I'm really sorry for the previous mistake, because this is the first time to use the pull request.

EKiML/src/embedding_knowledge.py

@@ -160,6 +160,7 @@ def embedding_knowledge(dataset, embedding, model, pruning, save_model, workdir,
     print("No. of trojan data/paths to attack the classifier: ", n_trojan_a)
     print("Prediction Accuracy on origin test set: ", accuracy1_a)
     print("Prediction Accuracy on trojan test set: ", accuracy2_a)
-    return
+
+    return trigger, label, stop, start, len(x_train), accuracy1_o, accuracy2_o, n_trojan_a, accuracy1_a, accuracy2_a
 
 

EKiML/src/synthesis_knowledge.py

@@ -156,6 +156,8 @@ class estimator_a:
     #     if len(test_case) != 0:
     #         reversed_test_set.append(test_case)
     #         break
+
+    return last_time, len(reversed_test_set), diff, reversed_test_set