthurlw]

Confirmed on main! Investigations and PRs are welcome.

Thanks for raising this!

[heoh]

I want to contribute to this. Thank you for explaining the issue.

[heoh]

take

[MarkusZimmerDLR]

Since this seems to be a very simple and minor fix, is it possible to not wait for the 3.0 release? Or is the release imminent?