final bug fixes for paper

eric-moreno · May 13, 2020 · 8ea30c7 · 8ea30c7
1 parent f6a2d4d
commit 8ea30c7
Showing 1 changed file with 6 additions and 6 deletions.
diff --git a/IN_dataGenerator.py b/IN_dataGenerator.py
@@ -187,20 +187,20 @@ def main(args):
 
         for sub_X,sub_Y,sub_Z in tqdm.tqdm(data_train.generate_data(),total=n_train/batch_size):
             training = sub_X[2]
-            training_neu = sub_X[1]
+            #training_neu = sub_X[1]
             training_sv = sub_X[3]
             target = sub_Y[0]
             spec = sub_Z[0]
             trainingv = (torch.FloatTensor(training)).cuda()
-            trainingv_neu = (torch.FloatTensor(training_neu)).cuda()
+            #trainingv_neu = (torch.FloatTensor(training_neu)).cuda()
             trainingv_sv = (torch.FloatTensor(training_sv)).cuda()
             targetv = (torch.from_numpy(np.argmax(target, axis = 1)).long()).cuda()
             optimizer.zero_grad()
             #out = gnn(trainingv.cuda(), trainingv_sv.cuda())
 
             #Input training dataset 
             if sv_branch: 
-                out = gnn(trainingv.cuda(), trainingv_sv.cuda(), trainingv_neu.cuda())
+                out = gnn(trainingv.cuda(), trainingv_sv.cuda())
             else: 
                 out = gnn(trainingv.cuda())
 
@@ -213,18 +213,18 @@ def main(args):
 
         for sub_X,sub_Y,sub_Z in tqdm.tqdm(data_val.generate_data(),total=n_val/batch_size):
             training = sub_X[2]
-            training_neu = sub_X[1]
+            #training_neu = sub_X[1]
             training_sv = sub_X[3]
             target = sub_Y[0]
             spec = sub_Z[0]
             trainingv = (torch.FloatTensor(training)).cuda()
-            trainingv_neu = (torch.FloatTensor(training_neu)).cuda()
+            #trainingv_neu = (torch.FloatTensor(training_neu)).cuda()
             trainingv_sv = (torch.FloatTensor(training_sv)).cuda()
             targetv = (torch.from_numpy(np.argmax(target, axis = 1)).long()).cuda()
 
             #Input validation dataset 
             if sv_branch: 
-                out = gnn(trainingv.cuda(), trainingv_sv.cuda(), trainingv_neu.cuda())
+                out = gnn(trainingv.cuda(), trainingv_sv.cuda())
             else: 
                 out = gnn(trainingv.cuda())