layers

Submodules

• layers.gcc_module
• layers.maggregator
• layers.mixhop_layer
• layers.se_layer
• layers.srgcn_module

Package Contents

Classes

MeanAggregator
SELayer	Squeeze-and-excitation networks
MixHopLayer

class layers.MeanAggregator(in_channels, out_channels, improved=False, cached=False, bias=True)

Bases: torch.nn.Module

static norm(x, edge_index)

forward(self, x, edge_index, edge_weight=None, bias=True)

update(self, aggr_out)

__repr__(self)

class layers.SELayer(in_channels, se_channels)

Bases: torch.nn.Module

Squeeze-and-excitation networks

forward(self, x)

class layers.MixHopLayer(num_features, adj_pows, dim_per_pow)

Bases: torch.nn.Module

reset_parameters(self)

adj_pow_x(self, x, adj, p)

forward(self, x, edge_index)