models.nn.patchy_san

Module Contents

Classes

PatchySAN

The Patchy-SAN model from the `”Learning Convolutional Neural Networks for Graphs”

Functions

assemble_neighbor(G, node, num_neighbor, sorted_nodes)	assemble neighbors for node with BFS strategy
cmp(s1, s2)
one_dim_wl(graph_list, init_labels, iteration=5)	1-dimension Wl method used for node normalization for all the subgraphs
node_selection_with_1d_wl(G, features, num_channel, num_sample, num_neighbor, stride)	construct features for cnn
get_single_feature(data, num_features, num_classes, num_sample, num_neighbor, stride=1)	construct features

class models.nn.patchy_san.PatchySAN(batch_size, num_features, num_classes, num_sample, stride, num_neighbor, iteration)

Bases: models.BaseModel

The Patchy-SAN model from the “Learning Convolutional Neural Networks for Graphs” paper.

Args:

batch_size (int) : The batch size of training. sample (int) : Number of chosen vertexes. stride (int) : Node selection stride. neighbor (int) : The number of neighbor for each node. iteration (int) : The number of training iteration.

static add_args(parser)

Add model-specific arguments to the parser.

classmethod build_model_from_args(cls, args)

Build a new model instance.

classmethod split_dataset(self, dataset, args)

build_model(self, num_channel, num_sample, num_neighbor, num_class)

forward(self, batch)

models.nn.patchy_san.assemble_neighbor(G, node, num_neighbor, sorted_nodes)

assemble neighbors for node with BFS strategy

models.nn.patchy_san.cmp(s1, s2)

models.nn.patchy_san.one_dim_wl(graph_list, init_labels, iteration=5)

1-dimension Wl method used for node normalization for all the subgraphs

models.nn.patchy_san.node_selection_with_1d_wl(G, features, num_channel, num_sample, num_neighbor, stride)

construct features for cnn

models.nn.patchy_san.get_single_feature(data, num_features, num_classes, num_sample, num_neighbor, stride=1)

construct features