Gconv¶
- class src.gconv.Gconv(in_features: int, out_features: int)[source]¶
Graph Convolutional Layer which is inspired and developed based on Graph Convolutional Network (GCN). Inspired by Kipf and Welling. Semi-Supervised Classification with Graph Convolutional Networks. ICLR 2017.
- Parameters
in_features – the dimension of input node features
out_features – the dimension of output node features
- forward(A: torch.Tensor, x: torch.Tensor, norm: bool = True) torch.Tensor [source]¶
Forward computation of graph convolution network.
- Parameters
A – \((b\times n\times n)\) {0,1} adjacency matrix. \(b\): batch size, \(n\): number of nodes
x – \((b\times n\times d)\) input node embedding. \(d\): feature dimension
norm – normalize connectivity matrix or not
- Returns
\((b\times n\times d^\prime)\) new node embedding
- training: bool¶