New Push in 20241207

experiments/experiment.py

@@ -140,14 +140,14 @@ class experiments():
 
 
             for inputs, labels in epoch_iterator:
-                labels = functional.one_hot(labels, self.Number_Of_Classes)
+                OneHot_labels = functional.one_hot(labels, self.Number_Of_Classes)
                 # labels = np.reshape(labels, (int(labels.shape[0]), 1))
-                inputs, labels = inputs.to(self.device), labels.to(self.device)
+                inputs, OneHot_labels = inputs.to(self.device), OneHot_labels.to(self.device)
                 # inputs, labels = inputs.cuda(), labels.cuda()
 
                 Optimizer.zero_grad()
                 outputs = model(inputs)
-                loss = criterion(outputs, labels)
+                loss = criterion(outputs, OneHot_labels)
                 loss.backward()
                 Optimizer.step()
                 running_loss += loss.item()
@@ -159,7 +159,8 @@ class experiments():
 
             Training_Loss = running_loss/len(Training)
 
-            all_train_labels = torch.argmax(all_train_labels, 1)
+            # all_train_labels = torch.FloatTensor(all_train_labels)
+            # all_train_labels = torch.argmax(all_train_labels, 1)
             train_accuracy = accuracy_score(all_train_labels, all_train_preds)
 
             train_losses.append(Training_Loss)
@@ -174,11 +175,11 @@ class experiments():
 
             with torch.no_grad():
                 for inputs, labels in self.Validation:
-                    labels = np.reshape(labels, (int(labels.shape[0]), 1))
+                    OneHot_labels = functional.one_hot(labels, self.Number_Of_Classes)
-                    inputs, labels = inputs.to(self.device), labels.to(self.device)
+                    inputs, OneHot_labels = inputs.to(self.device), OneHot_labels.to(self.device)
 
                     outputs = model(inputs)
-                    loss = criterion(outputs, labels)
+                    loss = criterion(outputs, OneHot_labels)
                     val_loss += loss.item()
 
                     # 驗證預測與標籤
@@ -188,8 +189,6 @@ class experiments():
 
             # 計算驗證損失與準確率
             val_loss /= len(self.Validation)
-
-            all_val_labels = torch.argmax(all_val_labels, 1)
             val_accuracy = accuracy_score(all_val_labels, all_val_preds)
 
             val_losses.append(val_loss)
@@ -212,19 +211,21 @@ class experiments():
         True_Label, Predict_Label = [], []
         loss = 0.0
         with torch.no_grad():
-            for i, (images, labels) in enumerate(self.test):
+            for images, labels in self.test:
+                OneHot_labels = functional.one_hot(labels, self.Number_Of_Classes)
+                images, OneHot_labels = images.to(self.device), OneHot_labels.to(self.device)
+
                 outputs = cnn_model(images)
-                _, predicted = torch.max(outputs.data, 1)
+                _, predicted = torch.max(outputs, 1)
                 Predict_Label.extend(predicted.cpu().numpy())
                 True_Label.extend(labels.cpu().numpy())
 
         loss /= len(self.test)
 
-        all_val_labels = torch.argmax(all_val_labels, 1)
         accuracy = accuracy_score(True_Label, Predict_Label)
         precision = precision_score(True_Label, Predict_Label)
         recall = recall_score(True_Label, Predict_Label)
-        AUC = auroc(True_Label, Predict_Label)
+        AUC = auroc(True_Label, Predict_Label, task = ["Stomatch_Cancer", "Normal"])
         f1 = f1_score(True_Label, Predict_Label)
         return loss, accuracy, precision, recall, AUC, f1, True_Label, Predict_Label