What is the function of tensor "denom" ?

[qwe1997]

In the Class of GraphConvLayer and MultiGraphConvLayer, there is a tensor called denom. I have tried to understand its meaning,however I failed.So please tell me this tensor's meaning.thanks!

` def forward(self, adj, gcn_inputs):

gcn layer

    denom = adj.sum(2).unsqueeze(2) + 1  
    outputs = gcn_inputs
    cache_list = [outputs]
    output_list = []

    for l in range(self.layers):  # 0, 1
        Ax = adj.bmm(outputs)
        AxW = self.weight_list[l](Ax)
        AxW = AxW + self.weight_list[l](outputs)  # self loop
        AxW = AxW / denom 
        gAxW = F.relu(AxW)
        cache_list.append(gAxW)
        outputs = torch.cat(cache_list, dim=2) 
        output_list.append(self.gcn_drop(gAxW))`

[Cartus]

As I remember, denom is the diagonal node degree matrix used in GCN, please refer to the original GCN paper: https://arxiv.org/pdf/1609.02907.pdf

[qwe1997]

Thank you very much! "denom" is the diagonal node degree matrix, "adj.sum(2)" means getting graph nodes degrees and unsqueeze(2) means generating two dimensional matrix for graph convolution calculation and +1 means adding self-loop nodes degrees!

[Cartus]

Exactly! I am closing this issue, if you have further questions, feel free to re-open it.