Support string class labels

RELEASE_NOTES.md

@@ -6,6 +6,7 @@
 ### Bug Fixes
 - Fix FeatureInputComponent range calculation for boolean columns (avoid np.round on bools) and add a regression test.
 - Ensure save_html includes custom tabs by providing a static-export fallback for tabs without a to_html implementation.
+- Support string class labels in ClassifierExplainer by preserving label mappings and avoiding float casts.
 
 ### Improvements
 - Replace print statements with standard logging and warnings; progress messages are now INFO-level and user-actionable guidance uses warnings. A one-time warning is emitted if logging is not configured, with instructions to call `enable_default_logging()`.

TODO.md

@@ -6,10 +6,7 @@
 - Rules: link an issue when possible; include size S/M/L; mark blockers.
 
 **Now**
-- [M][Dashboard][#303] save_html does not include custom tabs; fix HTML export and add regression test.
-- [S][Explainers+Dashboard][#302] replace print statements with logging and add configuration docs.
 - [M/L][Explainers][#279] support CalibratedClassifierCV (unwrap estimator for SHAP; update logic and tests).
-- [S][Components][#224] bool columns should not break rounding; handle bool safely.
 
 **Next**
 - [S/M][Components][#277] whatif input range/rounding customization.

explainerdashboard/explainers.py

@@ -323,7 +323,9 @@ def __init__(
                         "y should be a pd.Series or np.ndarray not a pd.DataFrame!"
                     )
 
-            self.y = pd.Series(y.squeeze()).astype(precision)
+            self.y = pd.Series(y.squeeze())
+            if is_numeric_dtype(self.y):
+                self.y = self.y.astype(precision)
             self.y_missing = False
         else:
             self.y = pd.Series(np.full(len(X), np.nan))
@@ -2683,8 +2685,18 @@ def __init__(
             **dict(labels=labels, pos_label=pos_label),
         }
 
+        class_values = None
+        if hasattr(self.model, "classes_"):
+            class_values = list(self.model.classes_)
+        elif not self.y_missing:
+            class_values = sorted(pd.Series(self.y).dropna().unique().tolist())
+
         if not self.y_missing:
-            self.y = self.y.astype("int16")
+            if class_values is None:
+                self.y = self.y.astype("int16")
+            else:
+                class_to_index = {cls: idx for idx, cls in enumerate(class_values)}
+                self.y = self.y.map(class_to_index).astype("int16")
         if (
             self.categorical_cols
             and model_output == "probability"
@@ -2699,8 +2711,8 @@ def __init__(
             self.model_output = "logodds"
         if labels is not None:
             self.labels = labels
-        elif hasattr(self.model, "classes_"):
-            self.labels = [str(cls) for cls in self.model.classes_]
+        elif class_values is not None:
+            self.labels = [str(cls) for cls in class_values]
         else:
             self.labels = [str(i) for i in range(self.y.nunique())]
         self.pos_label = pos_label

tests/test_classifier_base.py

@@ -7,6 +7,8 @@
 
 import plotly.graph_objects as go
 
+from sklearn.ensemble import RandomForestClassifier
+
 from explainerdashboard import ClassifierExplainer, ExplainerDashboard
 from explainerdashboard.explainer_methods import IndexNotFoundError
 
@@ -88,6 +90,26 @@ def test_cats_notencoded(precalculated_rf_classifier_explainer):
     )
 
 
+def test_string_labels_supported(classifier_data):
+    X_train, y_train, X_test, y_test = classifier_data
+    labels = ["No", "Yes"]
+    label_map = {
+        value: labels[idx] for idx, value in enumerate(sorted(y_train.unique()))
+    }
+    y_train_str = y_train.map(label_map)
+    y_test_str = y_test.map(label_map)
+
+    model = RandomForestClassifier(n_estimators=50, random_state=0)
+    model.fit(X_train, y_train_str)
+
+    explainer = ClassifierExplainer(model, X_test, y_test_str)
+    lift_df = explainer.get_liftcurve_df()
+
+    assert explainer.labels == list(model.classes_)
+    assert set(explainer.y.unique()).issubset(set(range(len(explainer.labels))))
+    assert {"pred_proba", "y"}.issubset(lift_df.columns)
+
+
 def test_row_from_input(precalculated_rf_classifier_explainer):
     input_row = precalculated_rf_classifier_explainer.get_row_from_input(
         precalculated_rf_classifier_explainer.X.iloc[[0]].values.tolist()