BUG: indexing with monotonic decreasing DTI

[ruiyuanlu]

Hi there, I'm trying to add a column of time using pandas. But some dates doesn't work fine in string format. I'm working on jupyter notebook with Anaconda 4.3.0

Here's what I'm trying to do:

data1 = data.head()
date_list = ["2018-01-02", "2017-02-10","2016-03-10", "2015-03-15", "2014-03-16"]
date_index = pd.to_datetime(date_list)
data1['date'] = date_index
data1

And the output is shown below.

Then I try to set 'date' as the index of this data frame.

data1 = data1.set_index("date")
print(data1.index)
data1

This looks reasonable, too.

Something weird happens when I try to index data using:

data1.loc['2018-01-02']

It raises KeyError!!!

But other dates work fine. For example:

data1.loc['2015-03-15']

Why is that?

Output of pd.show_versions()

[paste the output of pd.show_versions() here below this line]
INSTALLED VERSIONS

commit: None
python: 3.6.0.final.0
python-bits: 64
OS: Windows
OS-release: 10
machine: AMD64
processor: Intel64 Family 6 Model 94 Stepping 3, GenuineIntel
byteorder: little
LC_ALL: None
LANG: None
LOCALE: None.None

pandas: 0.19.2
nose: 1.3.7
pip: 9.0.1
setuptools: 27.2.0
Cython: 0.25.2
numpy: 1.13.3
scipy: 0.18.1
statsmodels: 0.6.1
xarray: None
IPython: 5.1.0
sphinx: 1.5.1
patsy: 0.4.1
dateutil: 2.6.0
pytz: 2016.10
blosc: None
bottleneck: 1.2.0
tables: 3.2.2
numexpr: 2.6.1
matplotlib: 2.0.0
openpyxl: 2.4.1
xlrd: 1.0.0
xlwt: 1.2.0
xlsxwriter: 0.9.6
lxml: 3.7.2
bs4: 4.5.3
html5lib: 0.9999999
httplib2: None
apiclient: None
sqlalchemy: 1.1.5
pymysql: None
psycopg2: None
jinja2: 2.9.4
boto: 2.45.0
pandas_datareader: None

[TomAugspurger]

Can you try on a more recent version of pandas to see if your issue is already fixed?

FYI, text tracebacks are generally more helpful than images.

[jreback]

Here's a repro. looks like buggy on 1st element of a montononic descending partial string indexing (though here its an exact match).

In [1]: df = pd.DataFrame({'A': [1,2,3]},index=pd.date_range('20170101', periods=3)[::-1])

In [2]: df
Out[2]: 
            A
2017-01-03  1
2017-01-02  2
2017-01-01  3

In [3]: df.loc['2017-01-03']
---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
~/pandas/pandas/core/indexing.py in _has_valid_type(self, key, axis)
   1501                 if not ax.contains(key):
-> 1502                     error()
   1503             except TypeError as e:

~/pandas/pandas/core/indexing.py in error()
   1496                                .format(key=key,
-> 1497                                        axis=self.obj._get_axis_name(axis)))
   1498 

KeyError: 'the label [2017-01-03] is not in the [index]'

During handling of the above exception, another exception occurred:

KeyError                                  Traceback (most recent call last)
<ipython-input-3-5dd0bbdc6e4e> in <module>()
----> 1 df.loc['2017-01-03']

~/pandas/pandas/core/indexing.py in __getitem__(self, key)
   1367 
   1368             maybe_callable = com._apply_if_callable(key, self.obj)
-> 1369             return self._getitem_axis(maybe_callable, axis=axis)
   1370 
   1371     def _is_scalar_access(self, key):

~/pandas/pandas/core/indexing.py in _getitem_axis(self, key, axis)
   1620 
   1621         # fall thru to straight lookup
-> 1622         self._has_valid_type(key, axis)
   1623         return self._get_label(key, axis=axis)
   1624 

~/pandas/pandas/core/indexing.py in _has_valid_type(self, key, axis)
   1508                 raise
   1509             except:
-> 1510                 error()
   1511 
   1512         return True

~/pandas/pandas/core/indexing.py in error()
   1495                 raise KeyError(u"the label [{key}] is not in the [{axis}]"
   1496                                .format(key=key,
-> 1497                                        axis=self.obj._get_axis_name(axis)))
   1498 
   1499             try:

KeyError: 'the label [2017-01-03] is not in the [index]'

In [4]: df.loc['2017-01-02']
Out[4]: 
            A
2017-01-02  2

[swyoon]

may i work on this issue?

[jreback]

@swyoon absolutely!

[mapehe]

Seems that this is caused by the fact that np.any([0]) is False. Added a suggested solution in #20677.