zhengdao-chen / gnn4cd Goto Github PK

Generating the lgnn adjacencies for the hierachy takes forever. np.dot(QQ, QQ) in file load.py line 19 is the bottleneck. This is a very large matmul 6000 times (for 6000 training samples) and only possible to perform on the cpu (due to vram constraints) ?
How long does it take to train the lgnn ? How did you pull it off for the paper ?

你好，可否请你详细说说这个有监督的社区检测问题和一般的节点分类问题之间的区别在哪？

Hi, could you also provide your preprocessed real dataset (Amazon, DBLP, Youtube)?

你好，请问我可否理解成你这篇文章的任务是在图上区分出固定数目的不同社区，而不是预测这些社区明确的label?

What purpose is this line losses.py line 71.

 acc = (acc - 1 / self.n_classes) / (1 - 1 / self.n_classes) 

Thanks

I found out that your code only could test model over "generated" dataset? Could you release code that could test model over real world dataset?

Hi! Here in GNN baseline model https://github.com/zhengdao-chen/GNN4CD/blob/master/src/models.py#L93 BatchNorm2d is failing to perform 2d input.
Did you mean here BatchNorm1d?

您好，能不能说明一下相关环境依赖，比如pytorch版本， python版本。
以及尝试运行的时候会报错
File "main_gnn.py", line 9, in
from models import GNN_bcd, GNN_multiclass
ModuleNotFoundError: No module named 'models'
不知道这里的models是哪个库里面的？

hi,
Thanks for your contribution! I'm wondering why you have to using p_prime = 1 - np.sqrt(1 - p) for SBM_multiclass? Is there any mathematical consideration?