update frame and series tests

jorisvandenbossche · jorisvandenbossche · commit 21954fbb7cef · 2025-08-12T10:37:59.000+02:00
diff --git a/pandas/conftest.py b/pandas/conftest.py
@@ -935,7 +935,7 @@ def rand_series_with_duplicate_datetimeindex() -> Series:
         (Period("2012-02-01", freq="D"), "period[D]"),
         (
             Timestamp("2011-01-01", tz="US/Eastern"),
-            DatetimeTZDtype(unit="s", tz="US/Eastern"),
+            DatetimeTZDtype(unit="us", tz="US/Eastern"),
         ),
         (Timedelta(seconds=500), "timedelta64[ns]"),
     ]
diff --git a/pandas/tests/frame/indexing/test_indexing.py b/pandas/tests/frame/indexing/test_indexing.py
@@ -825,7 +825,7 @@ def test_setitem_single_column_mixed_datetime(self):
         # check our dtypes
         result = df.dtypes
         expected = Series(
-            [np.dtype("float64")] * 3 + [np.dtype("datetime64[s]")],
+            [np.dtype("float64")] * 3 + [np.dtype("datetime64[us]")],
             index=["foo", "bar", "baz", "timestamp"],
         )
         tm.assert_series_equal(result, expected)
diff --git a/pandas/tests/frame/indexing/test_setitem.py b/pandas/tests/frame/indexing/test_setitem.py
@@ -260,7 +260,7 @@ def test_setitem_dict_preserves_dtypes(self):
             (Interval(left=0, right=5), IntervalDtype("int64", "right")),
             (
                 Timestamp("2011-01-01", tz="US/Eastern"),
-                DatetimeTZDtype(unit="s", tz="US/Eastern"),
+                DatetimeTZDtype(unit="us", tz="US/Eastern"),
             ),
         ],
     )
diff --git a/pandas/tests/frame/methods/test_combine_first.py b/pandas/tests/frame/methods/test_combine_first.py
@@ -197,7 +197,7 @@ def test_combine_first_align_nan(self):
         # GH 7509 (not fixed)
         dfa = DataFrame([[pd.Timestamp("2011-01-01"), 2]], columns=["a", "b"])
         dfb = DataFrame([[4], [5]], columns=["b"])
-        assert dfa["a"].dtype == "datetime64[s]"
+        assert dfa["a"].dtype == "datetime64[us]"
         assert dfa["b"].dtype == "int64"
 
         res = dfa.combine_first(dfb)
@@ -206,7 +206,7 @@ def test_combine_first_align_nan(self):
             columns=["a", "b"],
         )
         tm.assert_frame_equal(res, exp)
-        assert res["a"].dtype == "datetime64[s]"
+        assert res["a"].dtype == "datetime64[us]"
         # TODO: this must be int64
         assert res["b"].dtype == "int64"
 
diff --git a/pandas/tests/frame/methods/test_get_numeric_data.py b/pandas/tests/frame/methods/test_get_numeric_data.py
@@ -21,7 +21,7 @@ def test_get_numeric_data_preserve_dtype(self):
         tm.assert_frame_equal(result, expected)
 
     def test_get_numeric_data(self, using_infer_string):
-        datetime64name = np.dtype("M8[s]").name
+        datetime64name = np.dtype("M8[us]").name
         objectname = np.dtype(np.object_).name
 
         df = DataFrame(
diff --git a/pandas/tests/frame/methods/test_reindex.py b/pandas/tests/frame/methods/test_reindex.py
@@ -141,7 +141,7 @@ def test_reindex_tzaware_fill_value(self):
 
         ts = pd.Timestamp("2023-04-10 17:32", tz="US/Pacific")
         res = df.reindex([0, 1], axis=1, fill_value=ts)
-        assert res.dtypes[1] == pd.DatetimeTZDtype(unit="s", tz="US/Pacific")
+        assert res.dtypes[1] == pd.DatetimeTZDtype(unit="us", tz="US/Pacific")
         expected = DataFrame({0: [1], 1: [ts]})
         expected[1] = expected[1].astype(res.dtypes[1])
         tm.assert_frame_equal(res, expected)
@@ -154,7 +154,7 @@ def test_reindex_tzaware_fill_value(self):
 
         interval = pd.Interval(ts, ts + pd.Timedelta(seconds=1))
         res = df.reindex([0, 1], axis=1, fill_value=interval)
-        assert res.dtypes[1] == pd.IntervalDtype("datetime64[s, US/Pacific]", "right")
+        assert res.dtypes[1] == pd.IntervalDtype("datetime64[us, US/Pacific]", "right")
         expected = DataFrame({0: [1], 1: [interval]})
         expected[1] = expected[1].astype(res.dtypes[1])
         tm.assert_frame_equal(res, expected)
diff --git a/pandas/tests/frame/methods/test_to_csv.py b/pandas/tests/frame/methods/test_to_csv.py
@@ -50,7 +50,7 @@ def test_to_csv_from_csv1_datetime(self, temp_file, datetime_frame):
         datetime_frame.to_csv(path)
         recons = self.read_csv(path, parse_dates=True)
         expected = datetime_frame.copy()
-        expected.index = expected.index.as_unit("s")
+        expected.index = expected.index.as_unit("us")
         tm.assert_frame_equal(expected, recons)
 
         datetime_frame.to_csv(path, index_label="index")
@@ -240,8 +240,8 @@ def make_dtnat_arr(n, nnat=None):
         result = self.read_csv(path).apply(to_datetime)
 
         expected = df[:]
-        expected["a"] = expected["a"].astype("M8[s]")
-        expected["b"] = expected["b"].astype("M8[s]")
+        expected["a"] = expected["a"].astype("M8[us]")
+        expected["b"] = expected["b"].astype("M8[us]")
         tm.assert_frame_equal(result, expected, check_names=False)
 
     def _return_result_expected(
@@ -579,7 +579,7 @@ def test_to_csv_multiindex(self, temp_file, float_frame, datetime_frame):
 
         # TODO to_csv drops column name
         expected = tsframe.copy()
-        expected.index = MultiIndex.from_arrays([old_index.as_unit("s"), new_index[1]])
+        expected.index = MultiIndex.from_arrays([old_index.as_unit("us"), new_index[1]])
         tm.assert_frame_equal(recons, expected, check_names=False)
 
         # do not load index
diff --git a/pandas/tests/frame/methods/test_to_numpy.py b/pandas/tests/frame/methods/test_to_numpy.py
@@ -71,8 +71,8 @@ def test_to_numpy_datetime_with_na(self):
         expected = np.array(
             [
                 [0.00e00, 0.00e00, 1.00e00],
-                [8.64e04, np.nan, np.nan],
-                [np.nan, 8.64e04, 2.00e00],
+                [8.64e10, np.nan, np.nan],
+                [np.nan, 8.64e10, 2.00e00],
             ]
         )
         result = df.to_numpy(float, na_value=np.nan)
diff --git a/pandas/tests/frame/test_block_internals.py b/pandas/tests/frame/test_block_internals.py
@@ -192,7 +192,7 @@ def test_construction_with_conversions(self):
 
         expected = DataFrame(
             {
-                "dt1": Timestamp("20130101"),
+                "dt1": Timestamp("20130101").as_unit("s"),
                 "dt2": date_range("20130101", periods=3).astype("M8[s]"),
                 # 'dt3' : date_range('20130101 00:00:01',periods=3,freq='s'),
                 # FIXME: don't leave commented-out
diff --git a/pandas/tests/frame/test_constructors.py b/pandas/tests/frame/test_constructors.py
@@ -879,9 +879,9 @@ def create_data(constructor):
         )
 
         result_datetime64 = DataFrame(data_datetime64)
+        assert result_datetime64.index.unit == "s"
+        result_datetime64.index = result_datetime64.index.as_unit("us")
         result_datetime = DataFrame(data_datetime)
-        assert result_datetime.index.unit == "us"
-        result_datetime.index = result_datetime.index.as_unit("s")
         result_Timestamp = DataFrame(data_Timestamp)
         tm.assert_frame_equal(result_datetime64, expected)
         tm.assert_frame_equal(result_datetime, expected)
@@ -945,7 +945,7 @@ def test_constructor_dict_extension_scalar(self, ea_scalar_and_dtype):
             (Interval(left=0, right=5), IntervalDtype("int64", "right")),
             (
                 Timestamp("2011-01-01", tz="US/Eastern"),
-                DatetimeTZDtype(unit="s", tz="US/Eastern"),
+                DatetimeTZDtype(unit="us", tz="US/Eastern"),
             ),
         ],
     )
@@ -1863,7 +1863,7 @@ def test_constructor_with_datetimes(self, using_infer_string):
                 else pd.StringDtype(na_value=np.nan)
             ]
             * 2
-            + [np.dtype("M8[s]"), np.dtype("M8[us]")],
+            + [np.dtype("M8[us]")] * 2,
             index=list("ABCDE"),
         )
         tm.assert_series_equal(result, expected)
@@ -3076,9 +3076,9 @@ def test_from_tzaware_mixed_object_array(self):
         res = DataFrame(arr, columns=["A", "B", "C"])
 
         expected_dtypes = [
-            "datetime64[s]",
-            "datetime64[s, US/Eastern]",
-            "datetime64[s, CET]",
+            "datetime64[us]",
+            "datetime64[us, US/Eastern]",
+            "datetime64[us, CET]",
         ]
         assert (res.dtypes == expected_dtypes).all()
 
diff --git a/pandas/tests/series/accessors/test_dt_accessor.py b/pandas/tests/series/accessors/test_dt_accessor.py
@@ -369,7 +369,7 @@ def test_dt_round_tz_nonexistent(self, method, ts_str, freq):
         tm.assert_series_equal(result, expected)
 
         result = getattr(ser.dt, method)(freq, nonexistent="NaT")
-        expected = Series([pd.NaT]).dt.tz_localize(result.dt.tz)
+        expected = Series([pd.NaT], dtype="datetime64[us]").dt.tz_localize(result.dt.tz)
         tm.assert_series_equal(result, expected)
 
         with pytest.raises(ValueError, match="2018-03-11 02:00:00"):
diff --git a/pandas/tests/series/indexing/test_setitem.py b/pandas/tests/series/indexing/test_setitem.py
@@ -1483,7 +1483,7 @@ class TestCoercionDatetime64HigherReso(CoercionTest):
     def obj(self, exp_dtype):
         idx = date_range("2011-01-01", freq="D", periods=4, unit="s")
         if exp_dtype == "m8[ms]":
-            idx = idx - Timestamp("1970-01-01")
+            idx = idx - Timestamp("1970-01-01").as_unit("s")
             assert idx.dtype == "m8[s]"
         elif exp_dtype == "M8[ms, UTC]":
             idx = idx.tz_localize("UTC")
@@ -1493,7 +1493,7 @@ def obj(self, exp_dtype):
     def val(self, exp_dtype):
         ts = Timestamp("2011-01-02 03:04:05.678").as_unit("ms")
         if exp_dtype == "m8[ms]":
-            return ts - Timestamp("1970-01-01")
+            return ts - Timestamp("1970-01-01").as_unit("ms")
         elif exp_dtype == "M8[ms, UTC]":
             return ts.tz_localize("UTC")
         return ts
diff --git a/pandas/tests/series/methods/test_combine_first.py b/pandas/tests/series/methods/test_combine_first.py
@@ -79,7 +79,10 @@ def test_combine_first_dt64(self, unit):
         s1 = Series([np.nan, "2011"])
         rs = s0.combine_first(s1)
 
-        xp = Series([datetime(2010, 1, 1), "2011"], dtype=f"datetime64[{unit}]")
+        xp = Series(
+            [datetime(2010, 1, 1), "2011"],
+            dtype=f"datetime64[{'ns' if unit == 'ns' else 'us'}]",
+        )
 
         tm.assert_series_equal(rs, xp)
 
diff --git a/pandas/tests/series/methods/test_fillna.py b/pandas/tests/series/methods/test_fillna.py
@@ -313,7 +313,7 @@ def test_datetime64_fillna(self):
     def test_datetime64_fillna_mismatched_reso_no_rounding(self, tz, scalar):
         # GH#56410
         dti = date_range("2016-01-01", periods=3, unit="s", tz=tz)
-        item = Timestamp("2016-02-03 04:05:06.789", tz=tz)
+        item = Timestamp("2016-02-03 04:05:06.789", tz=tz).as_unit("ms")
         vec = date_range(item, periods=3, unit="ms")
 
         exp_dtype = "M8[ms]" if tz is None else "M8[ms, UTC]"
@@ -340,8 +340,12 @@ def test_datetime64_fillna_mismatched_reso_no_rounding(self, tz, scalar):
     )
     def test_timedelta64_fillna_mismatched_reso_no_rounding(self, scalar):
         # GH#56410
-        tdi = date_range("2016-01-01", periods=3, unit="s") - Timestamp("1970-01-01")
-        item = Timestamp("2016-02-03 04:05:06.789") - Timestamp("1970-01-01")
+        tdi = date_range("2016-01-01", periods=3, unit="s") - Timestamp(
+            "1970-01-01"
+        ).as_unit("s")
+        item = (Timestamp("2016-02-03 04:05:06.789") - Timestamp("1970-01-01")).as_unit(
+            "ms"
+        )
         vec = timedelta_range(item, periods=3, unit="ms")
 
         expected = Series([item, tdi[1], tdi[2]], dtype="m8[ms]")
diff --git a/pandas/tests/series/methods/test_to_csv.py b/pandas/tests/series/methods/test_to_csv.py
@@ -32,7 +32,7 @@ def test_from_csv(self, datetime_series, string_series, temp_file):
         datetime_series.to_csv(path, header=False)
         ts = self.read_csv(path, parse_dates=True)
         expected = datetime_series.copy()
-        expected.index = expected.index.as_unit("s")
+        expected.index = expected.index.as_unit("us")
         tm.assert_series_equal(expected, ts, check_names=False)
 
         assert ts.name is None
@@ -59,7 +59,6 @@ def test_from_csv(self, datetime_series, string_series, temp_file):
 
         series = self.read_csv(path, sep="|", parse_dates=True)
         check_series = Series({datetime(1998, 1, 1): 1.0, datetime(1999, 1, 1): 2.0})
-        check_series.index = check_series.index.as_unit("s")
         tm.assert_series_equal(check_series, series)
 
         series = self.read_csv(path, sep="|", parse_dates=False)
diff --git a/pandas/tests/series/test_constructors.py b/pandas/tests/series/test_constructors.py
@@ -1163,7 +1163,7 @@ def test_constructor_with_datetime_tz4(self):
                 Timestamp("2013-01-02 14:00:00-0800", tz="US/Pacific"),
             ]
         )
-        assert ser.dtype == "datetime64[s, US/Pacific]"
+        assert ser.dtype == "datetime64[us, US/Pacific]"
         assert lib.infer_dtype(ser, skipna=True) == "datetime64"
 
     def test_constructor_with_datetime_tz3(self):
@@ -1408,10 +1408,10 @@ def create_data(constructor):
         result_datetime = Series(data_datetime)
         result_Timestamp = Series(data_Timestamp)
 
-        tm.assert_series_equal(result_datetime64, expected)
         tm.assert_series_equal(
-            result_datetime, expected.set_axis(expected.index.as_unit("us"))
+            result_datetime64, expected.set_axis(expected.index.as_unit("s"))
         )
+        tm.assert_series_equal(result_datetime, expected)
         tm.assert_series_equal(result_Timestamp, expected)
 
     def test_constructor_dict_tuple_indexer(self):

Original file line number	Diff line number	Diff line change
`@@ -935,7 +935,7 @@ def rand_series_with_duplicate_datetimeindex() -> Series:`
`935`	`935`	`(Period("2012-02-01", freq="D"), "period[D]"),`
`936`	`936`	`(`
`937`	`937`	`Timestamp("2011-01-01", tz="US/Eastern"),`
`938`		`- DatetimeTZDtype(unit="s", tz="US/Eastern"),`
	`938`	`+ DatetimeTZDtype(unit="us", tz="US/Eastern"),`
`939`	`939`	`),`
`940`	`940`	`(Timedelta(seconds=500), "timedelta64[ns]"),`
`941`	`941`	`]`
Original file line number	Diff line number	Diff line change
`@@ -825,7 +825,7 @@ def test_setitem_single_column_mixed_datetime(self):`
`825`	`825`	`# check our dtypes`
`826`	`826`	`result = df.dtypes`
`827`	`827`	`expected = Series(`
`828`		`- [np.dtype("float64")] * 3 + [np.dtype("datetime64[s]")],`
	`828`	`+ [np.dtype("float64")] * 3 + [np.dtype("datetime64[us]")],`
`829`	`829`	`index=["foo", "bar", "baz", "timestamp"],`
`830`	`830`	`)`
`831`	`831`	`tm.assert_series_equal(result, expected)`
Original file line number	Diff line number	Diff line change
`@@ -260,7 +260,7 @@ def test_setitem_dict_preserves_dtypes(self):`
`260`	`260`	`(Interval(left=0, right=5), IntervalDtype("int64", "right")),`
`261`	`261`	`(`
`262`	`262`	`Timestamp("2011-01-01", tz="US/Eastern"),`
`263`		`- DatetimeTZDtype(unit="s", tz="US/Eastern"),`
	`263`	`+ DatetimeTZDtype(unit="us", tz="US/Eastern"),`
`264`	`264`	`),`
`265`	`265`	`],`
`266`	`266`	`)`
Original file line number	Diff line number	Diff line change
`@@ -71,8 +71,8 @@ def test_to_numpy_datetime_with_na(self):`
`71`	`71`	`expected = np.array(`
`72`	`72`	`[`
`73`	`73`	`[0.00e00, 0.00e00, 1.00e00],`
`74`		`- [8.64e04, np.nan, np.nan],`
`75`		`- [np.nan, 8.64e04, 2.00e00],`
	`74`	`+ [8.64e10, np.nan, np.nan],`
	`75`	`+ [np.nan, 8.64e10, 2.00e00],`
`76`	`76`	`]`
`77`	`77`	`)`
`78`	`78`	`result = df.to_numpy(float, na_value=np.nan)`
Original file line number	Diff line number	Diff line change
`@@ -192,7 +192,7 @@ def test_construction_with_conversions(self):`
`192`	`192`
`193`	`193`	`expected = DataFrame(`
`194`	`194`	`{`
`195`		`- "dt1": Timestamp("20130101"),`
	`195`	`+ "dt1": Timestamp("20130101").as_unit("s"),`
`196`	`196`	`"dt2": date_range("20130101", periods=3).astype("M8[s]"),`
`197`	`197`	`# 'dt3' : date_range('20130101 00:00:01',periods=3,freq='s'),`
`198`	`198`	`# FIXME: don't leave commented-out`