BUG: to_datetime with out-of-bounds np.datetime64 and errors='ignore' doesn't return input (#50992)

MarcoGorelli · mroeschke · web-flow · commit 24a9073273e7 · 2023-02-02T08:42:31.000-08:00
* fixup ignore inconsistency

* add unit test, clarify whatsnew

* simplify

Co-authored-by: Matthew Roeschke &lt;10647082+mroeschke@users.noreply.github.com&gt;

---------

Co-authored-by: MarcoGorelli &lt;&gt;
Co-authored-by: Matthew Roeschke &lt;10647082+mroeschke@users.noreply.github.com&gt;
diff --git a/doc/source/whatsnew/v2.0.0.rst b/doc/source/whatsnew/v2.0.0.rst
@@ -1102,6 +1102,7 @@ Datetimelike
 - Bug in :func:`to_datetime` was raising ``ValueError`` when parsing mixed-offset :class:`Timestamp` with ``errors='ignore'`` (:issue:`50585`)
 - Bug in :func:`to_datetime` was incorrectly handling floating-point inputs within 1 ``unit`` of the overflow boundaries (:issue:`50183`)
 - Bug in :func:`to_datetime` with unit of "Y" or "M" giving incorrect results, not matching pointwise :class:`Timestamp` results (:issue:`50870`)
+- Bug in :func:`to_datetime` was not returning input with ``errors='ignore'`` when input was out-of-bounds (:issue:`50587`)
 -
 
 Timedelta
diff --git a/pandas/_libs/tslib.pyx b/pandas/_libs/tslib.pyx
@@ -34,7 +34,6 @@ from pandas._libs.tslibs.np_datetime cimport (
     NPY_DATETIMEUNIT,
     NPY_FR_ns,
     check_dts_bounds,
-    get_datetime64_value,
     npy_datetimestruct,
     npy_datetimestruct_to_datetime,
     pandas_datetime_to_datetimestruct,
@@ -542,16 +541,6 @@ cpdef array_to_datetime(
 
             cnp.PyArray_MultiIter_NEXT(mi)
 
-        except OutOfBoundsDatetime as ex:
-            ex.args = (f"{ex}, at position {i}",)
-            if is_coerce:
-                iresult[i] = NPY_NAT
-                cnp.PyArray_MultiIter_NEXT(mi)
-                continue
-            elif is_raise:
-                raise
-            return ignore_errors_out_of_bounds_fallback(values), tz_out
-
         except (TypeError, OverflowError, ValueError) as ex:
             ex.args = (f"{ex}, at position {i}",)
             if is_coerce:
@@ -578,55 +567,6 @@ cpdef array_to_datetime(
     return result, tz_out
 
 
-@cython.wraparound(False)
-@cython.boundscheck(False)
-cdef ndarray ignore_errors_out_of_bounds_fallback(ndarray values):
-    """
-    Fallback for array_to_datetime if an OutOfBoundsDatetime is raised
-    and errors == "ignore"
-
-    Parameters
-    ----------
-    values : ndarray[object]
-
-    Returns
-    -------
-    ndarray[object]
-    """
-    cdef:
-        Py_ssize_t i, n = values.size
-        object val
-        cnp.broadcast mi
-        ndarray[object] oresult
-        ndarray oresult_nd
-
-    oresult_nd = cnp.PyArray_EMPTY(values.ndim, values.shape, cnp.NPY_OBJECT, 0)
-    mi = cnp.PyArray_MultiIterNew2(oresult_nd, values)
-    oresult = oresult_nd.ravel()
-
-    for i in range(n):
-        # Analogous to `val = values[i]`
-        val = <object>(<PyObject**>cnp.PyArray_MultiIter_DATA(mi, 1))[0]
-
-        # set as nan except if its a NaT
-        if checknull_with_nat_and_na(val):
-            if isinstance(val, float):
-                oresult[i] = np.nan
-            else:
-                oresult[i] = <object>NaT
-        elif is_datetime64_object(val):
-            if get_datetime64_value(val) == NPY_NAT:
-                oresult[i] = <object>NaT
-            else:
-                oresult[i] = val.item()
-        else:
-            oresult[i] = val
-
-        cnp.PyArray_MultiIter_NEXT(mi)
-
-    return oresult
-
-
 @cython.wraparound(False)
 @cython.boundscheck(False)
 cdef _array_to_datetime_object(
diff --git a/pandas/tests/tools/test_to_datetime.py b/pandas/tests/tools/test_to_datetime.py
@@ -1084,9 +1084,15 @@ def test_to_datetime_array_of_dt64s(self, cache, unit):
         # numpy is either a python datetime.datetime or datetime.date
         tm.assert_index_equal(
             to_datetime(dts_with_oob, errors="ignore", cache=cache),
-            Index([dt.item() for dt in dts_with_oob]),
+            Index(dts_with_oob),
         )
 
+    def test_out_of_bounds_errors_ignore(self):
+        # https://github.com/pandas-dev/pandas/issues/50587
+        result = to_datetime(np.datetime64("9999-01-01"), errors="ignore")
+        expected = np.datetime64("9999-01-01")
+        assert result == expected
+
     def test_to_datetime_tz(self, cache):
 
         # xref 8260

Original file line number	Diff line number	Diff line change
`@@ -1102,6 +1102,7 @@ Datetimelike`
`1102`	`1102`	- Bug in :func:`to_datetime` was raising ``ValueError`` when parsing mixed-offset :class:`Timestamp` with ``errors='ignore'`` (:issue:`50585`)
`1103`	`1103`	- Bug in :func:`to_datetime` was incorrectly handling floating-point inputs within 1 ``unit`` of the overflow boundaries (:issue:`50183`)
`1104`	`1104`	- Bug in :func:`to_datetime` with unit of "Y" or "M" giving incorrect results, not matching pointwise :class:`Timestamp` results (:issue:`50870`)
	`1105`	+- Bug in :func:`to_datetime` was not returning input with ``errors='ignore'`` when input was out-of-bounds (:issue:`50587`)
`1105`	`1106`	`-`
`1106`	`1107`
`1107`	`1108`	`Timedelta`