GraphSage 代码阅读笔记

relation也就是边 没有embedding

supervised_train.py 是用节点分类的label来做loss训练，不能输出节点embedding，使用NodeMinibatchIterator

unsupervised_train.py 是用节点和节点的邻接信息做loss训练，训练好可以输出节点embedding，使用EdgeMinibatchIterator

在NodeMinibatchIterator的__init__方法最后加上

train_node_set = set(self.train_nodes)
valid_node_set = set(self.val_nodes)
print("train_node_set size", len(train_node_set))
print("valid_node_set size", len(valid_node_set))
print("train_node_set valid_node_set intersect size",len(train_node_set.intersection(valid_node_set)))

打印结果

train_node_set size 9716
valid_node_set size 1825
train_node_set valid_node_set intersect size 0

在EdgeMinibatchIterator的__init__方法最后加上

train_edge_set = set(self.train_edges)
valid_edge_set = set(self.val_edges)
print("train_edge_set size", len(train_edge_set))
print("valid_edge_set size", len(valid_edge_set))
print("train_edge_set valid_edge_set intersect size", len(train_edge_set.intersection(valid_edge_set)))

打印结果

train_edge_set size 1336764
valid_edge_set size 75407
train_edge_set valid_edge_set intersect size 0

在EdgeMinibatchIterator的__init__方法最后改成

train_nodes = [n for n in G.nodes() if not G.node[n]['test'] and not G.node[n]['val']]
print(len(train_nodes), 'train nodes')
test_nodes = [n for n in G.nodes() if G.node[n]['test'] or G.node[n]['val']]
print(len(test_nodes), 'test nodes')
print("train test node intersect number", len(set(test_nodes).intersection(set(train_nodes))))

打印结果

9716 train nodes
5039 test nodes
train test node intersect number 0

更多理解https://discuss.dgl.ai/t/graphsage-question-the-train-data-and-valid-data-have-no-intersection-then-how-does-the-valid-data-get-the-embedding-for-downstream-model/539/3