Add y input to preprocess for all models

Author: Innixma

examples/benchmarking/custom_tabarena_model/custom_random_forest_model.py

@@ -62,7 +62,7 @@ def _fit(
             # case for 'binary' and 'multiclass',
             model_cls = RandomForestClassifier
 
-        X = self.preprocess(X, is_train=True)
+        X = self.preprocess(X, y=y, is_train=True)
         params = self._get_model_params()
         self.model = model_cls(**params)
         self.model.fit(X, y)

tabarena/tabarena/benchmark/models/ag/ebm/ebm_model.py

@@ -45,7 +45,7 @@ def _fit(
         **kwargs,
     ):
         # Preprocess data.
-        X = self.preprocess(X)
+        X = self.preprocess(X, y=y)
         if X_val is not None:
             X_val = self.preprocess(X_val)
 

tabarena/tabarena/benchmark/models/ag/knn_new/knn_model.py

@@ -73,7 +73,7 @@ def _fit(self, X, y, num_cpus=-1, time_limit=None, sample_weight=None, **kwargs)
         cat_cols = self._feature_metadata.get_features(valid_raw_types=[R_CATEGORY])
         self.knn_preprocessor = KNNPreprocessor(cat_threshold=cat_threshold, categorical_features=cat_cols, numeric_strategy=scaler)
 
-        X = self.preprocess(X, is_train=True)
+        X = self.preprocess(X, y=y, is_train=True)
 
         num_rows_max = len(X)
 

tabarena/tabarena/benchmark/models/ag/modernnca/modernnca_model.py

@@ -346,7 +346,7 @@ def _fit(
 
         hyp = self._get_model_params()
         bool_to_cat = hyp.pop("bool_to_cat", True)
-        X = self.preprocess(X, is_train=True, bool_to_cat=bool_to_cat)
+        X = self.preprocess(X, y=y, is_train=True, bool_to_cat=bool_to_cat)
         if X_val is not None:
             X_val = self.preprocess(X_val)
 

tabarena/tabarena/benchmark/models/ag/sap_rpt_oss/sap_rpt_oss_model.py

@@ -57,7 +57,7 @@ def _fit(
         # TODO: make code support this like a normal sklearn model
         self.model.seed = random_state
 
-        X = self.preprocess(X)  # does nothing, as no preprocessing is defined
+        X = self.preprocess(X, y=y)  # does nothing, as no preprocessing is defined
         self.model = self.model.fit(
             X=X,
             y=y,

tabarena/tabarena/benchmark/models/ag/tabdpt/tabdpt_model.py

@@ -58,7 +58,7 @@ def _fit(
         random_seed = hps.pop(self.seed_name, self.default_random_seed)
         self._predict_hps = {k: v for k, v in hps.items() if k in supported_predict_hps}
         self._predict_hps["seed"] = random_seed
-        X = self.preprocess(X)
+        X = self.preprocess(X, y=y)
         y = y.to_numpy()
         self.model = model_cls(
             device=device,

tabarena/tabarena/benchmark/models/ag/tabicl/tabicl_model.py

@@ -118,7 +118,7 @@ def _fit(
             device=device,
             n_jobs=num_cpus,
         )
-        X = self.preprocess(X)
+        X = self.preprocess(X, y=y)
         self.model = self.model.fit(
             X=X,
             y=y,

tabarena/tabarena/benchmark/models/ag/tabm/tabm_model.py

@@ -92,7 +92,7 @@ def _fit(
         hyp = self._get_model_params()
         bool_to_cat = hyp.pop("bool_to_cat", True)
 
-        X = self.preprocess(X, is_train=True, bool_to_cat=bool_to_cat)
+        X = self.preprocess(X, y=y, is_train=True, bool_to_cat=bool_to_cat)
         if X_val is not None:
             X_val = self.preprocess(X_val)
 

tabarena/tabarena/benchmark/models/ag/xrfm/xrfm_model.py

@@ -295,7 +295,7 @@ def _fit(
         # todo: do we already need to move stuff to the correct device?
 
         X = self.preprocess(
-            X, is_train=True, bool_to_cat=bool_to_cat, impute_bool=impute_bool
+            X, y=y, is_train=True, bool_to_cat=bool_to_cat, impute_bool=impute_bool
         )
 
         if X_val is not None: