BUG: sort_index method ignores ascending parameter of type tuple

[vezeli]

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

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

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

Reproducible Example

import pandas as pd

df = pd.DataFrame({
    "animal": ["dog", "duck", "horse", "penguin", "kangaroo"], 
    "legs": [4, 2, 4, 2, 2], 
    "class": ["mammal", "bird", "mammal", "bird", "mammal"]
})

df.set_index(["class", "animal"], inplace=True)

sorted1 = df.sort_index(level=(0, 1), ascending=[True, False])
sorted2 = df.sort_index(level=(0, 1), ascending=(True, False))

Issue Description

In the above example, the sort_index method on DataFrame object does not sort MultiIndex properly when the ascending parameter is passed as tuple. The sorted2.index should be sorted in ascending order for level 0 (class) and descending order for level 1 (animal). However, this is not the case and level 1 is also ascending:

>>> print(sorted2.index)
MultiIndex([(  'bird',     'duck'),
            (  'bird',  'penguin'),
            ('mammal',      'dog'),
            ('mammal',    'horse'),
            ('mammal', 'kangaroo')],
           names=['class', 'animal'])

When the ascending parameter is list than the sort_index method works as expected such that level 0 is ascending and level 1 is descending:

>>> print(sorted1.index)
MultiIndex([(  'bird',  'penguin'),
            (  'bird',     'duck'),
            ('mammal', 'kangaroo'),
            ('mammal',    'horse'),
            ('mammal',      'dog')],
           names=['class', 'animal'])

Expected Behavior

I would expect sorted1.index and sorted2.index to be equal (i.e., include a test case with assert sorted1.index == sorted2.index that passes).

Installed Versions

INSTALLED VERSIONS

commit : b5958ee
python : 3.6.9.final.0
python-bits : 64
OS : Linux
OS-release : 5.4.0-84-generic
Version : #94~18.04.1-Ubuntu SMP Thu Aug 26 23:17:46 UTC 2021
machine : x86_64
processor : x86_64
byteorder : little
LC_ALL : None
LANG : en_US.UTF-8
LOCALE : en_US.UTF-8

pandas : 1.1.5
numpy : 1.19.5
pytz : 2021.3
dateutil : 2.8.2
pip : 21.2.4
setuptools : 58.2.0
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : None
IPython : 7.16.1
pandas_datareader: None
bs4 : None
bottleneck : None
fsspec : None
fastparquet : None
gcsfs : None
matplotlib : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : None
pytables : None
pyxlsb : None
s3fs : None
scipy : None
sqlalchemy : None
tables : None
tabulate : None
xarray : None
xlrd : None
xlwt : None
numba : None

[attack68]

The docs confirm list-like should be accepted. Would you be interested in taking a look and submitting a PR?

[phofl]

This works on master, may need tests

[phofl]

This started working on 1.2. @vezeli please don't check the box persist on the latest version of pandas if you did not check

[vezeli]

Sorry about that @phofl I thought I was using the latest version but I was probably in wrong virtual environment... If it's OK with you I could submit a PR with a test that provides ascending param with a tuple?

[phofl]

Yeah for sure. You could also check if there is an existing test and parametrize that one. Not sure if there is though. Have checked yesterday if we have a test for tuples but could not find anything