BUG: Pandas merge DFs on index and column

[carlosg-m]

• 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.

---

Code Sample, a copy-pastable example

# Your code here
df = pd.DataFrame([0.5,0.6,6,6,7,12],index=[0,0,1,1,1,2], columns=['distance'])

print(df)

   distance
0       0.5
0       0.6
1       6.0
1       6.0
1       7.0
2      12.0

result = df.merge(df.groupby(df.index)['distance'].min(), left_index=True, right_index=True, on='distance')

print(result)

   distance
0       0.5
0       0.6
1       6.0
1       6.0
1       7.0
2      12.0

Problem description

Merging two DataFrames using left and right index and a column does not seem to be possible.

I only have two options, reset the index on both DFs and merge on both columns which is slower, or create a multiindex which is unnecessary.

Expected Output

result = df.reset_index().merge(df.groupby(df.index)['distance'].min().reset_index(), on=['index','distance'])

print(result)

   index  distance
0      0       0.5
1      1       6.0
2      1       6.0
3      2      12.0

Output of pd.show_versions()

INSTALLED VERSIONS

commit : d9fff27
python : 3.6.9.final.0
python-bits : 64
OS : Linux
OS-release : 4.19.112+
Version : #1 SMP Thu Jul 23 08:00:38 PDT 2020
machine : x86_64
processor : x86_64
byteorder : little
LC_ALL : None
LANG : en_US.UTF-8
LOCALE : en_US.UTF-8

pandas : 1.1.0
numpy : 1.18.5
pytz : 2018.9
dateutil : 2.8.1
pip : 19.3.1
setuptools : 49.2.0
Cython : 0.29.21
pytest : 3.6.4
hypothesis : None
sphinx : 1.8.5
blosc : None
feather : 0.4.1
xlsxwriter : None
lxml.etree : 4.2.6
html5lib : 1.0.1
pymysql : None
psycopg2 : 2.7.6.1 (dt dec pq3 ext lo64)
jinja2 : 2.11.2
IPython : 5.5.0
pandas_datareader: None
bs4 : 4.6.3
bottleneck : 1.3.2
fsspec : 0.8.0
fastparquet : None
gcsfs : None
matplotlib : 3.2.2
numexpr : 2.7.1
odfpy : None
openpyxl : 2.5.9
pandas_gbq : 0.11.0
pyarrow : 0.14.1
pytables : None
pyxlsb : None
s3fs : 0.4.2
scipy : 1.4.1
sqlalchemy : 1.3.18
tables : 3.4.4
tabulate : 0.8.7
xarray : 0.15.1
xlrd : 1.1.0
xlwt : 1.3.0
numba : 0.48.0

[dsaxton]

I think this is just how the API is written, which would require making the index a column if that's what you want to do.

I'm not sure what the specific use case is here but a faster / equivalent way to do this is to use a groupby.transform window function:

mask = df.groupby(df.index)["distance"].transform(np.min) == df["distance"]                                                                                                                         
print(df[mask])                                                                                                                                                                                     
#    distance
# 0       0.5
# 1       6.0
# 1       6.0
# 2      12.0

[carlosg-m]

I think this is just how the API is written, which would require making the index a column if that's what you want to do.

I'm not sure what the specific use case is here but a faster / equivalent way to do this is to use a groupby.transform window function:

mask = df.groupby(df.index)["distance"].transform(np.min) == df["distance"]                                                                                                                         
print(df[mask])                                                                                                                                                                                     
#    distance
# 0       0.5
# 1       6.0
# 1       6.0
# 2      12.0

Thank you!

[simonjayhawkins]

closing as answered

[simonjayhawkins]

xref #16228 (comment)