update tests

Author: jorisvandenbossche

pandas/tests/io/parser/common/test_common_basic.py

@@ -65,7 +65,7 @@ def test_read_csv_local(all_parsers, csv1):
                 datetime(2000, 1, 10),
                 datetime(2000, 1, 11),
             ],
-            dtype="M8[s]",
+            dtype="M8[us]" if all_parsers.engine != "pyarrow" else "M8[s]",
             name="index",
         ),
     )
@@ -167,7 +167,7 @@ def test_read_csv_dataframe(all_parsers, csv1):
                 datetime(2000, 1, 10),
                 datetime(2000, 1, 11),
             ],
-            dtype="M8[s]",
+            dtype="M8[us]" if all_parsers.engine != "pyarrow" else "M8[s]",
             name="index",
         ),
     )

pandas/tests/io/parser/common/test_index.py

@@ -263,7 +263,7 @@ def test_read_csv_no_index_name(all_parsers, csv_dir_path):
                 datetime(2000, 1, 6),
                 datetime(2000, 1, 7),
             ],
-            dtype="M8[s]",
+            dtype="M8[us]",
         ),
     )
     tm.assert_frame_equal(result, expected)

pandas/tests/io/parser/test_multi_thread.py

@@ -155,5 +155,5 @@ def test_multi_thread_path_multipart_read_csv(all_parsers):
         result = _generate_multi_thread_dataframe(parser, path, num_rows, num_tasks)
 
     expected = df[:]
-    expected["date"] = expected["date"].astype("M8[s]")
+    expected["date"] = expected["date"].astype("M8[us]")
     tm.assert_frame_equal(result, expected)

pandas/tests/io/parser/test_parse_dates.py

@@ -61,7 +61,6 @@ def test_date_col_as_index_col(all_parsers):
             datetime(1999, 1, 27, 21, 0),
             datetime(1999, 1, 27, 22, 0),
         ],
-        dtype="M8[s]",
         name="X1",
     )
     expected = DataFrame(
@@ -126,7 +125,7 @@ def test_parse_dates_string(all_parsers):
     parser = all_parsers
     result = parser.read_csv(StringIO(data), index_col="date", parse_dates=["date"])
     # freq doesn't round-trip
-    index = date_range("1/1/2009", periods=3, name="date", unit="s")._with_freq(None)
+    index = date_range("1/1/2009", periods=3, name="date", unit="us")._with_freq(None)
 
     expected = DataFrame(
         {"A": ["a", "b", "c"], "B": [1, 3, 4], "C": [2, 4, 5]}, index=index
@@ -143,8 +142,6 @@ def test_parse_dates_column_list(all_parsers, parse_dates):
     expected = DataFrame(
         {"a": [datetime(2010, 1, 1)], "b": [1], "c": [datetime(2010, 2, 15)]}
     )
-    expected["a"] = expected["a"].astype("M8[s]")
-    expected["c"] = expected["c"].astype("M8[s]")
     expected = expected.set_index(["a", "b"])
 
     result = parser.read_csv(
@@ -168,7 +165,7 @@ def test_multi_index_parse_dates(all_parsers, index_col):
 20090103,three,c,4,5
 """
     parser = all_parsers
-    dti = date_range("2009-01-01", periods=3, freq="D", unit="s")
+    dti = date_range("2009-01-01", periods=3, freq="D", unit="us")
     index = MultiIndex.from_product(
         [
             dti,
@@ -218,6 +215,7 @@ def test_parse_tz_aware(all_parsers):
     if parser.engine == "pyarrow":
         pytz = pytest.importorskip("pytz")
         expected_tz = pytz.utc
+        expected.index = expected.index.as_unit("s")
     else:
         expected_tz = timezone.utc
     tm.assert_frame_equal(result, expected)
@@ -303,7 +301,7 @@ def test_parse_dates_empty_string(all_parsers):
     expected = DataFrame(
         [[datetime(2012, 1, 1), 1], [pd.NaT, 2]], columns=["Date", "test"]
     )
-    expected["Date"] = expected["Date"].astype("M8[s]")
+    expected["Date"] = expected["Date"].astype("M8[us]")
     tm.assert_frame_equal(result, expected)
 
 
@@ -314,22 +312,18 @@ def test_parse_dates_empty_string(all_parsers):
         (
             "a\n04.15.2016",
             {"parse_dates": ["a"]},
-            DataFrame([datetime(2016, 4, 15)], columns=["a"], dtype="M8[s]"),
+            DataFrame([datetime(2016, 4, 15)], columns=["a"]),
         ),
         (
             "a\n04.15.2016",
             {"parse_dates": True, "index_col": 0},
-            DataFrame(
-                index=DatetimeIndex(["2016-04-15"], dtype="M8[s]", name="a"), columns=[]
-            ),
+            DataFrame(index=DatetimeIndex(["2016-04-15"], name="a"), columns=[]),
         ),
         (
             "a,b\n04.15.2016,09.16.2013",
             {"parse_dates": ["a", "b"]},
             DataFrame(
-                [[datetime(2016, 4, 15), datetime(2013, 9, 16)]],
-                dtype="M8[s]",
-                columns=["a", "b"],
+                [[datetime(2016, 4, 15), datetime(2013, 9, 16)]], columns=["a", "b"]
             ),
         ),
         (
@@ -339,8 +333,8 @@ def test_parse_dates_empty_string(all_parsers):
                 index=MultiIndex.from_tuples(
                     [
                         (
-                            Timestamp(2016, 4, 15).as_unit("s"),
-                            Timestamp(2013, 9, 16).as_unit("s"),
+                            Timestamp(2016, 4, 15),
+                            Timestamp(2013, 9, 16),
                         )
                     ],
                     names=["a", "b"],
@@ -411,7 +405,7 @@ def test_parse_timezone(all_parsers):
         end="2018-01-04 09:05:00",
         freq="1min",
         tz=timezone(timedelta(minutes=540)),
-        unit="s",
+        unit="us",
     )._with_freq(None)
     expected_data = {"dt": dti, "val": [23350, 23400, 23400, 23400, 23400]}
 
@@ -450,7 +444,7 @@ def test_parse_delimited_date_swap_no_warning(
     all_parsers, date_string, dayfirst, expected, request
 ):
     parser = all_parsers
-    expected = DataFrame({0: [expected]}, dtype="datetime64[s]")
+    expected = DataFrame({0: [expected]}, dtype="datetime64[us]")
     if parser.engine == "pyarrow":
         if not dayfirst:
             # "CSV parse error: Empty CSV file or block"
@@ -483,7 +477,7 @@ def test_parse_delimited_date_swap_with_warning(
     all_parsers, date_string, dayfirst, expected
 ):
     parser = all_parsers
-    expected = DataFrame({0: [expected]}, dtype="datetime64[s]")
+    expected = DataFrame({0: [expected]}, dtype="datetime64[us]")
     warning_msg = (
         "Parsing dates in .* format when dayfirst=.* was specified. "
         "Pass `dayfirst=.*` or specify a format to silence this warning."
@@ -602,7 +596,6 @@ def test_date_parser_usecols_thousands(all_parsers):
         thousands="-",
     )
     expected = DataFrame({"B": [3, 4], "C": [Timestamp("20-09-2001 01:00:00")] * 2})
-    expected["C"] = expected["C"].astype("M8[s]")
     tm.assert_frame_equal(result, expected)
 
 
@@ -611,9 +604,7 @@ def test_dayfirst_warnings():
 
     # CASE 1: valid input
     input = "date\n31/12/2014\n10/03/2011"
-    expected = DatetimeIndex(
-        ["2014-12-31", "2011-03-10"], dtype="datetime64[s]", freq=None, name="date"
-    )
+    expected = DatetimeIndex(["2014-12-31", "2011-03-10"], freq=None, name="date")
     warning_msg = (
         "Parsing dates in .* format when dayfirst=.* was specified. "
         "Pass `dayfirst=.*` or specify a format to silence this warning."
@@ -672,9 +663,7 @@ def test_dayfirst_warnings():
 def test_dayfirst_warnings_no_leading_zero(date_string, dayfirst):
     # GH47880
     initial_value = f"date\n{date_string}"
-    expected = DatetimeIndex(
-        ["2014-01-31"], dtype="datetime64[s]", freq=None, name="date"
-    )
+    expected = DatetimeIndex(["2014-01-31"], freq=None, name="date")
     warning_msg = (
         "Parsing dates in .* format when dayfirst=.* was specified. "
         "Pass `dayfirst=.*` or specify a format to silence this warning."
@@ -729,7 +718,7 @@ def test_replace_nans_before_parsing_dates(all_parsers):
                 Timestamp("2017-09-09"),
             ]
         },
-        dtype="M8[s]",
+        dtype="M8[us]",
     )
     tm.assert_frame_equal(result, expected)
 
@@ -744,7 +733,6 @@ def test_parse_dates_and_string_dtype(all_parsers):
     result = parser.read_csv(StringIO(data), dtype="string", parse_dates=["b"])
     expected = DataFrame({"a": ["1"], "b": [Timestamp("2019-12-31")]})
     expected["a"] = expected["a"].astype("string")
-    expected["b"] = expected["b"].astype("M8[s]")
     tm.assert_frame_equal(result, expected)
 
 
@@ -763,9 +751,7 @@ def test_parse_dot_separated_dates(all_parsers):
         warn = None
     else:
         expected_index = DatetimeIndex(
-            ["2003-03-27 14:55:00", "2003-08-03 15:20:00"],
-            dtype="datetime64[ms]",
-            name="a",
+            ["2003-03-27 14:55:00", "2003-08-03 15:20:00"], name="a"
         )
         warn = UserWarning
     msg = r"when dayfirst=False \(the default\) was specified"
@@ -798,7 +784,7 @@ def test_parse_dates_dict_format(all_parsers):
             "a": [Timestamp("2019-12-31"), Timestamp("2020-12-31")],
             "b": [Timestamp("2019-12-31"), Timestamp("2020-12-31")],
         },
-        dtype="M8[s]",
+        dtype="M8[us]",
     )
     tm.assert_frame_equal(result, expected)
 
@@ -840,6 +826,8 @@ def test_parse_dates_arrow_engine(all_parsers):
             "b": 1,
         }
     )
+    if parser.engine == "pyarrow":
+        expected["a"] = expected["a"].dt.as_unit("s")
     tm.assert_frame_equal(result, expected)
 
 

pandas/tests/io/parser/test_read_fwf.py

@@ -308,7 +308,6 @@ def test_fwf_regression():
         parse_dates=True,
         date_format="%Y%j%H%M%S",
     )
-    expected.index = expected.index.astype("M8[s]")
     tm.assert_frame_equal(result, expected)
 
 

pandas/tests/io/parser/test_skiprows.py

@@ -43,7 +43,7 @@ def test_skip_rows_bug(all_parsers, skiprows):
     )
     index = Index(
         [datetime(2000, 1, 1), datetime(2000, 1, 2), datetime(2000, 1, 3)],
-        dtype="M8[s]",
+        dtype="M8[us]",
         name=0,
     )
 
@@ -88,7 +88,7 @@ def test_skip_rows_blank(all_parsers):
     )
     index = Index(
         [datetime(2000, 1, 1), datetime(2000, 1, 2), datetime(2000, 1, 3)],
-        dtype="M8[s]",
+        dtype="M8[us]",
         name=0,
     )
 

pandas/tests/io/pytables/test_append.py

@@ -826,7 +826,7 @@ def test_append_raise(setup_path, using_infer_string):
         msg = re.escape(
             "Cannot serialize the column [foo] "
             "because its data contents are not [string] "
-            "but [datetime64[s]] object dtype"
+            "but [datetime64[us]] object dtype"
         )
         with pytest.raises(ValueError, match=msg):
             store.append("df", df)

pandas/tests/io/pytables/test_round_trip.py

@@ -252,8 +252,7 @@ def test_table_values_dtypes_roundtrip(setup_path, using_infer_string):
                 "int8": 1,
                 "int64": 1,
                 str_dtype: 1,
-                "datetime64[s]": 2,
-                "datetime64[ms]": 1,
+                "datetime64[us]": 3,
                 "datetime64[ns]": 1,
             },
             name="count",

pandas/tests/io/test_gcs.py

@@ -117,8 +117,6 @@ def from_uri(path):
         df2 = df1
 
     expected = df1[:]
-    if format in ["csv", "excel"]:
-        expected["dt"] = expected["dt"].dt.as_unit("s")
 
     tm.assert_frame_equal(df2, expected)
 

pandas/tests/io/test_sql.py

@@ -1858,10 +1858,8 @@ def test_api_custom_dateparsing_error(
 
     if conn_name == "postgresql_adbc_types" and pa_version_under14p1:
         expected["DateCol"] = expected["DateCol"].astype("datetime64[ns]")
-    elif "postgres" in conn_name or "mysql" in conn_name:
-        expected["DateCol"] = expected["DateCol"].astype("datetime64[us]")
     else:
-        expected["DateCol"] = expected["DateCol"].astype("datetime64[s]")
+        expected["DateCol"] = expected["DateCol"].astype("datetime64[us]")
     tm.assert_frame_equal(result, expected)