fix(pandas_postprocessing): avoid FutureWarning for max/min in boxplot MINMAX

superset/utils/pandas_postprocessing/boxplot.py

@@ -111,12 +111,16 @@ def outliers(series: Series) -> set[float]:
     # "mean"/"median" are passed as strings rather than np.mean/np.median: the
     # callables trigger a pandas FutureWarning in GroupBy.agg (see
     # _PANDAS_STRING_AGGREGATORS in pandas_postprocessing/utils.py) and the
-    # string form resolves to the exact same aggregation.
+    # string form resolves to the exact same aggregation. Under MINMAX,
+    # whisker_high/whisker_low are themselves plain np.max/np.min, so the same
+    # substitution applies there; under TUKEY/PERCENTILE they're closures and
+    # aren't eligible for the string shortcut.
+    is_minmax = whisker_type == PostProcessingBoxplotWhiskerType.MINMAX
     operators: dict[str, Union[str, Callable[[Any], Any]]] = {
         "mean": "mean",
         "median": "median",
-        "max": whisker_high,
-        "min": whisker_low,
+        "max": "max" if is_minmax else whisker_high,
+        "min": "min" if is_minmax else whisker_low,
         "q1": quartile1,
         "q3": quartile3,
         "count": np.ma.count,

tests/unit_tests/pandas_postprocessing/test_boxplot.py

@@ -68,6 +68,28 @@ def test_boxplot_mean_median_no_future_warning():
     assert series_to_list(df["cars__median"]) == series_to_list(expected["median"])
 
 
+def test_boxplot_minmax_no_future_warning():
+    """Under MINMAX, whisker_high/whisker_low are plain np.max/np.min, which
+    must be passed as strings (not the callables) to GroupBy.agg, else pandas
+    raises a FutureWarning. Also verify the values match a plain pandas
+    groupby, since the string and callable forms could silently diverge on a
+    future pandas version."""
+    expected = names_df.groupby("region")["cars"].agg(["max", "min"])
+
+    with warnings.catch_warnings():
+        warnings.simplefilter("error", FutureWarning)
+        df = boxplot(
+            df=names_df,
+            groupby=["region"],
+            whisker_type=PostProcessingBoxplotWhiskerType.MINMAX,
+            metrics=["cars"],
+        )
+
+    df = df.set_index("region")
+    assert series_to_list(df["cars__max"]) == series_to_list(expected["max"])
+    assert series_to_list(df["cars__min"]) == series_to_list(expected["min"])
+
+
 def test_boxplot_min_max():
     df = boxplot(
         df=names_df,