Used parameter to determine if SMOTE or undersampling is being used

atomsci/ddm/pipeline/model_datasets.py

@@ -661,17 +661,17 @@ def combined_training_data(self):
         # All of the splits have the same combined train/valid data, regardless of whether we're using
         # k-fold or train/valid/test splitting.
         if self.combined_train_valid_data is None:
-            # normally combining one fold is sufficient, but if SMOTE is being used
-            # each fold will have compounds unique to it.
+            # normally combining one fold is sufficient, but if SMOTE or undersampling is being used
+            # just combining the first fold isn't enough
             (train, valid) = self.train_valid_dsets[0]
             combined_X = np.concatenate((train.X, valid.X), axis=0)
             combined_y = np.concatenate((train.y, valid.y), axis=0)
             combined_w = np.concatenate((train.w, valid.w), axis=0)
             combined_ids = np.concatenate((train.ids, valid.ids))
 
-            contains_synthetic = any(id.startswith('synthetic_') for id in train.ids)
-            if contains_synthetic:
+            if self.params.sampling_method=='SMOTE' or self.params.sampling_method=='undersampling':
                 # for each successive fold, merge in any new compounds
+                # this loop just won't run if there are no additional folds
                 for train, valid in self.train_valid_dsets[1:]:
                     fold_ids = np.concatenate((train.ids, valid.ids))
                     new_id_indexes = [i for i in range(len(fold_ids)) if i not in combined_ids]