BUG: DataFrameGroupBy.groups fails when Categorical indexer contains NaNs and dropna=False

[tehunter]

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

>>> df = DataFrame(
...         {
...             "cat": Categorical(["a", np.nan, "a"], categories=["a", "b", "d"]),
...             "vals": [1, 2, 3],
...         }
...     )
>>> g = df.groupby("cat", observed=True, dropna=False)
>>> result = g.groups
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/workspaces/pandas/pandas/core/groupby/groupby.py", line 569, in groups
    return self._grouper.groups
  File "properties.pyx", line 36, in pandas._libs.properties.CachedProperty.__get__
  File "/workspaces/pandas/pandas/core/groupby/ops.py", line 710, in groups
    return self.groupings[0].groups
  File "properties.pyx", line 36, in pandas._libs.properties.CachedProperty.__get__
  File "/workspaces/pandas/pandas/core/groupby/grouper.py", line 711, in groups
    return codes, uniques
  File "/workspaces/pandas/pandas/core/arrays/categorical.py", line 745, in from_codes
    dtype = CategoricalDtype._from_values_or_dtype(
  File "/workspaces/pandas/pandas/core/dtypes/dtypes.py", line 347, in _from_values_or_dtype
    dtype = CategoricalDtype(categories, ordered)
  File "/workspaces/pandas/pandas/core/dtypes/dtypes.py", line 230, in __init__
    self._finalize(categories, ordered, fastpath=False)
  File "/workspaces/pandas/pandas/core/dtypes/dtypes.py", line 387, in _finalize
    categories = self.validate_categories(categories, fastpath=fastpath)
  File "/workspaces/pandas/pandas/core/dtypes/dtypes.py", line 585, in validate_categories
    raise ValueError("Categorical categories cannot be null")
ValueError: Categorical categories cannot be null
>>>

Issue Description

When using df.groupby(cat, dropna=False).groups, we encounter a ValueError. This is counter-intuitive, as grouping operations work without an issue.

>>> df = DataFrame(
...         {
...             "cat": Categorical(["a", np.nan, "a"], categories=["a", "b", "d"]),
...             "vals": [1, 2, 3],
...         }
...     )
>>> g = df.groupby("cat", observed=True, dropna=False)
>>> g.sum()
     vals
cat      
a       4
NaN     2
>>> g.sum().index
CategoricalIndex(['a', nan], categories=['a', 'b', 'd'], ordered=False, dtype='category', name='cat')

Expected Behavior

.groups should return a dictionary which includes the NaN as the last entry.

Installed Versions

INSTALLED VERSIONS

commit : 41131a1
python : 3.10.8
python-bits : 64
OS : Linux
OS-release : 5.15.167.4-microsoft-standard-WSL2
Version : #1 SMP Tue Nov 5 00:21:55 UTC 2024
machine : x86_64
processor :
byteorder : little
LC_ALL : None
LANG : C.UTF-8
LOCALE : en_US.UTF-8

pandas : 3.0.0.dev0+2085.g41131a1432
numpy : 2.2.5
dateutil : 2.9.0.post0
pip : 25.0.1
Cython : 3.0.12
sphinx : 8.1.3
IPython : 8.35.0
adbc-driver-postgresql: None
adbc-driver-sqlite : None
bs4 : 4.13.4
blosc : None
bottleneck : 1.4.2
fastparquet : 2024.11.0
fsspec : 2025.3.2
html5lib : 1.1
hypothesis : 6.131.8
gcsfs : 2025.3.2
jinja2 : 3.1.6
lxml.etree : 5.4.0
matplotlib : 3.10.1
numba : 0.61.2
numexpr : 2.10.2
odfpy : None
openpyxl : 3.1.5
psycopg2 : 2.9.10
pymysql : 1.4.6
pyarrow : 19.0.1
pyreadstat : 1.2.8
pytest : 8.3.5
python-calamine : None
pytz : 2025.2
pyxlsb : 1.0.10
s3fs : 2025.3.2
scipy : 1.15.2
sqlalchemy : 2.0.40
tables : 3.10.1
tabulate : 0.9.0
xarray : 2024.9.0
xlrd : 2.0.1
xlsxwriter : 3.2.3
zstandard : 0.23.0
tzdata : 2025.2
qtpy : None
pyqt5 : None

[rhshadrach]

Thanks for the report. Confirmed on main. PR to fix is up.