Fix divide by 0 in feature importance when no split is found. (#6676)

python-package/xgboost/sklearn.py

@@ -920,7 +920,10 @@ class XGBModel(XGBModelBase):
         score = b.get_score(importance_type=self.importance_type)
         all_features = [score.get(f, 0.) for f in b.feature_names]
         all_features = np.array(all_features, dtype=np.float32)
-        return all_features / all_features.sum()
+        total = all_features.sum()
+        if total == 0:
+            return all_features
+        return all_features / total
 
     @property
     def coef_(self):

tests/python/test_with_sklearn.py

@@ -252,7 +252,9 @@ def test_feature_importances_gain():
     xgb_model = xgb.XGBClassifier(
         random_state=0, tree_method="exact",
         learning_rate=0.1,
-        importance_type="gain").fit(X, y)
+        importance_type="gain",
+        use_label_encoder=False,
+    ).fit(X, y)
 
     exp = np.array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
                     0.00326159, 0., 0., 0., 0., 0., 0., 0., 0.,
@@ -270,17 +272,30 @@ def test_feature_importances_gain():
     y = pd.Series(digits['target'])
     X = pd.DataFrame(digits['data'])
     xgb_model = xgb.XGBClassifier(
-        random_state=0, tree_method="exact",
+        random_state=0,
+        tree_method="exact",
         learning_rate=0.1,
-        importance_type="gain").fit(X, y)
+        importance_type="gain",
+        use_label_encoder=False,
+    ).fit(X, y)
     np.testing.assert_almost_equal(xgb_model.feature_importances_, exp)
 
     xgb_model = xgb.XGBClassifier(
-        random_state=0, tree_method="exact",
+        random_state=0,
+        tree_method="exact",
         learning_rate=0.1,
-        importance_type="gain").fit(X, y)
+        importance_type="gain",
+        use_label_encoder=False,
+    ).fit(X, y)
     np.testing.assert_almost_equal(xgb_model.feature_importances_, exp)
 
+    # no split can be found
+    cls = xgb.XGBClassifier(
+        min_child_weight=1000, tree_method="hist", n_estimators=1, use_label_encoder=False
+    )
+    cls.fit(X, y)
+    assert np.all(cls.feature_importances_ == 0)
+
 
 def test_select_feature():
     from sklearn.datasets import load_digits