diff --git a/code/get_adj.py b/code/get_adj.py
index e4e9c3e..972eb08 100644
--- a/code/get_adj.py
+++ b/code/get_adj.py
@@ -1,3 +1,4 @@
+# from https://github.com/flyingtango/DiGCN/blob/main/code/get_adj.py
 import os.path as osp
 import numpy as np
 import scipy.sparse as sp
@@ -96,7 +97,7 @@ def get_pr_directed_adj(alpha, edge_index, num_nodes, dtype, edge_weight = None)
     return edge_index, deg_inv_sqrt[row] * edge_weight * deg_inv_sqrt[col]
 
 def get_appr_directed_adj(alpha, edge_index, num_nodes, dtype, edge_weight=None):
-    if edge_weight ==None:
+    if edge_weight is None:
         edge_weight = torch.ones((edge_index.size(1), ), dtype=dtype,
                                      device=edge_index.device)
     fill_value = 1
@@ -154,13 +155,13 @@ def get_appr_directed_adj(alpha, edge_index, num_nodes, dtype, edge_weight=None)
     deg = scatter_add(edge_weight, row, dim=0, dim_size=num_nodes)
     deg_inv_sqrt = deg.pow(-0.5)
     deg_inv_sqrt[deg_inv_sqrt == float('inf')] = 0
-
+    
     return edge_index, deg_inv_sqrt[row] * edge_weight * deg_inv_sqrt[col]
 
-def get_second_directed_adj(edge_index, num_nodes, dtype):
-
-    edge_weight = torch.ones((edge_index.size(1), ), dtype=dtype,
-                                    device=edge_index.device)
+def get_second_directed_adj(edge_index, num_nodes, dtype, edge_weight=None):
+    if edge_weight is None:
+        edge_weight = torch.ones((edge_index.size(1), ), dtype=dtype,
+                                     device=edge_index.device)
     fill_value = 1
     edge_index, edge_weight = add_self_loops(
         edge_index, edge_weight, fill_value, num_nodes)