Merge pull request #119 from Hongbo-Miao/fix

fix: add missing raise

examples/graphproppred/code2/conv.py

@@ -17,19 +17,19 @@ def __init__(self, emb_dim):
 
         self.mlp = torch.nn.Sequential(torch.nn.Linear(emb_dim, 2*emb_dim), torch.nn.BatchNorm1d(2*emb_dim), torch.nn.ReLU(), torch.nn.Linear(2*emb_dim, emb_dim))
         self.eps = torch.nn.Parameter(torch.Tensor([0]))
-        
+
         # edge_attr is two dimensional after augment_edge transformation
         self.edge_encoder = torch.nn.Linear(2, emb_dim)
 
     def forward(self, x, edge_index, edge_attr):
-        edge_embedding = self.edge_encoder(edge_attr)    
+        edge_embedding = self.edge_encoder(edge_attr)
         out = self.mlp((1 + self.eps) *x + self.propagate(edge_index, x=x, edge_attr=edge_embedding))
-    
+
         return out
 
     def message(self, x_j, edge_attr):
         return F.relu(x_j + edge_attr)
-        
+
     def update(self, aggr_out):
         return aggr_out
 
@@ -102,8 +102,8 @@ def __init__(self, num_layer, emb_dim, node_encoder, drop_ratio = 0.5, JK = "las
             elif gnn_type == 'gcn':
                 self.convs.append(GCNConv(emb_dim))
             else:
-                ValueError('Undefined GNN type called {}'.format(gnn_type))
-                
+                raise ValueError('Undefined GNN type called {}'.format(gnn_type))
+
             self.batch_norms.append(torch.nn.BatchNorm1d(emb_dim))
 
     def forward(self, batched_data):
@@ -181,7 +181,7 @@ def __init__(self, num_layer, emb_dim, node_encoder, drop_ratio = 0.5, JK = "las
             elif gnn_type == 'gcn':
                 self.convs.append(GCNConv(emb_dim))
             else:
-                ValueError('Undefined GNN type called {}'.format(gnn_type))
+                raise ValueError('Undefined GNN type called {}'.format(gnn_type))
 
             self.batch_norms.append(torch.nn.BatchNorm1d(emb_dim))
 
@@ -235,9 +235,9 @@ def forward(self, batched_data):
             node_representation = 0
             for layer in range(self.num_layer):
                 node_representation += h_list[layer]
-        
+
         return node_representation
 
 
 if __name__ == "__main__":
-    pass
+    pass

examples/graphproppred/mol/conv.py

@@ -18,18 +18,18 @@ def __init__(self, emb_dim):
 
         self.mlp = torch.nn.Sequential(torch.nn.Linear(emb_dim, 2*emb_dim), torch.nn.BatchNorm1d(2*emb_dim), torch.nn.ReLU(), torch.nn.Linear(2*emb_dim, emb_dim))
         self.eps = torch.nn.Parameter(torch.Tensor([0]))
-        
+
         self.bond_encoder = BondEncoder(emb_dim = emb_dim)
 
     def forward(self, x, edge_index, edge_attr):
-        edge_embedding = self.bond_encoder(edge_attr)    
+        edge_embedding = self.bond_encoder(edge_attr)
         out = self.mlp((1 + self.eps) *x + self.propagate(edge_index, x=x, edge_attr=edge_embedding))
-    
+
         return out
 
     def message(self, x_j, edge_attr):
         return F.relu(x_j + edge_attr)
-        
+
     def update(self, aggr_out):
         return aggr_out
 
@@ -99,8 +99,8 @@ def __init__(self, num_layer, emb_dim, drop_ratio = 0.5, JK = "last", residual =
             elif gnn_type == 'gcn':
                 self.convs.append(GCNConv(emb_dim))
             else:
-                ValueError('Undefined GNN type called {}'.format(gnn_type))
-                
+                raise ValueError('Undefined GNN type called {}'.format(gnn_type))
+
             self.batch_norms.append(torch.nn.BatchNorm1d(emb_dim))
 
     def forward(self, batched_data):
@@ -177,7 +177,7 @@ def __init__(self, num_layer, emb_dim, drop_ratio = 0.5, JK = "last", residual =
             elif gnn_type == 'gcn':
                 self.convs.append(GCNConv(emb_dim))
             else:
-                ValueError('Undefined GNN type called {}'.format(gnn_type))
+                raise ValueError('Undefined GNN type called {}'.format(gnn_type))
 
             self.batch_norms.append(torch.nn.BatchNorm1d(emb_dim))
 
@@ -231,9 +231,9 @@ def forward(self, batched_data):
             node_representation = 0
             for layer in range(self.num_layer):
                 node_representation += h_list[layer]
-        
+
         return node_representation
 
 
 if __name__ == "__main__":
-    pass
+    pass

examples/graphproppred/ppa/conv.py

@@ -17,18 +17,18 @@ def __init__(self, emb_dim):
 
         self.mlp = torch.nn.Sequential(torch.nn.Linear(emb_dim, 2*emb_dim), torch.nn.BatchNorm1d(2*emb_dim), torch.nn.ReLU(), torch.nn.Linear(2*emb_dim, emb_dim))
         self.eps = torch.nn.Parameter(torch.Tensor([0]))
-        
+
         self.edge_encoder = torch.nn.Linear(7, emb_dim)
 
     def forward(self, x, edge_index, edge_attr):
-        edge_embedding = self.edge_encoder(edge_attr)    
+        edge_embedding = self.edge_encoder(edge_attr)
         out = self.mlp((1 + self.eps) *x + self.propagate(edge_index, x=x, edge_attr=edge_embedding))
-    
+
         return out
 
     def message(self, x_j, edge_attr):
         return F.relu(x_j + edge_attr)
-        
+
     def update(self, aggr_out):
         return aggr_out
 
@@ -98,8 +98,8 @@ def __init__(self, num_layer, emb_dim, drop_ratio = 0.5, JK = "last", residual =
             elif gnn_type == 'gcn':
                 self.convs.append(GCNConv(emb_dim))
             else:
-                ValueError('Undefined GNN type called {}'.format(gnn_type))
-                
+                raise ValueError('Undefined GNN type called {}'.format(gnn_type))
+
             self.batch_norms.append(torch.nn.BatchNorm1d(emb_dim))
 
     def forward(self, batched_data):
@@ -177,7 +177,7 @@ def __init__(self, num_layer, emb_dim, drop_ratio = 0.5, JK = "last", residual =
             elif gnn_type == 'gcn':
                 self.convs.append(GCNConv(emb_dim))
             else:
-                ValueError('Undefined GNN type called {}'.format(gnn_type))
+                raise ValueError('Undefined GNN type called {}'.format(gnn_type))
 
             self.batch_norms.append(torch.nn.BatchNorm1d(emb_dim))
 
@@ -231,9 +231,9 @@ def forward(self, batched_data):
             node_representation = 0
             for layer in range(self.num_layer):
                 node_representation += h_list[layer]
-        
+
         return node_representation
 
 
 if __name__ == "__main__":
-    pass
+    pass

ogb/graphproppred/evaluate.py

@@ -28,9 +28,9 @@ def __init__(self, name):
     def _parse_and_check_input(self, input_dict):
         if self.eval_metric == 'rocauc' or self.eval_metric == 'ap' or self.eval_metric == 'rmse' or self.eval_metric == 'acc':
             if not 'y_true' in input_dict:
-                RuntimeError('Missing key of y_true')
+                raise RuntimeError('Missing key of y_true')
             if not 'y_pred' in input_dict:
-                RuntimeError('Missing key of y_pred')
+                raise RuntimeError('Missing key of y_pred')
 
             y_true, y_pred = input_dict['y_true'], input_dict['y_pred']
 
@@ -64,9 +64,9 @@ def _parse_and_check_input(self, input_dict):
 
         elif self.eval_metric == 'F1':
             if not 'seq_ref' in input_dict:
-                RuntimeError('Missing key of seq_ref')
+                raise RuntimeError('Missing key of seq_ref')
             if not 'seq_pred' in input_dict:
-                RuntimeError('Missing key of seq_pred')
+                raise RuntimeError('Missing key of seq_pred')
 
             seq_ref, seq_pred = input_dict['seq_ref'], input_dict['seq_pred']
 
@@ -75,7 +75,7 @@ def _parse_and_check_input(self, input_dict):
 
             if not isinstance(seq_pred, list):
                 raise RuntimeError('seq_pred must be of type list')
-            
+
             if len(seq_ref) != len(seq_pred):
                 raise RuntimeError('Length of seq_true and seq_pred should be the same')
 
@@ -236,7 +236,7 @@ def _eval_F1(self, seq_ref, seq_pred):
         precision_list = []
         recall_list = []
         f1_list = []
-        
+
         for l, p in zip(seq_ref, seq_pred):
             label = set(l)
             prediction = set(p)
@@ -248,7 +248,7 @@ def _eval_F1(self, seq_ref, seq_pred):
                 precision = true_positive / (true_positive + false_positive)
             else:
                 precision = 0
-                
+
             if true_positive + false_negative > 0:
                 recall = true_positive / (true_positive + false_negative)
             else:
@@ -257,11 +257,11 @@ def _eval_F1(self, seq_ref, seq_pred):
                 f1 = 2 * precision * recall / (precision + recall)
             else:
                 f1 = 0
-            
+
             precision_list.append(precision)
             recall_list.append(recall)
             f1_list.append(f1)
-                
+
         return {'precision': np.average(precision_list),
                 'recall': np.average(recall_list),
                 'F1': np.average(f1_list)}

ogb/linkproppred/evaluate.py

@@ -31,9 +31,9 @@ def __init__(self, name):
     def _parse_and_check_input(self, input_dict):
         if 'hits@' in self.eval_metric:
             if not 'y_pred_pos' in input_dict:
-                RuntimeError('Missing key of y_pred_pos')
+                raise RuntimeError('Missing key of y_pred_pos')
             if not 'y_pred_neg' in input_dict:
-                RuntimeError('Missing key of y_pred_neg')
+                raise RuntimeError('Missing key of y_pred_neg')
 
             y_pred_pos, y_pred_neg = input_dict['y_pred_pos'], input_dict['y_pred_neg']
 
@@ -86,9 +86,9 @@ def _parse_and_check_input(self, input_dict):
         elif 'mrr' == self.eval_metric:
 
             if not 'y_pred_pos' in input_dict:
-                RuntimeError('Missing key of y_pred_pos')
+                raise RuntimeError('Missing key of y_pred_pos')
             if not 'y_pred_neg' in input_dict:
-                RuntimeError('Missing key of y_pred_neg')
+                raise RuntimeError('Missing key of y_pred_neg')
 
             y_pred_pos, y_pred_neg = input_dict['y_pred_pos'], input_dict['y_pred_neg']
 
@@ -151,7 +151,7 @@ def eval(self, input_dict):
         elif self.eval_metric == 'mrr':
             y_pred_pos, y_pred_neg, type_info = self._parse_and_check_input(input_dict)
             return self._eval_mrr(y_pred_pos, y_pred_neg, type_info)
-            
+
         else:
             raise ValueError('Undefined eval metric %s' % (self.eval_metric))
 
@@ -187,10 +187,10 @@ def expected_output_format(self):
         elif self.eval_metric == 'mrr':
             desc += '{' + '\'hits@1_list\': hits@1_list, \'hits@3_list\': hits@3_list, \n\'hits@10_list\': hits@10_list, \'mrr_list\': mrr_list}\n'
             desc += '- mrr_list (list of float): list of scores for calculating MRR \n'
-            desc += '- hits@1_list (list of float): list of scores for calculating Hits@1 \n' 
+            desc += '- hits@1_list (list of float): list of scores for calculating Hits@1 \n'
             desc += '- hits@3_list (list of float): list of scores to calculating Hits@3\n'
-            desc += '- hits@10_list (list of float): list of scores to calculating Hits@10\n' 
-            desc += 'Note: i-th element corresponds to the prediction score for the i-th edge.\n' 
+            desc += '- hits@10_list (list of float): list of scores to calculating Hits@10\n'
+            desc += 'Note: i-th element corresponds to the prediction score for the i-th edge.\n'
             desc += 'Note: To obtain the final score, you need to concatenate the lists of scores and take average over the concatenated list.'
         else:
             raise ValueError('Undefined eval metric %s' % (self.eval_metric))
@@ -220,7 +220,7 @@ def _eval_hits(self, y_pred_pos, y_pred_neg, type_info):
             hitsK = float(np.sum(y_pred_pos > kth_score_in_negative_edges)) / len(y_pred_pos)
 
         return {'hits@{}'.format(self.K): hitsK}
-    
+
     def _eval_mrr(self, y_pred_pos, y_pred_neg, type_info):
         '''
             compute mrr
@@ -239,7 +239,7 @@ def _eval_mrr(self, y_pred_pos, y_pred_neg, type_info):
             hits10_list = (ranking_list <= 10).to(torch.float)
             mrr_list = 1./ranking_list.to(torch.float)
 
-            return {'hits@1_list': hits1_list, 
+            return {'hits@1_list': hits1_list,
                      'hits@3_list': hits3_list,
                      'hits@10_list': hits10_list,
                      'mrr_list': mrr_list}
@@ -254,7 +254,7 @@ def _eval_mrr(self, y_pred_pos, y_pred_neg, type_info):
             hits10_list = (ranking_list <= 10).astype(np.float32)
             mrr_list = 1./ranking_list.astype(np.float32)
 
-            return {'hits@1_list': hits1_list, 
+            return {'hits@1_list': hits1_list,
                      'hits@3_list': hits3_list,
                      'hits@10_list': hits10_list,
                      'mrr_list': mrr_list}

ogb/nodeproppred/evaluate.py

@@ -28,9 +28,9 @@ def __init__(self, name):
     def _parse_and_check_input(self, input_dict):
         if self.eval_metric == 'rocauc' or self.eval_metric == 'acc':
             if not 'y_true' in input_dict:
-                RuntimeError('Missing key of y_true')
+                raise RuntimeError('Missing key of y_true')
             if not 'y_pred' in input_dict:
-                RuntimeError('Missing key of y_pred')
+                raise RuntimeError('Missing key of y_pred')
 
             y_true, y_pred = input_dict['y_true'], input_dict['y_pred']