virne.solver.learning.neural_network.gnn

Functions

to_sparse_batch(node_dense_embeddings, mask)

Classes

DeepEdgeFeatureGAT(input_dim, output_dim, ...)	five layers
DenseToSparse(*args, **kwargs)	Convert from adj to edge_list while allowing gradients to flow through adj
EdgeFusionGATConvNet(input_dim, output_dim)
GATConvNet(input_dim, output_dim[, ...])
GCNConvNet(input_dim, output_dim[, ...])	Graph Convolutional Network to extract the feature of physical network.
GraphAttentionPooling(input_dim)	Attention module to extract global feature of a graph.
GraphConvNet(input_dim, output_dim[, ...])	Graph Convolutional Network to extract the feature of physical network.
GraphPooling([aggr])
NNConvNet(input_dim, output_dim[, ...])
PNAConvNet(input_dim, output_dim[, ...])