DEPR: Timestamp(dt64obj, tz=tz) (#49381)

Author: jbrockmendel

doc/source/whatsnew/v2.0.0.rst

@@ -295,6 +295,7 @@ Removal of prior version deprecations/changes
 - Changed behavior of :class:`DataFrame` constructor given floating-point ``data`` and an integer ``dtype``, when the data cannot be cast losslessly, the floating point dtype is retained, matching :class:`Series` behavior (:issue:`41170`)
 - Changed behavior of :class:`DataFrame` constructor when passed a ``dtype`` (other than int) that the data cannot be cast to; it now raises instead of silently ignoring the dtype (:issue:`41733`)
 - Changed the behavior of :class:`Series` constructor, it will no longer infer a datetime64 or timedelta64 dtype from string entries (:issue:`41731`)
+- Changed behavior of :class:`Timestamp` constructor with a ``np.datetime64`` object and a ``tz`` passed to interpret the input as a wall-time as opposed to a UTC time (:issue:`42288`)
 - Changed behavior of :class:`Index` constructor when passed a ``SparseArray`` or ``SparseDtype`` to retain that dtype instead of casting to ``numpy.ndarray`` (:issue:`43930`)
 - Removed the deprecated ``base`` and ``loffset`` arguments from :meth:`pandas.DataFrame.resample`, :meth:`pandas.Series.resample` and :class:`pandas.Grouper`. Use ``offset`` or ``origin`` instead (:issue:`31809`)
 - Changed behavior of :meth:`DataFrame.any` and :meth:`DataFrame.all` with ``bool_only=True``; object-dtype columns with all-bool values will no longer be included, manually cast to ``bool`` dtype first (:issue:`46188`)

pandas/_libs/tslibs/timestamps.pyx

@@ -1637,18 +1637,9 @@ class Timestamp(_Timestamp):
 
         tzobj = maybe_get_tz(tz)
         if tzobj is not None and is_datetime64_object(ts_input):
-            # GH#24559, GH#42288 In the future we will treat datetime64 as
+            # GH#24559, GH#42288 As of 2.0 we treat datetime64 as
             #  wall-time (consistent with DatetimeIndex)
-            warnings.warn(
-                "In a future version, when passing a np.datetime64 object and "
-                "a timezone to Timestamp, the datetime64 will be interpreted "
-                "as a wall time, not a UTC time.  To interpret as a UTC time, "
-                "use `Timestamp(dt64).tz_localize('UTC').tz_convert(tz)`",
-                FutureWarning,
-                stacklevel=find_stack_level(),
-            )
-            # Once this deprecation is enforced, we can do
-            #  return Timestamp(ts_input).tz_localize(tzobj)
+            return cls(ts_input).tz_localize(tzobj)
 
         if nanosecond is None:
             nanosecond = 0

pandas/tests/scalar/timestamp/test_constructors.py

@@ -54,18 +54,13 @@ def test_constructor_datetime64_with_tz(self):
         dt = np.datetime64("1970-01-01 05:00:00")
         tzstr = "UTC+05:00"
 
-        msg = "interpreted as a wall time"
-        with tm.assert_produces_warning(FutureWarning, match=msg):
-            ts = Timestamp(dt, tz=tzstr)
+        # pre-2.0 this interpreted dt as a UTC time. in 2.0 this is treated
+        #  as a wall-time, consistent with DatetimeIndex behavior
+        ts = Timestamp(dt, tz=tzstr)
 
-        # Check that we match the old behavior
-        alt = Timestamp(dt).tz_localize("UTC").tz_convert(tzstr)
+        alt = Timestamp(dt).tz_localize(tzstr)
         assert ts == alt
-
-        # Check that we *don't* match the future behavior
-        assert ts.hour != 5
-        expected_future = Timestamp(dt).tz_localize(tzstr)
-        assert ts != expected_future
+        assert ts.hour == 5
 
     def test_constructor(self):
         base_str = "2014-07-01 09:00"