BUG: margin=True is affected by NaN in other columns in pivot_table

[nikhilpatwardhan]

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

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

• (optional) I have confirmed this bug exists on the master branch of pandas.

---

Note: Please read this guide detailing how to provide the necessary information for us to reproduce your bug.

Code Sample, a copy-pastable example

>>> import pandas as pd
>>> data = {'A': [6, 7, None], 'B': [None, 5, 8], 'name': ['Alice', 'Alice', 'Alice']}
>>> df = pd.DataFrame(data)
>>> df.pivot_table(index='name', values=['A', 'B'], aggfunc=['min', 'max'], margins=True)

Output:

       min       max     
         A    B    A    B
name                     
Alice  6.0  5.0  7.0  8.0
All    7.0  5.0  7.0  5.0

Problem description

The summary line 'All' shows min for A as 7.0 and max for B as 5.0 although min of A is 6.0 and max of B is 8.0

Expected Output

       min       max     
         A    B    A    B
name                     
Alice  6.0  5.0  7.0  8.0
All    6.0  5.0  7.0  8.0

Output of pd.show_versions()

INSTALLED VERSIONS

commit : 6bc8a49
python : 3.8.2.final.0
python-bits : 64
OS : Darwin
OS-release : 19.3.0
Version : Darwin Kernel Version 19.3.0: Thu Jan 9 20:58:23 PST 2020; root:xnu-6153.81.5~1/RELEASE_X86_64
machine : x86_64
processor : i386
byteorder : little
LC_ALL : None
LANG : en_US.UTF-8
LOCALE : en_US.UTF-8

pandas : 1.1.0.dev0+1225.g6bc8a49f2
numpy : 1.18.1
pytz : 2019.3
dateutil : 2.8.1
pip : 20.0.2
setuptools : 46.1.3.post20200325
Cython : 0.29.16
pytest : 5.4.1
hypothesis : 5.8.0
sphinx : 3.0.1
blosc : None
feather : None
xlsxwriter : 1.2.8
lxml.etree : 4.5.0
html5lib : 1.0.1
pymysql : None
psycopg2 : None
jinja2 : 2.11.1
IPython : 7.13.0
pandas_datareader: None
bs4 : 4.9.0
bottleneck : 1.3.2
fastparquet : 0.3.3
gcsfs : None
matplotlib : 3.2.1
numexpr : 2.7.1
odfpy : None
openpyxl : 3.0.3
pandas_gbq : None
pyarrow : 0.16.0
pytables : None
pyxlsb : None
s3fs : 0.4.2
scipy : 1.4.1
sqlalchemy : 1.3.16
tables : 3.6.1
tabulate : 0.8.7
xarray : 0.15.1
xlrd : 1.2.0
xlwt : 1.3.0
numba : 0.48.0

[dsaxton]

This looks to be the source of the bug (it's dropping rows with any missing data after calculating the pivot table but before getting the margins):

pandas/pandas/core/reshape/pivot.py

Line 159 in d106b81

data = data[data.notna().all(axis=1)]

which was itself added to fix a bug in crosstab: #12614. So somehow this would have to be fixed without breaking crosstab.

[theavey]

Still exists in pandas 1.4.3

>>> df = pd.DataFrame({"a": [1, 2, None, 4], "b": ["a", "a", "a", "b"], "c": [1] * 4})
>>> df
	a	b	c
0	1.00	a	1
1	2.00	a	1
2	NaN	a	1
3	4.00	b	1
>>> df.pivot_table(index="b", values=["a", "c"], aggfunc="sum", margins=True)
	a	c
b		
a	3.00	3
b	4.00	1
All	7.00	3

Would instead expect

>>> df.pivot_table(index="b", values=["a", "c"], aggfunc="sum", margins=True)
	a	c
b		
a	3.00	3
b	4.00	1
All	7.00	4