BUG: Chained loc indexing reversal on dataframe with datetime index has different results than on dataframes with other index types

[ChrisDarley]

Pandas version checks

• 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 main branch of pandas.

Reproducible Example

import pandas as pd
import numpy as np
a = pd.DataFrame(
 data=np.array([[1,2,3,4],[10,20,30,40]]).transpose(),
 index = pd.date_range(start='2020-01-01', periods=4,freq='D'))
print(a.loc[::-1].loc[::-1])

Issue Description

A common way to reverse dataframes is by using df.loc[::-1]. If the dataframe has a datetime index, and you create a reversed view of the dataframe using df.loc[::-1], if you reverse that view again (i.e. df.loc[::-1].loc[::-1]), it only returns the last row of the original dataframe. This issue does not happen if you use df.iloc[::-1].iloc[::-1], nor does it happen with basic indexing such as df[::-1][::-1]. I think that this issue is specific to datetime indexes, because it did not occur when doing df.loc[::-1].loc[::-1] when I had a string index. I'm not sure that df.loc[::-1] is the preferred way to reverse a dataframe, but if double reversal works on a dataframe with a string index it should theoretically work on a a dataframe with a datetime index.

Expected Behavior

The same operation works fine on a dataframe with a string index

import pandas as pd
import numpy as np
a = pd.DataFrame(
data=np.array([[1,2,3,4],[10,20,30,40]]).transpose(),
index = ['a', 'b', 'c', 'd'])
print(a.loc[::-1].loc[::-1])

Installed Versions

INSTALLED VERSIONS

commit : 2e218d1
python : 3.10.6.final.0
python-bits : 64
OS : Linux
OS-release : 5.15.0-53-generic
Version : #59-Ubuntu SMP Mon Oct 17 18:53:30 UTC 2022
machine : x86_64
processor : x86_64
byteorder : little
LC_ALL : None
LANG : en_US.UTF-8
LOCALE : en_US.UTF-8

pandas : 1.5.3
numpy : 1.24.1
pytz : 2022.7
dateutil : 2.8.2
setuptools : 59.6.0
pip : 22.0.2
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : 4.9.2
html5lib : 1.1
pymysql : None
psycopg2 : None
jinja2 : 3.1.2
IPython : 8.7.0
pandas_datareader: None
bs4 : 4.11.1
bottleneck : None
brotli : None
fastparquet : None
fsspec : None
gcsfs : None
matplotlib : 3.6.2
numba : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : None
pyreadstat : None
pyxlsb : None
s3fs : None
scipy : 1.9.3
snappy : None
sqlalchemy : 1.4.45
tables : None
tabulate : None
xarray : None
xlrd : None
xlwt : None
zstandard : None
tzdata : None

[phofl]

Hi, thanks for your report. I think this is expected. slicing on DatetimeIndexes is only supposed to work on monotonic indexes. cc @jbrockmendel

[jbrockmendel]

Do they need to be monotonic increasing or any monotonic? seems like ::-1 is pretty unambiguous as to what we'd expect

[phofl]

Sorry, should have been more clear: monotonic increasing. We could certainly make this work, but the whole conversion into an indexer assumes monotonic increasing indexes

[topper-123]

If it doesn't/shouldn't work, then the operation should raise, e.g. a KeyError?