BUG: Revisit Index.delete preserves freq

[sinhrks]

Fix the issue that freq is preserved inappropriately, caused by #7302. Fix preserves freq if edge element is being deleted (same as #7299).
CC: @rosnfeld

[jreback]

ok looks good thanks

[rosnfeld]

I tried the same exercise as before, and it didn't work for me:

In [1]: ts = pd.Series(1, index=pd.date_range('2000-01-01', periods=10, freq='D'))

In [2]: ts[:5].index
Out[2]: 
<class 'pandas.tseries.index.DatetimeIndex'>
[2000-01-01, ..., 2000-01-05]
Length: 5, Freq: D, Timezone: None

In [3]: ts[[1,2,3]].index
Out[3]: 
<class 'pandas.tseries.index.DatetimeIndex'>
[2000-01-02, ..., 2000-01-04]
Length: 3, Freq: None, Timezone: None

In [4]: ts.drop(ts.index[[5,6,7]]).index
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-4-3de5d002e589> in <module>()
----> 1 ts.drop(ts.index[[5,6,7]]).index

/home/andrew/git/pandas-rosnfeld/pandas/core/generic.py in drop(self, labels, axis, level, inplace, **kwargs)
   1462                 new_axis = axis.drop(labels, level=level)
   1463             else:
-> 1464                 new_axis = axis.drop(labels)
   1465             dropped = self.reindex(**{axis_name: new_axis})
   1466             try:

/home/andrew/git/pandas-rosnfeld/pandas/core/index.py in drop(self, labels)
   1801         if mask.any():
   1802             raise ValueError('labels %s not contained in axis' % labels[mask])
-> 1803         return self.delete(indexer)
   1804 
   1805 

/home/andrew/git/pandas-rosnfeld/pandas/tseries/index.py in delete(self, loc)
   1618         """
   1619         new_dates = np.delete(self.asi8, loc)
-> 1620         if loc in (0, -len(self), -1, len(self) - 1):
   1621             freq = self.freq
   1622         else:

ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

[sinhrks]

Thanks for your confirmation. I didn't imagine loc can be list or array.. I'll reconsider the logic.

[jreback]

@sinhrks is this updated with @rosnfeld example?

[sinhrks]

@jreback @rosnfeld Updated to accept int, slice and np.array as the same as np.delete, and added test cases for these.

[jreback]

@rosnfeld can you give this a try?

[rosnfeld]

Looks good.

[jreback]

thanks @sinhrks !