Merge pull request #10292 from sinhrks/tdi_freq

BUG: TimedeltaIndex slicing may reset freq

Author: jreback

doc/source/whatsnew/v0.16.2.txt

@@ -150,7 +150,7 @@ Bug Fixes
 
 - Bug in ``Series.align`` resets ``name`` when ``fill_value`` is specified (:issue:`10067`)
 - Bug in ``SparseSeries.abs`` resets ``name`` (:issue:`10241`)
-
+- Bug in ``TimedeltaIndex`` slicing may reset freq (:issue:`10292`)
 
 - Bug in GroupBy.get_group raises ValueError when group key contains NaT (:issue:`6992`)
 - Bug in ``SparseSeries`` constructor ignores input data name (:issue:`10258`)

pandas/tseries/base.py

@@ -69,6 +69,35 @@ def __contains__(self, key):
         except (KeyError, TypeError, ValueError):
             return False
 
+    def __getitem__(self, key):
+        getitem = self._data.__getitem__
+        if np.isscalar(key):
+            val = getitem(key)
+            return self._box_func(val)
+        else:
+            if com.is_bool_indexer(key):
+                key = np.asarray(key)
+                if key.all():
+                    key = slice(0, None, None)
+                else:
+                    key = lib.maybe_booleans_to_slice(key.view(np.uint8))
+
+            attribs = self._get_attributes_dict()
+
+            freq = None
+            if isinstance(key, slice):
+                if self.freq is not None and key.step is not None:
+                    freq = key.step * self.freq
+                else:
+                    freq = self.freq
+            attribs['freq'] = freq
+
+            result = getitem(key)
+            if result.ndim > 1:
+                return result
+
+            return self._simple_new(result, **attribs)
+
     @property
     def freqstr(self):
         """ return the frequency object as a string if its set, otherwise None """

pandas/tseries/index.py

@@ -1349,32 +1349,6 @@ def slice_indexer(self, start=None, end=None, step=None, kind=None):
             else:
                 raise
 
-    def __getitem__(self, key):
-        getitem = self._data.__getitem__
-        if np.isscalar(key):
-            val = getitem(key)
-            return Timestamp(val, offset=self.offset, tz=self.tz)
-        else:
-            if com.is_bool_indexer(key):
-                key = np.asarray(key)
-                if key.all():
-                    key = slice(0,None,None)
-                else:
-                    key = lib.maybe_booleans_to_slice(key.view(np.uint8))
-
-            new_offset = None
-            if isinstance(key, slice):
-                if self.offset is not None and key.step is not None:
-                    new_offset = key.step * self.offset
-                else:
-                    new_offset = self.offset
-
-            result = getitem(key)
-            if result.ndim > 1:
-                return result
-
-            return self._simple_new(result, self.name, new_offset, self.tz)
-
     # alias to offset
     def _get_freq(self):
         return self.offset

pandas/tseries/tdi.py

@@ -244,7 +244,7 @@ def _generate(cls, start, end, periods, name, offset, closed=None):
 
     @property
     def _box_func(self):
-        return lambda x: Timedelta(x,unit='ns')
+        return lambda x: Timedelta(x, unit='ns')
 
     @classmethod
     def _simple_new(cls, values, name=None, freq=None, **kwargs):
@@ -747,25 +747,6 @@ def _partial_td_slice(self, key, freq, use_lhs=True, use_rhs=True):
         # try to find a the dates
         return (lhs_mask & rhs_mask).nonzero()[0]
 
-    def __getitem__(self, key):
-        getitem = self._data.__getitem__
-        if np.isscalar(key):
-            val = getitem(key)
-            return Timedelta(val)
-        else:
-            if com.is_bool_indexer(key):
-                key = np.asarray(key)
-                if key.all():
-                    key = slice(0,None,None)
-                else:
-                    key = lib.maybe_booleans_to_slice(key.view(np.uint8))
-
-            result = getitem(key)
-            if result.ndim > 1:
-                return result
-
-            return self._simple_new(result, self.name)
-
     def searchsorted(self, key, side='left'):
         if isinstance(key, (np.ndarray, Index)):
             key = np.array(key, dtype=_TD_DTYPE, copy=False)

pandas/tseries/tests/test_base.py

@@ -297,6 +297,38 @@ def test_nonunique_contains(self):
                                        ['2015', '2015', '2016'], ['2015', '2015', '2014'])):
             tm.assertIn(idx[0], idx)
 
+    def test_getitem(self):
+        idx1 = pd.date_range('2011-01-01', '2011-01-31', freq='D', name='idx')
+        idx2 = pd.date_range('2011-01-01', '2011-01-31', freq='D', tz='Asia/Tokyo', name='idx')
+
+        for idx in [idx1, idx2]:
+            result = idx[0]
+            self.assertEqual(result, pd.Timestamp('2011-01-01', tz=idx.tz))
+
+            result = idx[0:5]
+            expected = pd.date_range('2011-01-01', '2011-01-05', freq='D',
+                                     tz=idx.tz, name='idx')
+            self.assert_index_equal(result, expected)
+            self.assertEqual(result.freq, expected.freq)
+
+            result = idx[0:10:2]
+            expected = pd.date_range('2011-01-01', '2011-01-09', freq='2D',
+                                     tz=idx.tz, name='idx')
+            self.assert_index_equal(result, expected)
+            self.assertEqual(result.freq, expected.freq)
+
+            result = idx[-20:-5:3]
+            expected = pd.date_range('2011-01-12', '2011-01-25', freq='3D',
+                                     tz=idx.tz, name='idx')
+            self.assert_index_equal(result, expected)
+            self.assertEqual(result.freq, expected.freq)
+
+            result = idx[4::-1]
+            expected = DatetimeIndex(['2011-01-05', '2011-01-04', '2011-01-03',
+                                      '2011-01-02', '2011-01-01'],
+                                     freq='-1D', tz=idx.tz, name='idx')
+            self.assert_index_equal(result, expected)
+            self.assertEqual(result.freq, expected.freq)
 
 class TestTimedeltaIndexOps(Ops):
 
@@ -742,6 +774,33 @@ def test_unknown_attribute(self):
         self.assertNotIn('foo',ts.__dict__.keys())
         self.assertRaises(AttributeError,lambda : ts.foo)
 
+    def test_getitem(self):
+        idx1 = pd.timedelta_range('1 day', '31 day', freq='D', name='idx')
+
+        for idx in [idx1]:
+            result = idx[0]
+            self.assertEqual(result, pd.Timedelta('1 day'))
+
+            result = idx[0:5]
+            expected = pd.timedelta_range('1 day', '5 day', freq='D', name='idx')
+            self.assert_index_equal(result, expected)
+            self.assertEqual(result.freq, expected.freq)
+
+            result = idx[0:10:2]
+            expected = pd.timedelta_range('1 day', '9 day', freq='2D', name='idx')
+            self.assert_index_equal(result, expected)
+            self.assertEqual(result.freq, expected.freq)
+
+            result = idx[-20:-5:3]
+            expected = pd.timedelta_range('12 day', '25 day', freq='3D', name='idx')
+            self.assert_index_equal(result, expected)
+            self.assertEqual(result.freq, expected.freq)
+
+            result = idx[4::-1]
+            expected = TimedeltaIndex(['5 day', '4 day', '3 day', '2 day', '1 day'],
+                                      freq='-1D', name='idx')
+            self.assert_index_equal(result, expected)
+            self.assertEqual(result.freq, expected.freq)
 
 class TestPeriodIndexOps(Ops):