BUG: inconsistent behaviour of Groupby (probably a regression)

[TheoSimier]

First, thank you very much for your work ! I love pandas !

The problem

Code Sample

import pandas as pd
# intialise data of lists. 
data = {'Fruit':['Apple', 'Orange', 'Apple'],
        'Origin':['France', 'France', 'Spain'],
        'Price':[10, 15, 20]}
  
# Create DataFrame without categorical variable
df = pd.DataFrame(data)
# Same dataframe with categorical variable
df_category = df.copy()
df_category['Origin'] = df_category['Origin'].astype('category')
df

# Normal behavior
df.groupby(['Fruit','Origin'])['Price'].mean()

# Abnormal behavior
df_category.groupby(['Fruit','Origin'])['Price'].mean()

Return

# Normal behavior
Fruit   Origin
Apple   France    10
        Spain     20
Orange  France    15
Name: Price, dtype: int64

# Abnormal behavior
Fruit   Origin
Apple   France    10.0
        Spain     20.0
Orange  France    15.0
        Spain      NaN
Name: Price, dtype: float64

Problem description

Bug that should have been solved with #20583. I supposed it is a regression. No new discussion since May 2018, so I prefer to reopen a issue.

Brief description: When a column of a groupby is a categorical variable, the output include rows with NA.

Expected Output

Fruit   Origin
Apple   France    10
        Spain     20
Orange  France    15
Name: Price, dtype: int64

Output of pd.show_versions()

INSTALLED VERSIONS ------------------ commit : None python : 3.7.0.final.0 python-bits : 64 OS : Darwin OS-release : 18.7.0 machine : x86_64 processor : i386 byteorder : little LC_ALL : None LANG : fr_FR.UTF-8 LOCALE : fr_FR.UTF-8

pandas : 0.25.1
numpy : 1.16.5
pytz : 2019.2
dateutil : 2.8.0
pip : 19.2.2
setuptools : 41.0.1
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : 4.4.1
html5lib : None
pymysql : None
psycopg2 : 2.7.5 (dt dec pq3 ext lo64)
jinja2 : 2.10.1
IPython : 7.8.0
pandas_datareader: None
bs4 : 4.8.0
bottleneck : None
fastparquet : None
gcsfs : None
lxml.etree : 4.4.1
matplotlib : 3.1.1
numexpr : None
odfpy : None
openpyxl : 2.6.2
pandas_gbq : None
pyarrow : None
pytables : None
s3fs : None
scipy : 1.3.1
sqlalchemy : None
tables : None
xarray : None
xlrd : 1.2.0
xlwt : None
xlsxwriter : None

[TheoSimier]

After taking more time to read the documentation the answer is in the documentation of the groupby function: https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.groupby.html. Seems strange to me that the defaut value of observed is False but okay.

observed : bool, default False

    This only applies if any of the groupers are Categoricals. If True: only show observed values for categorical groupers. If False: show all values for categorical groupers.

    New in version 0.23.0.