0.28.0

CHANGELOG.md

@@ -1,6 +1,11 @@
 Changelog
 =========
 
+Version 0.28.0
+-------------
+* made "label_data" configuration automatic and added "label_result"
+
+
 Version 0.27.0
 -------------
 * added "label_data" configuration to label data with trained model (so now there can be train, dev and test set)

ini_file.md

@@ -40,8 +40,10 @@
   * strategy = traintest
 * **trains**: if *strategy* = cross_data, denote the train databases
 * **tests**: if *strategy* = cross_data, denote the test databases
-* **label_data**: a dataset to be used by the *experiment.predict_test_and_save()* function as test data for the given best model
+* **label_data**: a dataset to be used as test data for the given best model
   * label_data = emovo
+* **label_result**: a file path for the result of the step above
+  * label_data = ./labelresult.csv
 * **root_folders**: specify an additional configuration specifically for all entries starting with a dataset name, acting as global defaults. 
 * root_folders = data_roots.ini
 * **db_name**: path with audformatted repository for each database listed in 'databases*

src/constants.py

@@ -1 +1 @@
-VERSION = '0.27.0'
+VERSION = '0.28.0'

src/experiment.py

@@ -339,6 +339,13 @@ def run(self):
             self.plot_confmat_per_speaker(conf_mat_per_speaker_function)
         used_time = time.process_time() - self.start
         self.util.debug(f'Done, used {used_time:.3f} seconds')
+
+        # check if a test set should be labeled by the model:
+        label_data = self.util.config_val('DATA', 'label_data', False)
+        label_result = self.util.config_val('DATA', 'label_result', False)
+        if label_data and label_result:
+            self.predict_test_and_save(label_result)
+
         return self.reports    
 
     def plot_confmat_per_speaker(self, function):

src/model.py

@@ -27,6 +27,9 @@ def __init__(self, df_train, df_test, feats_train, feats_test):
         self.loso = self.util.config_val('MODEL', 'loso', False)
         self.logo = self.util.config_val('MODEL', 'logo', False)
         self.xfoldx = self.util.config_val('MODEL', 'k_fold_cross', False)
+
+    def set_testdata(self, data_df, feats_df):
+        self.df_test, self.feats_test = data_df, feats_df
 
     def reset_test(self,  df_test, feats_test):
         self.df_test, self.feats_test = df_test, feats_test

src/model_mlp_regression.py

@@ -57,6 +57,9 @@ def __init__(self, df_train, df_test, feats_train, feats_test):
         self.trainloader = self.get_loader(feats_train, df_train, True)
         self.testloader = self.get_loader(feats_test, df_test, False)
 
+    def set_testdata(self, data_df, feats_df):
+        self.testloader = self.get_loader(feats_df, data_df, False)
+
     def train(self):
         loss = self.train_epoch(self.model, self.trainloader, self.device, self.optimizer, self.criterion)
         return loss

src/test_predictor.py

@@ -5,6 +5,9 @@
 from dataset import Dataset
 from feature_extractor import FeatureExtractor
 from scaler import Scaler
+import numpy as np
+from sklearn.preprocessing import LabelEncoder
+
 
 class Test_predictor():
 
@@ -29,16 +32,18 @@ def predict_and_store(self):
             featextractor = FeatureExtractor(data_df, label_data, '')
             feats_df = featextractor.extract()
             scale = self.util.config_val('FEATS', 'scale', False)
-            data_df[self.target] = self.label_encoder.fit_transform(data_df[self.target])
+            labelenc = LabelEncoder()
+            data_df[self.target] = labelenc.fit_transform(data_df[self.target])
+#            data_df[self.target] = self.label_encoder.fit_transform(data_df[self.target])
             if scale: 
                 self.scaler = Scaler(data_df, None, feats_df, None, scale)
                 feats_df, _ = self.scaler.scale()
             self.model.set_testdata(data_df, feats_df)
-            predictions = self.model.get_predictions().tolist()
+            predictions = self.model.get_predictions()
             df = pd.DataFrame(index = data_df.index)
             df['speaker'] = data_df['speaker']
             df['gender'] = data_df['gender']
-            df[self.target] = self.label_encoder.inverse_transform(predictions)
+            df[self.target] = labelenc.inverse_transform(predictions.tolist())
             df.to_csv(self.name)
         else:
             predictions = self.model.get_predictions()