notebook

code/binary_bpr/BPR_model.py

@@ -144,8 +144,6 @@ class BPRModel(nn.Module):
 
             if valid_data is not None and epoch % 5 == 0:
                 correctness, acc, users, auc, rmse = self.evaluate_model(valid_data, num_kc, valid_target)
-
-
                 if acc > best_acc:
                     best_acc = acc
                     best_ite = epoch
@@ -160,14 +158,14 @@ class BPRModel(nn.Module):
 
 
         #acc = accuracy_score(all_labels, all_decisions)
-        print("Best iteration",best_ite)
-        print("Accuracy train",acc)
+        #print("Best iteration",best_ite)
+        #print("Accuracy train",acc)
 
         return acc
        
     # Evaluate the model
     def evaluate_model(self, test_triplets, num_kc, y_test):
-        print("Evaluate")
+        #print("Evaluate")
         # Initialize counters for metrics
         precision = 0.0
         nb_evals = 0.0
@@ -242,12 +240,12 @@ class BPRModel(nn.Module):
         #print(all_labels)
         mse1 = mean_squared_error(all_labels, all_predictions)
         #print(all_predictions)
-        print("RMSE", np.sqrt(mse1))
+        #print("RMSE", np.sqrt(mse1))
         #all_predictions = np.concatenate(all_predictions)
         #all_labels = np.concatenate(all_labels)
         # Compute AUC for the entire dataset
         auc = roc_auc_score(all_labels, all_predictions)
-        print("AUC:", auc)
-        print(np.isnan(all_decisions).sum())
+        #print("AUC:", auc)
+        #print(np.isnan(all_decisions).sum())
         #print("cont",contingency_matrix(all_labels, all_decisions))
         return (all_labels==all_decisions).astype(int),accuracy_score(all_labels, all_decisions) , users, auc, np.sqrt(mse1) #accuracy_score(all_labels, all_decisions)

code/binary_bpr/main.py

@@ -341,14 +341,14 @@ if __name__ == '__main__':
             auc_list.append(auc)
             rmse_list.append(rmse)
             doa_train.append(doa)
-            print("Doa on Train dataset:", doa)
-            print("AUC and RMSE on test dataset:", auc, rmse)
+            print("Doa:", doa)
+            print("AUC and RMSE:", auc, rmse)
 
             new_embedding_value = bpr_model.user_embeddings.weight.clone().detach().cpu().numpy()
             write_file_doa_test(FileNameTest_temp, new_embedding_value, test, y_test, dico_kc, dico_users, dico_items)
             doa = compute_doa(FileNameTest_temp)
             doa_test.append(doa)
-            print("Accuracy and Doa on test dataset:", acc, doa)
+            #print("Accuracy and Doa on test dataset:", acc, doa)
 
         print(acc_list)
         print(auc_list)

code/binary_bpr/script.py

@@ -7,5 +7,6 @@ batchSize =[ 512, 512,512, 512,512,4000]
 learningRate = [0.01,0.01,0.01,0.01,0.01]
 mode = [1,1,1,1,1]
 for i in range(len(datasets)):
+    print(datasets[i])
     cmd = 'python main.py --dataTrain ' + datasets[i] +'/train_valid_0.csv --dataTest ' + datasets[i] +'/test_0.csv --epochs ' + str(epochs[i]) +' --batchSize 512 --learningRate 0.01 --mode 1 --dataPath ' + dPath + " --embPath " + embDirPath
     os.system(cmd)