DEPR: Series.astype(np.datetime64) (pandas-dev#48555)

doc/source/whatsnew/v1.5.0.rst

@@ -936,6 +936,7 @@ Other Deprecations
 - Deprecated the ``sort_columns`` argument in :meth:`DataFrame.plot` and :meth:`Series.plot` (:issue:`47563`).
 - Deprecated positional arguments for all but the first argument of :meth:`DataFrame.to_stata` and :func:`read_stata`, use keyword arguments instead (:issue:`48128`).
 - Deprecated the ``mangle_dupe_cols`` argument in :func:`read_csv`, :func:`read_fwf`, :func:`read_table` and :func:`read_excel`. The argument was never implemented, and a new argument where the renaming pattern can be specified will be added instead (:issue:`47718`)
+- Deprecated allowing ``dtype='datetime64'`` or ``dtype=np.datetime64`` in :meth:`Series.astype`, use "datetime64[ns]" instead (:issue:`47844`)
 
 .. ---------------------------------------------------------------------------
 .. _whatsnew_150.performance:

pandas/core/arrays/datetimes.py

@@ -632,6 +632,22 @@ def astype(self, dtype, copy: bool = True):
             return type(self)._simple_new(res_values, dtype=dtype)
             # TODO: preserve freq?
 
+        elif (
+            self.tz is None
+            and is_datetime64_dtype(dtype)
+            and dtype != self.dtype
+            and is_unitless(dtype)
+        ):
+            # TODO(2.0): just fall through to dtl.DatetimeLikeArrayMixin.astype
+            warnings.warn(
+                "Passing unit-less datetime64 dtype to .astype is deprecated "
+                "and will raise in a future version. Pass 'datetime64[ns]' instead",
+                FutureWarning,
+                stacklevel=find_stack_level(inspect.currentframe()),
+            )
+            # unit conversion e.g. datetime64[s]
+            return self._ndarray.astype(dtype)
+
         elif is_period_dtype(dtype):
             return self.to_period(freq=dtype.freq)
         return dtl.DatetimeLikeArrayMixin.astype(self, dtype, copy)

pandas/core/indexes/base.py

@@ -42,6 +42,7 @@
     IncompatibleFrequency,
     OutOfBoundsDatetime,
     Timestamp,
+    is_unitless,
     tz_compare,
 )
 from pandas._typing import (
@@ -1086,6 +1087,11 @@ def astype(self, dtype, copy: bool = True):
 
         values = self._data
         if isinstance(values, ExtensionArray):
+            if isinstance(dtype, np.dtype) and dtype.kind == "M" and is_unitless(dtype):
+                # TODO(2.0): remove this special-casing once this is enforced
+                #  in DTA.astype
+                raise TypeError(f"Cannot cast {type(self).__name__} to dtype")
+
             with rewrite_exception(type(values).__name__, type(self).__name__):
                 new_values = values.astype(dtype, copy=copy)
 

pandas/tests/series/methods/test_astype.py

@@ -30,6 +30,19 @@
 
 
 class TestAstypeAPI:
+    def test_astype_unitless_dt64_deprecated(self):
+        # GH#47844
+        ser = Series(["1970-01-01", "1970-01-01", "1970-01-01"], dtype="datetime64[ns]")
+
+        msg = "Passing unit-less datetime64 dtype to .astype is deprecated and "
+        with tm.assert_produces_warning(FutureWarning, match=msg):
+            res = ser.astype(np.datetime64)
+        tm.assert_series_equal(ser, res)
+
+        with tm.assert_produces_warning(FutureWarning, match=msg):
+            res = ser.astype("datetime64")
+        tm.assert_series_equal(ser, res)
+
     def test_arg_for_errors_in_astype(self):
         # see GH#14878
         ser = Series([1, 2, 3])