BUG: Incorrect index with groupby and groupby.agg when observed=False using categorical columns with/without as_index=False

[RogerThomas]

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

#!/usr/bin/env python
import argparse
import pandas as pd


def main(args):
    df_data = {
        "A": ["A1", "A2", "A3", "A1"],
        "B": ["B1", "B2", "B3", "B1"],
        "C": [1.1, 1.2, 1.3, 1.4]
    }
    df = pd.DataFrame(df_data)

    as_index = args.as_index

    print("df:")
    print(df)
    print('---------')

    print(f"Non-Category: Groupby non-agg, as_index={as_index}")
    gdf = df.groupby(["A", "B"], as_index=as_index)["C"].sum()
    print(gdf)
    print('---------')

    print(f"Non-Category: Groupby agg, as_index={as_index}")
    gdf = df.groupby(["A", "B"], as_index=as_index).agg({"C": "sum"})
    print(gdf)
    print('---------')

    df["A"] = df["A"].astype("category")
    df["B"] = df["B"].astype("category")

    print(f"Category: Groupby non-agg, as_index={as_index}")
    gdf = df.groupby(["A", "B"], as_index=as_index)["C"].sum()
    print(gdf)
    print('---------')

    print(f"Category: Groupby agg, as_index={as_index}")
    gdf = df.groupby(["A", "B"], as_index=as_index).agg({"C": "sum"})
    print(gdf)
    print('---------')


if __name__ == "__main__":
    parser = argparse.ArgumentParser(
        description="Main script",
        epilog="Example usage: python test_gb_with_cats.py filename",
    )
    parser.add_argument("--as-index", default=False, action="store_true")
    args = parser.parse_args()

    main(args)

Issue Description

When using cateogry types as gb columns with as_index=True

Both df.groupby and df.groupby.agg work but the result seems to have all permutations of the underyling categories.

See output of reproducible example below

df:
    A   B    C
0  A1  B1  1.1
1  A2  B2  1.2
2  A3  B3  1.3
3  A1  B1  1.4
---------
Non-Category: Groupby non-agg, as_index=True
A   B
A1  B1    2.5
A2  B2    1.2
A3  B3    1.3
Name: C, dtype: float64
---------
Non-Category: Groupby agg, as_index=True
         C
A  B
A1 B1  2.5
A2 B2  1.2
A3 B3  1.3
---------
Category: Groupby non-agg, as_index=True
A   B
A1  B1    2.5
    B2    0.0
    B3    0.0
A2  B1    0.0
    B2    1.2
    B3    0.0
A3  B1    0.0
    B2    0.0
    B3    1.3
Name: C, dtype: float64
---------
Category: Groupby agg, as_index=True
         C
A  B
A1 B1  2.5
   B2  0.0
   B3  0.0
A2 B1  0.0
   B2  1.2
   B3  0.0
A3 B1  0.0
   B2  0.0
   B3  1.3
---------

When using cateogry types as gb columns with as_index=False

df.groupby works but again the result seems to have all permutations of the underyling categories.
However, df.groupby.agg fails with a crpytic error message like

ValueError: Length of values (3) does not match length of index (9)

Expected Behavior

Both groupby and groupby.agg should work with/without as_index and shouldn't do a cross join on the underlying categories

Installed Versions

INSTALLED VERSIONS

commit : 66e3805
python : 3.7.13.final.0
python-bits : 64
OS : Darwin
OS-release : 20.6.0
Version : Darwin Kernel Version 20.6.0: Mon Aug 30 06:12:21 PDT 2021; root:xnu-7195.141.6~3/RELEASE_X86_64
machine : x86_64
processor : i386
byteorder : little
LC_ALL : None
LANG : en_GB.UTF-8
LOCALE : en_GB.UTF-8

pandas : 1.3.5
numpy : 1.21.1
pytz : 2021.1
dateutil : 2.8.2
pip : 22.0.4
setuptools : 57.0.0
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : 1.3.7
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : 2.9.3 (dt dec pq3 ext lo64)
jinja2 : 3.0.1
IPython : None
pandas_datareader: None
bs4 : None
bottleneck : None
fsspec : 2021.07.0
fastparquet : None
gcsfs : None
matplotlib : 3.5.0
numexpr : None
odfpy : None
openpyxl : 3.0.7
pandas_gbq : None
pyarrow : 4.0.1
pyxlsb : None
s3fs : None
scipy : 1.7.0
sqlalchemy : 1.4.32
tables : None
tabulate : 0.8.9
xarray : None
xlrd : 1.1.0
xlwt : None
numba : 0.53.1

[samukweku]

@RogerThomas if you pass observed=True for categorical group by, is there any change?

[RogerThomas]

Yes indeed @samukweku passing observed=True for the categorical group bys does indeed fix the issue, i.e.:

#!/usr/bin/env python
import argparse
import pandas as pd


def main(args):
    df_data = {
        "A": ["A1", "A2", "A3", "A1"],
        "B": ["B1", "B2", "B3", "B1"],
        "C": [1.1, 1.2, 1.3, 1.4]
    }
    df = pd.DataFrame(df_data)

    as_index = args.as_index

    print("df:")
    print(df)
    print('---------')

    print(f"Non-Category: Groupby non-agg, as_index={as_index}")
    gdf = df.groupby(["A", "B"], as_index=as_index)["C"].sum()
    print(gdf)
    print('---------')

    print(f"Non-Category: Groupby agg, as_index={as_index}")
    gdf = df.groupby(["A", "B"], as_index=as_index).agg({"C": "sum"})
    print(gdf)
    print('---------')

    df["A"] = df["A"].astype("category")
    df["B"] = df["B"].astype("category")

    print(f"Category: Groupby non-agg, as_index={as_index}")
    gdf = df.groupby(["A", "B"], as_index=as_index, observed=True)["C"].sum()
    print(gdf)
    print('---------')

    print(f"Category: Groupby agg, as_index={as_index}")
    gdf = df.groupby(["A", "B"], as_index=as_index, observed=True).agg({"C": "sum"})
    print(gdf)
    print('---------')


if __name__ == "__main__":
    parser = argparse.ArgumentParser(
        description="Main script",
        epilog="Example usage: python test_gb_with_cats.py filename",
    )
    parser.add_argument("--as-index", default=False, action="store_true")
    args = parser.parse_args()

    main(args)

[rhshadrach]

Thanks for the report! For observed=False (the default), this is expected behavior. For agg raising on categories with as_index=False, this is a duplicate of #36698. Closing.