BUG: DatetimeIndex comparison handles NaT incorrectly

doc/source/v0.14.1.txt

@@ -237,6 +237,7 @@ Bug Fixes
 - Bug in when writing Stata files where the encoding was ignored (:issue:`7286`)
 
 
+- Bug in ``DatetimeIndex`` comparison doesn't handle ``NaT`` properly (:issue:`7529`)
 
 
 - Bug in passing input with ``tzinfo`` to some offsets ``apply``, ``rollforward`` or ``rollback`` resets ``tzinfo`` or raises ``ValueError`` (:issue:`7465`)

pandas/tseries/index.py

@@ -74,22 +74,35 @@ def wrapper(left, right):
     return wrapper
 
 
-def _dt_index_cmp(opname):
+def _dt_index_cmp(opname, nat_result=False):
     """
     Wrap comparison operations to convert datetime-like to datetime64
     """
     def wrapper(self, other):
         func = getattr(super(DatetimeIndex, self), opname)
-        if isinstance(other, datetime):
+        if isinstance(other, datetime) or isinstance(other, compat.string_types):
             other = _to_m8(other, tz=self.tz)
-        elif isinstance(other, list):
-            other = DatetimeIndex(other)
-        elif isinstance(other, compat.string_types):
-            other = _to_m8(other, tz=self.tz)
-        elif not isinstance(other, (np.ndarray, ABCSeries)):
-            other = _ensure_datetime64(other)
-        result = func(other)
+            result = func(other)
+            if com.isnull(other):
+                result.fill(nat_result)
+        else:
+            if isinstance(other, list):
+                other = DatetimeIndex(other)
+            elif not isinstance(other, (np.ndarray, ABCSeries)):
+                other = _ensure_datetime64(other)
+            result = func(other)
 
+            if isinstance(other, Index):
+                o_mask = other.values.view('i8') == tslib.iNaT
+            else:
+                o_mask = other.view('i8') == tslib.iNaT
+
+            if o_mask.any():
+                result[o_mask] = nat_result
+
+        mask = self.asi8 == tslib.iNaT
+        if mask.any():
+            result[mask] = nat_result
         return result.view(np.ndarray)
 
     return wrapper
@@ -142,7 +155,7 @@ class DatetimeIndex(DatetimeIndexOpsMixin, Int64Index):
     _arrmap = None
 
     __eq__ = _dt_index_cmp('__eq__')
-    __ne__ = _dt_index_cmp('__ne__')
+    __ne__ = _dt_index_cmp('__ne__', nat_result=True)
     __lt__ = _dt_index_cmp('__lt__')
     __gt__ = _dt_index_cmp('__gt__')
     __le__ = _dt_index_cmp('__le__')

pandas/tseries/period.py

@@ -498,16 +498,11 @@ def dt64arr_to_periodarr(data, freq, tz):
 
 # --- Period index sketch
 
-def _period_index_cmp(opname):
+def _period_index_cmp(opname, nat_result=False):
     """
     Wrap comparison operations to convert datetime-like to datetime64
     """
     def wrapper(self, other):
-        if opname == '__ne__':
-            fill_value = True
-        else:
-            fill_value = False
-
         if isinstance(other, Period):
             func = getattr(self.values, opname)
             if other.freq != self.freq:
@@ -523,7 +518,7 @@ def wrapper(self, other):
             mask = (com.mask_missing(self.values, tslib.iNaT) |
                     com.mask_missing(other.values, tslib.iNaT))
             if mask.any():
-                result[mask] = fill_value
+                result[mask] = nat_result
 
             return result
         else:
@@ -532,10 +527,10 @@ def wrapper(self, other):
             result = func(other.ordinal)
 
         if other.ordinal == tslib.iNaT:
-            result.fill(fill_value)
+            result.fill(nat_result)
         mask = self.values == tslib.iNaT
         if mask.any():
-            result[mask] = fill_value
+            result[mask] = nat_result
 
         return result
     return wrapper
@@ -595,7 +590,7 @@ class PeriodIndex(DatetimeIndexOpsMixin, Int64Index):
     _allow_period_index_ops = True
 
     __eq__ = _period_index_cmp('__eq__')
-    __ne__ = _period_index_cmp('__ne__')
+    __ne__ = _period_index_cmp('__ne__', nat_result=True)
     __lt__ = _period_index_cmp('__lt__')
     __gt__ = _period_index_cmp('__gt__')
     __le__ = _period_index_cmp('__le__')

pandas/tseries/tests/test_timeseries.py

@@ -2179,6 +2179,93 @@ def test_comparisons_coverage(self):
         exp = rng == rng
         self.assert_numpy_array_equal(result, exp)
 
+    def test_comparisons_nat(self):
+        fidx1 = pd.Index([1.0, np.nan, 3.0, np.nan, 5.0, 7.0])
+        fidx2 = pd.Index([2.0, 3.0, np.nan, np.nan, 6.0, 7.0])
+
+        didx1 = pd.DatetimeIndex(['2014-01-01', pd.NaT, '2014-03-01', pd.NaT,
+                                  '2014-05-01', '2014-07-01'])
+        didx2 = pd.DatetimeIndex(['2014-02-01', '2014-03-01', pd.NaT, pd.NaT,
+                                  '2014-06-01', '2014-07-01'])
+        darr = np.array([np.datetime64('2014-02-01 00:00Z'),
+                         np.datetime64('2014-03-01 00:00Z'),
+                         np.datetime64('nat'), np.datetime64('nat'),
+                         np.datetime64('2014-06-01 00:00Z'),
+                         np.datetime64('2014-07-01 00:00Z')])
+
+        if _np_version_under1p7:
+            # cannot test array because np.datetime('nat') returns today's date
+            cases = [(fidx1, fidx2), (didx1, didx2)]
+        else:
+            cases = [(fidx1, fidx2), (didx1, didx2), (didx1, darr)]
+
+        # Check pd.NaT is handles as the same as np.nan
+        for idx1, idx2 in cases:
+            result = idx1 < idx2
+            expected = np.array([True, False, False, False, True, False])
+            self.assert_numpy_array_equal(result, expected)
+            result = idx2 > idx1
+            expected = np.array([True, False, False, False, True, False])
+            self.assert_numpy_array_equal(result, expected)
+
+            result = idx1 <= idx2
+            expected = np.array([True, False, False, False, True, True])
+            self.assert_numpy_array_equal(result, expected)
+            result = idx2 >= idx1
+            expected = np.array([True, False, False, False, True, True])
+            self.assert_numpy_array_equal(result, expected)
+
+            result = idx1 == idx2
+            expected = np.array([False, False, False, False, False, True])
+            self.assert_numpy_array_equal(result, expected)
+
+            result = idx1 != idx2
+            expected = np.array([True, True, True, True, True, False])
+            self.assert_numpy_array_equal(result, expected)
+
+        for idx1, val in [(fidx1, np.nan), (didx1, pd.NaT)]:
+            result = idx1 < val
+            expected = np.array([False, False, False, False, False, False])
+            self.assert_numpy_array_equal(result, expected)
+            result = idx1 > val
+            self.assert_numpy_array_equal(result, expected)
+
+            result = idx1 <= val
+            self.assert_numpy_array_equal(result, expected)
+            result = idx1 >= val
+            self.assert_numpy_array_equal(result, expected)
+
+            result = idx1 == val
+            self.assert_numpy_array_equal(result, expected)
+
+            result = idx1 != val
+            expected = np.array([True, True, True, True, True, True])
+            self.assert_numpy_array_equal(result, expected)
+
+        # Check pd.NaT is handles as the same as np.nan
+        for idx1, val in [(fidx1, 3), (didx1, datetime(2014, 3, 1))]:
+            result = idx1 < val
+            expected = np.array([True, False, False, False, False, False])
+            self.assert_numpy_array_equal(result, expected)
+            result = idx1 > val
+            expected = np.array([False, False, False, False, True, True])
+            self.assert_numpy_array_equal(result, expected)
+
+            result = idx1 <= val
+            expected = np.array([True, False, True, False, False, False])
+            self.assert_numpy_array_equal(result, expected)
+            result = idx1 >= val
+            expected = np.array([False, False, True, False, True, True])
+            self.assert_numpy_array_equal(result, expected)
+
+            result = idx1 == val
+            expected = np.array([False, False, True, False, False, False])
+            self.assert_numpy_array_equal(result, expected)
+
+            result = idx1 != val
+            expected = np.array([True, True, False, True, True, True])
+            self.assert_numpy_array_equal(result, expected)
+
     def test_map(self):
         rng = date_range('1/1/2000', periods=10)