BUG: Get wrong result when groupby category column with dropna=False

[zxymath]

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

# Your code here
ser = pd.Series([1., 1., 1., 1.])
cat = pd.Categorical(['a', 'b', 'c', np.nan])
print(ser.groupby(cat, dropna=False).sum())

# output:
a    1.0
b    1.0
c    1.0
dtype: float64

Expected Output

a    1.0
b    1.0
c    1.0
NaN  1.0
dtype: float64

Output of pd.show_versions()

INSTALLED VERSIONS

commit : 2a7d332
python : 3.7.6.final.0
python-bits : 64
OS : Windows
OS-release : 10
Version : 10.0.18362
machine : AMD64
processor : Intel64 Family 6 Model 158 Stepping 13, GenuineIntel
byteorder : little
LC_ALL : None
LANG : None
LOCALE : None.None

pandas : 1.1.2
numpy : 1.18.1
pytz : 2019.3
dateutil : 2.8.1
pip : 20.0.2
setuptools : 45.2.0.post20200210
Cython : 0.29.15
pytest : 5.3.5
hypothesis : 5.5.4
sphinx : 2.4.0
blosc : None
feather : None
xlsxwriter : 1.2.7
lxml.etree : 4.5.0
html5lib : 1.0.1
pymysql : 0.9.3
psycopg2 : None
jinja2 : 2.11.1
IPython : 7.12.0
pandas_datareader: None
bs4 : 4.8.2
bottleneck : 1.3.2
fsspec : 0.6.2
fastparquet : None
gcsfs : None
matplotlib : 3.1.3
numexpr : 2.7.1
odfpy : None
openpyxl : 3.0.3
pandas_gbq : None
pyarrow : None
pytables : None
pyxlsb : None
s3fs : None
scipy : 1.4.1
sqlalchemy : 1.3.13
tables : 3.6.1
tabulate : None
xarray : None
xlrd : 1.2.0
xlwt : 1.3.0
numba : 0.48.0

[phofl]

Hey,

have you checked #35646? This looks like a duplicate

[FenderJazz]

It looks different from #35646 to me -- is specifically about categoricals, and happens with >1 groupby columns?

Your code here

cat = pd.Categorical(['a', 'b', 'c', np.nan])
df = pd.DataFrame({'ser': ser, 'var1': cat, 'var2': cat})
print(df.groupby(['var1', 'var2'], observed=True, dropna=False).sum())

# Same but not categorical
noncat = pd.Series(['a', 'b', 'c', np.nan])
df = pd.DataFrame({'ser': ser, 'var1': noncat, 'var2': noncat})
print(df.groupby(['var1', 'var2'], observed=True, dropna=False).sum())

Output

var1 var2     
a    a     1.0
b    b     1.0
c    c     1.0
           ser
var1 var2     
a    a     1.0
b    b     1.0
c    c     1.0
NaN  NaN   1.0

Expected output

var1 var2     
a    a     1.0
b    b     1.0
c    c     1.0
NaN  NaN   1.0
           ser
var1 var2     
a    a     1.0
b    b     1.0
c    c     1.0
NaN  NaN   1.0

[timstaley]

Just a +1 report of hitting this issue, still present on 1.2.0rc0

[rhshadrach]

Thanks for the report! Further investigations and PRs to fix are welcome.