BUG: scalar indexing on 2D DTA/TDA/PA (#33342)

Author: jbrockmendel

pandas/core/arrays/datetimelike.py

@@ -574,6 +574,8 @@ def __getitem__(self, key):
                 freq = self.freq
 
         result = getitem(key)
+        if lib.is_scalar(result):
+            return self._box_func(result)
         return self._simple_new(result, dtype=self.dtype, freq=freq)
 
     def __setitem__(

pandas/core/internals/blocks.py

@@ -6,7 +6,7 @@
 
 import numpy as np
 
-from pandas._libs import NaT, Timestamp, algos as libalgos, lib, writers
+from pandas._libs import NaT, algos as libalgos, lib, writers
 import pandas._libs.internals as libinternals
 from pandas._libs.tslibs import Timedelta, conversion
 from pandas._libs.tslibs.timezones import tz_compare
@@ -2009,12 +2009,7 @@ def array_values(self):
     def iget(self, key):
         # GH#31649 we need to wrap scalars in Timestamp/Timedelta
         # TODO(EA2D): this can be removed if we ever have 2D EA
-        result = super().iget(key)
-        if isinstance(result, np.datetime64):
-            result = Timestamp(result)
-        elif isinstance(result, np.timedelta64):
-            result = Timedelta(result)
-        return result
+        return self.array_values().reshape(self.shape)[key]
 
     def shift(self, periods, axis=0, fill_value=None):
         # TODO(EA2D) this is unnecessary if these blocks are backed by 2D EAs

pandas/tests/arrays/test_datetimelike.py

@@ -222,6 +222,11 @@ def test_getitem_2d(self, arr1d):
         result = arr2d[:3, 0]
         tm.assert_equal(result, expected)
 
+        # Scalar lookup
+        result = arr2d[-1, 0]
+        expected = arr1d[-1]
+        assert result == expected
+
     def test_setitem(self):
         data = np.arange(10, dtype="i8") * 24 * 3600 * 10 ** 9
         arr = self.array_cls(data, freq="D")