GH1246 Upgrade to pandas 2.3.0

Author: loicdiridollou

pyproject.toml

@@ -37,7 +37,7 @@ ty = "^0.0.1a8"
 
 [tool.poetry.group.dev.dependencies]
 mypy = "1.16.0"
-pandas = "2.2.3"
+pandas = "2.3.0"
 pyarrow = ">=10.0.1"
 pytest = ">=7.1.2"
 pyright = ">=1.1.400"

tests/__init__.py

@@ -50,6 +50,7 @@
 
 TYPE_CHECKING_INVALID_USAGE: Final = TYPE_CHECKING
 WINDOWS = os.name == "nt" or "cygwin" in platform.system().lower()
+PD_LTE_23 = Version(pd.__version__) < Version("2.3.999")
 PD_LTE_22 = Version(pd.__version__) < Version("2.2.999")
 NUMPY20 = np.lib.NumpyVersion(np.__version__) >= "2.0.0"
 

tests/test_errors.py

@@ -4,7 +4,7 @@
 import pytest
 
 from tests import (
-    PD_LTE_22,
+    PD_LTE_23,
     WINDOWS,
 )
 
@@ -108,13 +108,13 @@ def test_specification_error() -> None:
 
 
 def test_setting_with_copy_error() -> None:
-    if PD_LTE_22:
+    if PD_LTE_23:
         with pytest.raises(errors.SettingWithCopyError):
             raise errors.SettingWithCopyError()
 
 
 def test_setting_with_copy_warning() -> None:
-    if PD_LTE_22:
+    if PD_LTE_23:
         with pytest.warns(errors.SettingWithCopyWarning):
             warnings.warn("", errors.SettingWithCopyWarning)
 

tests/test_frame.py

@@ -49,6 +49,7 @@
 
 from tests import (
     PD_LTE_22,
+    PD_LTE_23,
     TYPE_CHECKING_INVALID_USAGE,
     check,
     ensure_clean,
@@ -450,7 +451,7 @@ def test_types_drop_duplicates() -> None:
         pd.DataFrame,
     )
 
-    if not PD_LTE_22:
+    if not PD_LTE_23:
         check(assert_type(df.drop_duplicates({"AAA"}), pd.DataFrame), pd.DataFrame)
         check(
             assert_type(df.drop_duplicates({"AAA": None}), pd.DataFrame), pd.DataFrame
@@ -1602,7 +1603,7 @@ def wrapped_min(x: Any) -> Any:
     with pytest_warns_bounded(
         FutureWarning,
         r"The provided callable <built-in function (min|max)> is currently using",
-        upper="2.2.99",
+        upper="2.3.99",
     ):
         check(assert_type(df.groupby("col1")["col3"].agg(min), pd.Series), pd.Series)
         check(
@@ -1751,7 +1752,7 @@ def test_types_window() -> None:
     with pytest_warns_bounded(
         FutureWarning,
         r"The provided callable <built-in function (min|max)> is currently using",
-        upper="2.2.99",
+        upper="2.3.99",
     ):
         check(
             assert_type(df.rolling(2).agg(max), pd.DataFrame),
@@ -1860,7 +1861,7 @@ def test_types_agg() -> None:
     with pytest_warns_bounded(
         FutureWarning,
         r"The provided callable <(built-in function (min|max|mean)|function mean at 0x\w+)> is currently using",
-        upper="2.2.99",
+        upper="2.3.99",
     ):
         check(assert_type(df.agg(min), pd.Series), pd.Series)
         check(assert_type(df.agg([min, max]), pd.DataFrame), pd.DataFrame)
@@ -1887,7 +1888,7 @@ def test_types_aggregate() -> None:
     with pytest_warns_bounded(
         FutureWarning,
         r"The provided callable <built-in function (min|max)> is currently using",
-        upper="2.2.99",
+        upper="2.3.99",
     ):
         check(assert_type(df.aggregate(min), pd.Series), pd.Series)
         check(assert_type(df.aggregate([min, max]), pd.DataFrame), pd.DataFrame)
@@ -2020,7 +2021,7 @@ def test_types_resample() -> None:
         FutureWarning,
         "'M' is deprecated",
         lower="2.1.99",
-        upper="2.2.99",
+        upper="2.3.99",
         upper_exception=ValueError,
     ):
         df.resample("M", on="date")
@@ -2163,7 +2164,7 @@ def resampler_foo(resampler: Resampler[pd.DataFrame]) -> pd.DataFrame:
         FutureWarning,
         "'M' is deprecated",
         lower="2.1.99",
-        upper="2.2.99",
+        upper="2.3.99",
         upper_exception=ValueError,
     ):
         (
@@ -2504,7 +2505,7 @@ def test_types_regressions() -> None:
     tslist = list(pd.to_datetime(["2022-01-01", "2022-01-02"]))
     check(assert_type(tslist, list[pd.Timestamp]), list, pd.Timestamp)
     sseries = pd.Series(tslist)
-    with pytest_warns_bounded(FutureWarning, "'d' is deprecated", lower="2.2.99"):
+    with pytest_warns_bounded(FutureWarning, "'d' is deprecated", lower="2.3.99"):
         sseries + pd.Timedelta(1, "d")
 
     check(
@@ -2518,7 +2519,7 @@ def test_types_regressions() -> None:
         FutureWarning,
         "'H' is deprecated",
         lower="2.1.99",
-        upper="2.2.99",
+        upper="2.3.99",
         upper_exception=ValueError,
     ):
         pd.date_range(start="2021-12-01", periods=24, freq="H")
@@ -2644,7 +2645,7 @@ class ReadCsvKwargs(TypedDict):
             ),
             pd.DataFrame,
         )
-        if PD_LTE_22:
+        if PD_LTE_23:
             parse_dates_2 = {"combined_date": ["Year", "Month", "Day"]}
             with pytest_warns_bounded(
                 FutureWarning,
@@ -3098,7 +3099,7 @@ def test_frame_stack() -> None:
     with pytest_warns_bounded(
         FutureWarning,
         "The previous implementation of stack is deprecated",
-        upper="2.2.99",
+        upper="2.3.99",
     ):
         check(
             assert_type(
@@ -3113,7 +3114,7 @@ def test_frame_stack() -> None:
             ),
             pd.Series,
         )
-        if PD_LTE_22:
+        if PD_LTE_23:
             check(
                 assert_type(
                     df_multi_level_cols2.stack(0, future_stack=False),
@@ -3145,7 +3146,7 @@ def test_frame_reindex_like() -> None:
         FutureWarning,
         "the 'method' keyword is deprecated and will be removed in a future version. Please take steps to stop the use of 'method'",
         lower="2.2.99",
-        upper="3.0.99",
+        upper="2.2.99",
     ):
         check(
             assert_type(
@@ -3277,7 +3278,7 @@ def test_groupby_result() -> None:
     index, value = next(iterator)
     assert_type((index, value), tuple[tuple, pd.DataFrame])
 
-    if PD_LTE_22:
+    if PD_LTE_23:
         check(assert_type(index, tuple), tuple, np.integer)
     else:
         check(assert_type(index, tuple), tuple, int)
@@ -3389,7 +3390,7 @@ def test_groupby_result_for_ambiguous_indexes() -> None:
     with pytest_warns_bounded(
         FutureWarning,
         "The default of observed=False is deprecated",
-        upper="2.2.99",
+        upper="2.3.99",
     ):
         categorical_index = pd.CategoricalIndex(df.a)
         iterator2 = df.groupby(categorical_index).__iter__()
@@ -3448,41 +3449,72 @@ def test_groupby_apply() -> None:
     def sum_mean(x: pd.DataFrame) -> float:
         return x.sum().mean()
 
-    with pytest_warns_bounded(
-        DeprecationWarning,
-        "DataFrameGroupBy.apply operated on the grouping columns.",
-        upper="2.99",
+    with (
+        pytest_warns_bounded(
+            DeprecationWarning,
+            "DataFrameGroupBy.apply operated on the grouping columns.",
+            upper="2.2.99",
+        ),
+        pytest_warns_bounded(
+            FutureWarning,
+            "DataFrameGroupBy.apply operated on the grouping columns.",
+            lower="2.2.99",
+            upper="2.3.99",
+        ),
     ):
         check(
             assert_type(df.groupby("col1").apply(sum_mean), pd.Series),
             pd.Series,
         )
 
     lfunc: Callable[[pd.DataFrame], float] = lambda x: x.sum().mean()
-    with pytest_warns_bounded(
-        DeprecationWarning,
-        "DataFrameGroupBy.apply operated on the grouping columns.",
-        upper="2.99",
+    with (
+        pytest_warns_bounded(
+            DeprecationWarning,
+            "DataFrameGroupBy.apply operated on the grouping columns.",
+            upper="2.2.99",
+        ),
+        pytest_warns_bounded(
+            FutureWarning,
+            "DataFrameGroupBy.apply operated on the grouping columns.",
+            lower="2.2.99",
+        ),
     ):
         check(assert_type(df.groupby("col1").apply(lfunc), pd.Series), pd.Series)
 
     def sum_to_list(x: pd.DataFrame) -> list:
         return x.sum().tolist()
 
-    with pytest_warns_bounded(
-        DeprecationWarning,
-        "DataFrameGroupBy.apply operated on the grouping columns.",
-        upper="2.99",
+    with (
+        pytest_warns_bounded(
+            DeprecationWarning,
+            "DataFrameGroupBy.apply operated on the grouping columns.",
+            upper="2.2.99",
+        ),
+        pytest_warns_bounded(
+            FutureWarning,
+            "DataFrameGroupBy.apply operated on the grouping columns.",
+            lower="2.2.99",
+            upper="2.3.99",
+        ),
     ):
         check(assert_type(df.groupby("col1").apply(sum_to_list), pd.Series), pd.Series)
 
     def sum_to_series(x: pd.DataFrame) -> pd.Series:
         return x.sum()
 
-    with pytest_warns_bounded(
-        DeprecationWarning,
-        "DataFrameGroupBy.apply operated on the grouping columns.",
-        upper="2.99",
+    with (
+        pytest_warns_bounded(
+            DeprecationWarning,
+            "DataFrameGroupBy.apply operated on the grouping columns.",
+            upper="2.2.99",
+        ),
+        pytest_warns_bounded(
+            FutureWarning,
+            "DataFrameGroupBy.apply operated on the grouping columns.",
+            lower="2.2.99",
+            upper="2.3.99",
+        ),
     ):
         check(
             assert_type(df.groupby("col1").apply(sum_to_series), pd.DataFrame),
@@ -3492,10 +3524,18 @@ def sum_to_series(x: pd.DataFrame) -> pd.Series:
     def sample_to_df(x: pd.DataFrame) -> pd.DataFrame:
         return x.sample()
 
-    with pytest_warns_bounded(
-        DeprecationWarning,
-        "DataFrameGroupBy.apply operated on the grouping columns.",
-        upper="2.99",
+    with (
+        pytest_warns_bounded(
+            DeprecationWarning,
+            "DataFrameGroupBy.apply operated on the grouping columns.",
+            upper="2.2.99",
+        ),
+        pytest_warns_bounded(
+            FutureWarning,
+            "DataFrameGroupBy.apply operated on the grouping columns.",
+            lower="2.2.99",
+            upper="2.3.99",
+        ),
     ):
         check(
             assert_type(
@@ -3769,7 +3809,7 @@ def test_resample_150_changes() -> None:
         FutureWarning,
         "'M' is deprecated",
         lower="2.1.99",
-        upper="2.2.99",
+        upper="2.3.99",
         upper_exception=ValueError,
     ):
         frame.resample("M", group_keys=True)
@@ -3872,7 +3912,7 @@ def test_getattr_and_dataframe_groupby() -> None:
     with pytest_warns_bounded(
         FutureWarning,
         r"The provided callable <built-in function (min|max)> is currently using",
-        upper="2.2.99",
+        upper="2.3.99",
     ):
         check(assert_type(df.groupby("col1").col3.agg(min), pd.Series), pd.Series)
         check(
@@ -4390,6 +4430,7 @@ def test_transpose() -> None:
         DeprecationWarning,
         msg,
         lower="2.2.99",
+        upper="2.2.99",
     ):
         check(assert_type(df.transpose(copy=True), pd.DataFrame), pd.DataFrame)