Rename in MultiIndex flattens index

[hayd]

When trying to rename something in a MultiIndex it doesn't work and flattens the index:

df = pd.DataFrame({'Data1': {0: 'present', 1: 'absent'}, 'Data2': {0: 'absent', 1: 'present'}, 'Location': {0: 'loc1', 1: 'loc2'}, 'Name': {0: 'Foo', 1: 'Foo2'}, 'Position': {0: 12345, 1: 67890}}).set_index(['Name', 'Location', 'Position'])

In [2]: df
Out[2]: 
                          Data1    Data2
Name Location Position                  
Foo  loc1     12345     present   absent
Foo2 loc2     67890      absent  present

In [3]: df.rename({'Foo2': 'Bar'})
Out[3]: 
                       Data1    Data2
(Foo, loc1, 12345)   present   absent
(Foo2, loc2, 67890)   absent  present

In [4]: df.rename_axis({'Foo2': 'Bar'}, axis=1)
Out[4]: 
                       Data1    Data2
(Foo, loc1, 12345)   present   absent
(Foo2, loc2, 67890)   absent  present

Taken from this SO question.

[waitingkuo]

Seems the name of the index is gone after rename for both Series and DataFrame

In [1]: import pandas as pd
In [2]: sr = pd.Series([1, 2, 3], pd.Index(['a', 'b', 'c'], name='test'))

In [3]: sr
Out[4]: 
test
a       1
b       2
c       3

In [4]: sr.rename({'a': 'd'})
Out[4]: 
d    1
b    2
c    3

[hayd]

I've added it to my list of "things to fix with index/column names" :)