TST: Fix test warnings (#47257)

Author: mroeschke

pandas/tests/frame/test_arithmetic.py

@@ -2007,11 +2007,11 @@ def test_bool_frame_mult_float():
     tm.assert_frame_equal(result, expected)
 
 
-def test_frame_sub_nullable_int(any_int_dtype):
+def test_frame_sub_nullable_int(any_int_ea_dtype):
     # GH 32822
-    series1 = Series([1, 2, np.nan], dtype=any_int_dtype)
-    series2 = Series([1, 2, 3], dtype=any_int_dtype)
-    expected = DataFrame([0, 0, np.nan], dtype=any_int_dtype)
+    series1 = Series([1, 2, None], dtype=any_int_ea_dtype)
+    series2 = Series([1, 2, 3], dtype=any_int_ea_dtype)
+    expected = DataFrame([0, 0, None], dtype=any_int_ea_dtype)
     result = series1.to_frame() - series2.to_frame()
     tm.assert_frame_equal(result, expected)
 

pandas/tests/frame/test_constructors.py

@@ -2574,7 +2574,8 @@ def check_views(c_only: bool = False):
 
         # FIXME(GH#35417): until GH#35417, iloc.setitem into EA values does not preserve
         #  view, so we have to check in the other direction
-        df.iloc[:, 2] = pd.array([45, 46], dtype=c.dtype)
+        with tm.assert_produces_warning(FutureWarning, match="will attempt to set"):
+            df.iloc[:, 2] = pd.array([45, 46], dtype=c.dtype)
         assert df.dtypes.iloc[2] == c.dtype
         if not copy:
             check_views(True)

pandas/tests/groupby/aggregate/test_numba.py

@@ -47,7 +47,7 @@ def incorrect_function(values, index):
 
 
 @td.skip_if_no("numba")
-@pytest.mark.filterwarnings("ignore:\n")
+@pytest.mark.filterwarnings("ignore")
 # Filter warnings when parallel=True and the function can't be parallelized by Numba
 @pytest.mark.parametrize("jit", [True, False])
 @pytest.mark.parametrize("pandas_obj", ["Series", "DataFrame"])
@@ -76,7 +76,7 @@ def func_numba(values, index):
 
 
 @td.skip_if_no("numba")
-@pytest.mark.filterwarnings("ignore:\n")
+@pytest.mark.filterwarnings("ignore")
 # Filter warnings when parallel=True and the function can't be parallelized by Numba
 @pytest.mark.parametrize("jit", [True, False])
 @pytest.mark.parametrize("pandas_obj", ["Series", "DataFrame"])

pandas/tests/groupby/test_numba.py

@@ -10,7 +10,7 @@
 
 
 @td.skip_if_no("numba")
-@pytest.mark.filterwarnings("ignore:\n")
+@pytest.mark.filterwarnings("ignore")
 # Filter warnings when parallel=True and the function can't be parallelized by Numba
 class TestEngine:
     def test_cython_vs_numba_frame(

pandas/tests/groupby/transform/test_numba.py

@@ -44,7 +44,7 @@ def incorrect_function(values, index):
 
 
 @td.skip_if_no("numba")
-@pytest.mark.filterwarnings("ignore:\n")
+@pytest.mark.filterwarnings("ignore")
 # Filter warnings when parallel=True and the function can't be parallelized by Numba
 @pytest.mark.parametrize("jit", [True, False])
 @pytest.mark.parametrize("pandas_obj", ["Series", "DataFrame"])
@@ -73,7 +73,7 @@ def func(values, index):
 
 
 @td.skip_if_no("numba")
-@pytest.mark.filterwarnings("ignore:\n")
+@pytest.mark.filterwarnings("ignore")
 # Filter warnings when parallel=True and the function can't be parallelized by Numba
 @pytest.mark.parametrize("jit", [True, False])
 @pytest.mark.parametrize("pandas_obj", ["Series", "DataFrame"])

pandas/tests/io/parser/test_parse_dates.py

@@ -1538,7 +1538,12 @@ def test_date_parser_resolution_if_not_ns(all_parsers):
 """
 
     def date_parser(dt, time):
-        return np.array(dt + "T" + time, dtype="datetime64[s]")
+        try:
+            arr = dt + "T" + time
+        except TypeError:
+            # dt & time are date/time objects
+            arr = [datetime.combine(d, t) for d, t in zip(dt, time)]
+        return np.array(arr, dtype="datetime64[s]")
 
     result = parser.read_csv(
         StringIO(data),

pandas/tests/strings/test_find_replace.py

@@ -25,7 +25,11 @@ def test_contains(any_string_dtype):
     values = Series(values, dtype=any_string_dtype)
     pat = "mmm[_]+"
 
-    result = values.str.contains(pat)
+    with tm.maybe_produces_warning(
+        PerformanceWarning,
+        any_string_dtype == "string[pyarrow]" and pa_version_under4p0,
+    ):
+        result = values.str.contains(pat)
     expected_dtype = "object" if any_string_dtype == "object" else "boolean"
     expected = Series(
         np.array([False, np.nan, True, True, False], dtype=np.object_),
@@ -88,7 +92,11 @@ def test_contains(any_string_dtype):
     )
     tm.assert_series_equal(result, expected)
 
-    result = values.str.contains(pat, na=False)
+    with tm.maybe_produces_warning(
+        PerformanceWarning,
+        any_string_dtype == "string[pyarrow]" and pa_version_under4p0,
+    ):
+        result = values.str.contains(pat, na=False)
     expected_dtype = np.bool_ if any_string_dtype == "object" else "boolean"
     expected = Series(np.array([False, False, True, True]), dtype=expected_dtype)
     tm.assert_series_equal(result, expected)
@@ -181,7 +189,11 @@ def test_contains_moar(any_string_dtype):
         dtype=any_string_dtype,
     )
 
-    result = s.str.contains("a")
+    with tm.maybe_produces_warning(
+        PerformanceWarning,
+        any_string_dtype == "string[pyarrow]" and pa_version_under4p0,
+    ):
+        result = s.str.contains("a")
     expected_dtype = "object" if any_string_dtype == "object" else "boolean"
     expected = Series(
         [False, False, False, True, True, False, np.nan, False, False, True],
@@ -619,7 +631,11 @@ def test_replace_moar(any_string_dtype):
         dtype=any_string_dtype,
     )
 
-    result = ser.str.replace("A", "YYY")
+    with tm.maybe_produces_warning(
+        PerformanceWarning,
+        any_string_dtype == "string[pyarrow]" and pa_version_under4p0,
+    ):
+        result = ser.str.replace("A", "YYY")
     expected = Series(
         ["YYY", "B", "C", "YYYaba", "Baca", "", np.nan, "CYYYBYYY", "dog", "cat"],
         dtype=any_string_dtype,
@@ -727,7 +743,11 @@ def test_replace_regex_single_character(regex, any_string_dtype):
         ):
             result = s.str.replace(".", "a", regex=regex)
     else:
-        result = s.str.replace(".", "a", regex=regex)
+        with tm.maybe_produces_warning(
+            PerformanceWarning,
+            any_string_dtype == "string[pyarrow]" and pa_version_under4p0,
+        ):
+            result = s.str.replace(".", "a", regex=regex)
 
     expected = Series(["aab", "a", "b", np.nan, ""], dtype=any_string_dtype)
     tm.assert_series_equal(result, expected)

pandas/tests/strings/test_strings.py

@@ -562,7 +562,11 @@ def test_slice_replace(start, stop, repl, expected, any_string_dtype):
 def test_strip_lstrip_rstrip(any_string_dtype, method, exp):
     ser = Series(["  aa   ", " bb \n", np.nan, "cc  "], dtype=any_string_dtype)
 
-    result = getattr(ser.str, method)()
+    with tm.maybe_produces_warning(
+        PerformanceWarning,
+        any_string_dtype == "string[pyarrow]" and pa_version_under4p0,
+    ):
+        result = getattr(ser.str, method)()
     expected = Series(exp, dtype=any_string_dtype)
     tm.assert_series_equal(result, expected)
 

pandas/tests/window/test_numba.py

@@ -50,7 +50,7 @@ def arithmetic_numba_supported_operators(request):
 
 
 @td.skip_if_no("numba")
-@pytest.mark.filterwarnings("ignore:\n")
+@pytest.mark.filterwarnings("ignore")
 # Filter warnings when parallel=True and the function can't be parallelized by Numba
 class TestEngine:
     @pytest.mark.parametrize("jit", [True, False])
@@ -331,7 +331,7 @@ def test_invalid_kwargs_nopython():
 
 @td.skip_if_no("numba")
 @pytest.mark.slow
-@pytest.mark.filterwarnings("ignore:\n")
+@pytest.mark.filterwarnings("ignore")
 # Filter warnings when parallel=True and the function can't be parallelized by Numba
 class TestTableMethod:
     def test_table_series_valueerror(self):

pandas/tests/window/test_online.py

@@ -24,7 +24,8 @@
 
 
 @td.skip_if_no("numba")
-@pytest.mark.filterwarnings("ignore:\n")
+@pytest.mark.filterwarnings("ignore")
+# Filter warnings when parallel=True and the function can't be parallelized by Numba
 class TestEWM:
     def test_invalid_update(self):
         df = DataFrame({"a": range(5), "b": range(5)})