Groupby Agg not working if different partials with same name on the same column

[charlesdong1991]

related to #28426 from PR #28428

from functools import partial

import pandas as pd
import numpy as np

quant50 = partial(np.percentile, q=50)
quant70 = partial(np.percentile, q=70)

test = pd.DataFrame({'col1': ['a', 'a', 'b', 'b', 'b'], 'col2': [1,2,3,4,5]})
test.groupby('col1').agg({'col2': [quant50, quant70]})

However, quant50 result is rewritten by quant70, so will get the same output on both columns.

output:

col1	percentile	percentile
a	1.7	1.7
b	4.4	4.4

expected output:

col1	percentile	percentile
a	1.5	1.7
b	4.0	4.4

[rileypeterson]

This works fine for me:

from functools import partial

import pandas as pd
import numpy as np

quant50 = partial(np.percentile, q=50)
quant50.__name__ = "quant50"
quant70 = partial(np.percentile, q=70)
quant70.__name__ = "quant70"

test = pd.DataFrame({'col1': ['a', 'a', 'b', 'b', 'b'], 'col2': [1,2,3,4,5]})
test.groupby('col1').agg({'col2': [quant50, quant70]})

Out[110]: 
        col2        
     quant50 quant70
col1                
a        1.5     1.7
b        4.0     4.4

Output of pd.show_versions()

INSTALLED VERSIONS
------------------
commit: None
python: 3.5.4.final.0
python-bits: 64
OS: Darwin
OS-release: 18.0.0
machine: x86_64
processor: i386
byteorder: little
LC_ALL: None
LANG: en_US.UTF-8
LOCALE: en_US.UTF-8

pandas: 0.24.2
pytest: 3.3.2
pip: 9.0.1
setuptools: 38.4.0
Cython: 0.27.3
numpy: 1.15.2
scipy: 1.1.0
pyarrow: None
xarray: None
IPython: 6.2.1
sphinx: 1.6.6
patsy: 0.5.1
dateutil: 2.6.1
pytz: 2017.3
blosc: None
bottleneck: None
tables: None
numexpr: None
feather: None
matplotlib: 3.0.3
openpyxl: 2.4.10
xlrd: 1.1.0
xlwt: 1.2.0
xlsxwriter: 1.0.2
lxml.etree: 4.1.1
bs4: 4.6.0
html5lib: 1.0.1
sqlalchemy: 1.2.1
pymysql: None
psycopg2: 2.7.4 (dt dec pq3 ext lo64)
jinja2: 2.10
s3fs: None
fastparquet: None
pandas_gbq: None
pandas_datareader: None
gcsfs: None

[aliraeisdanaei]

take

[rhshadrach]

I view specifying the __name__ as a temporary workaround, I don't think the behavior identified in the OP is correct, and needs to be fixed. It is quite subtle and silently leads to incorrect results.

@mroeschke for any thoughts.

[mroeschke]

That's a good point. Agreed it probably doesn't fix the original naming issue.