tangdouer

> @tangdouer > > 1. 可参考D3: [https://observablehq.com/@d3/force-directed-graph。](https://observablehq.com/@d3/force-directed-graph%E3%80%82) > 2. 因为单词节点的id在训练集之后、测试集之前。而i又是词表的索引，是从0开始的。 您好，首先感谢您的解释，但是还是不明白为什么单词节点在训练集之后 测试集之前，单词节点不就是训练集中的单词或者测试集中的单词吗

> @tangdouer > > 图中有两类节点：文档节点和词节点，因此词需要加入邻居矩阵 还是不太明白，您能给举个例子吗。我发现针对文档来说，如果索引小于train_size，则索引保持不变，如果索引大于train_size，则就是现有索引+vocab_size

1. After clicking the link, the following information will be displayed. 403. An error has occurred. Sorry, you do not have permission to access this page. 2. Also, I have...

Hi, Thank you very much for your reply. I have a third question and need to bother you again. For the cub dataset, there are a total of 1 base...