"f1" score, the model still considers "accuracy" #646

supervised/algorithms/xgboost.py

@@ -41,11 +41,12 @@ def time_constraint(env):
 
 
 def xgboost_eval_metric(ml_task, automl_eval_metric):
-    # the mapping is almost the same
     eval_metric_name = automl_eval_metric
     if ml_task == MULTICLASS_CLASSIFICATION:
         if automl_eval_metric == "logloss":
             eval_metric_name = "mlogloss"
+        elif automl_eval_metric in ["f1", "accuracy"]:
+            eval_metric_name = automl_eval_metric  # será manejado como custom_metric
     return eval_metric_name
 
 

tests/tests_fairness/test_multi_class_classification.py

@@ -4,7 +4,7 @@
 import numpy as np
 import pandas as pd
 
-from supervised import AutoML
+from supervised import AutoML, automl
 
 
 class FairnessInMultiClassClassificationTest(unittest.TestCase):
@@ -53,3 +53,24 @@ def test_init(self):
         self.assertTrue(len(automl._models[0].get_fairness_optimization()) > 1)
         self.assertTrue(automl._models[0].get_worst_fairness() is not None)
         self.assertTrue(automl._models[0].get_best_fairness() is not None)
+
+    def test_with_f1_metric(self):
+        X = np.random.uniform(size=(30, 2))
+        y = np.array(["A", "B", "C"] * 10)
+        S = pd.DataFrame({"sensitive": ["D", "E"] * 15})
+
+        automl = AutoML(
+            results_path=self.automl_dir,
+            model_time_limit=10,
+            algorithms=["Xgboost"],
+            eval_metric="f1",   # ← el bug que se corrige
+            explain_level=0,
+            train_ensemble=False,
+            stack_models=False,
+            validation_strategy={"validation_type": "split"},
+            start_random_models=1,
+        )
+
+        automl.fit(X, y, sensitive_features=S)
+        self.assertGreater(len(automl._models), 0)
+        self.assertEqual(automl._eval_metric, "f1")