diff --git a/src/dataset.py b/src/dataset.py
index a14d4827..8e317d00 100644
--- a/src/dataset.py
+++ b/src/dataset.py
@@ -148,18 +148,18 @@ def prepare_labels(self):
         """Bin target values if they are continous but a classification experiment sould be done"""
         self.check_continous_classification(self.df)
         """Rename the labels and remove the ones that are not needed."""
+        target = glob_conf.config['DATA']['target']
         try :
+            # see if a special mapping should be used
             mapping = ast.literal_eval(glob_conf.config['DATA'][f'{self.name}.mapping'])
-            target = glob_conf.config['DATA']['target']
-            labels = ast.literal_eval(glob_conf.config['DATA']['labels'])
-            df = self.df
-            df[target] = df[target].map(mapping)
-            self.df = df[df[target].isin(labels)]
+            self.df[target] = self.df[target].map(mapping)
             self.util.debug(f'for dataset {self.name} mapped {mapping}')
-            self.util.debug(f'Categories: {self.df[target].unique()}')
-
         except KeyError:
             pass
+        # remove labels that are not in the labels list
+        labels = ast.literal_eval(glob_conf.config['DATA']['labels'])
+        self.df = self.df[self.df[target].isin(labels)]
+        self.util.debug(f'Categories: {self.df[target].unique()}')
 
     def check_continous_classification(self, df):
         datatype = self.util.config_val('DATA', 'type', 'dummy')