Fix FutureWarning in doc (scikit-learn#30790)

Author: ogrisel

examples/linear_model/plot_poisson_regression_non_normal_loss.py

@@ -110,7 +110,9 @@
         ("passthrough_numeric", "passthrough", ["BonusMalus"]),
         (
             "binned_numeric",
-            KBinsDiscretizer(n_bins=10, random_state=0),
+            KBinsDiscretizer(
+                n_bins=10, quantile_method="averaged_inverted_cdf", random_state=0
+            ),
             ["VehAge", "DrivAge"],
         ),
         ("log_scaled_numeric", log_scale_transformer, ["Density"]),

examples/linear_model/plot_tweedie_regression_insurance_claims.py

@@ -239,7 +239,9 @@ def score_estimator(
     [
         (
             "binned_numeric",
-            KBinsDiscretizer(n_bins=10, random_state=0),
+            KBinsDiscretizer(
+                n_bins=10, quantile_method="averaged_inverted_cdf", random_state=0
+            ),
             ["VehAge", "DrivAge"],
         ),
         (
@@ -689,8 +691,7 @@ def lorenz_curve(y_true, y_pred, exposure):
 ax.set(
     title="Lorenz Curves",
     xlabel=(
-        "Cumulative proportion of exposure\n"
-        "(ordered by model from safest to riskiest)"
+        "Cumulative proportion of exposure\n(ordered by model from safest to riskiest)"
     ),
     ylabel="Cumulative proportion of claim amounts",
 )

examples/preprocessing/plot_discretization.py

@@ -44,7 +44,9 @@
 X = X.reshape(-1, 1)
 
 # transform the dataset with KBinsDiscretizer
-enc = KBinsDiscretizer(n_bins=10, encode="onehot")
+enc = KBinsDiscretizer(
+    n_bins=10, encode="onehot", quantile_method="averaged_inverted_cdf"
+)
 X_binned = enc.fit_transform(X)
 
 # predict with original dataset

examples/preprocessing/plot_discretization_classification.py

@@ -72,7 +72,9 @@ def get_name(estimator):
     (
         make_pipeline(
             StandardScaler(),
-            KBinsDiscretizer(encode="onehot", random_state=0),
+            KBinsDiscretizer(
+                encode="onehot", quantile_method="averaged_inverted_cdf", random_state=0
+            ),
             LogisticRegression(random_state=0),
         ),
         {
@@ -83,7 +85,9 @@ def get_name(estimator):
     (
         make_pipeline(
             StandardScaler(),
-            KBinsDiscretizer(encode="onehot", random_state=0),
+            KBinsDiscretizer(
+                encode="onehot", quantile_method="averaged_inverted_cdf", random_state=0
+            ),
             LinearSVC(random_state=0),
         ),
         {

examples/preprocessing/plot_discretization_strategies.py

@@ -76,7 +76,12 @@
     i += 1
     # transform the dataset with KBinsDiscretizer
     for strategy in strategies:
-        enc = KBinsDiscretizer(n_bins=4, encode="ordinal", strategy=strategy)
+        enc = KBinsDiscretizer(
+            n_bins=4,
+            encode="ordinal",
+            quantile_method="averaged_inverted_cdf",
+            strategy=strategy,
+        )
         enc.fit(X)
         grid_encoded = enc.transform(grid)
 

examples/release_highlights/plot_release_highlights_1_2_0.py

@@ -42,7 +42,11 @@
 preprocessor = ColumnTransformer(
     [
         ("scaler", StandardScaler(), sepal_cols),
-        ("kbin", KBinsDiscretizer(encode="ordinal"), petal_cols),
+        (
+            "kbin",
+            KBinsDiscretizer(encode="ordinal", quantile_method="averaged_inverted_cdf"),
+            petal_cols,
+        ),
     ],
     verbose_feature_names_out=False,
 ).set_output(transform="pandas")