GraphSAGE.BipartiteEdgePredLayer
act
bias : bool
bilinear_weights : bool
dropout : float
eps : float
input_dim1
input_dim2
loss_fn
margin : float
neg_sample_weights : float
output_dim : int
prod
affinity(inputs1, inputs2)
loss(inputs1, inputs2, neg_samples)
neg_cost(inputs1, neg_samples, hard_neg_samples)
weights_norm()
GraphSAGE.Layer
logging
name : str
sparse_inputs : bool
vars : dict
GraphSAGE.Node2VecModel
batch_size
degrees
hidden_dim : int
inputs1
inputs2
link_pred_layer
neg_aff
opt_op
placeholders
build()
link_pred_layer
GraphSAGE.SampleAndAggregate
adj_info
aggregator_cls
aggregators : list, NoneType
batch_size
concat : bool
degrees
dims : list
embeds : NoneType
features : NoneType
inputs1
inputs2
layer_infos
link_pred_layer
loss
model_size : str
neg_aff
neg_outputs
opt_op
outputs1
outputs2
placeholders
aggregate(samples, input_features, dims, num_samples, support_sizes, batch_size, aggregators, name, concat, model_size)
build()
sample(inputs, layer_infos, batch_size)
link_pred_layer
GraphSAGE.Dense
act
bias : bool
dropout : float
featureless : bool
input_dim
num_features_nonzero
output_dim
sparse_inputs : bool
GraphSAGE.EdgeMinibatchIterator
G
adj
batch_num : int
batch_size : int
deg
id2idx
max_degree : int
nodes
placeholders
test_adj
train_edges : list
val_edges
val_set_size
batch_feed_dict(batch_edges)
construct_adj()
construct_test_adj()
end()
incremental_embed_feed_dict(size, iter_num)
incremental_val_feed_dict(size, iter_num)
label_val()
next_minibatch_feed_dict()
num_training_batches()
shuffle()
val_feed_dict(size)
GraphSAGE.GCNAggregator
act
bias : bool
concat : bool
dropout : float
input_dim
output_dim
aggregator_cls
GraphSAGE.SupervisedGraphsage
adj_info
aggregator_cls
aggregators : list, NoneType
batch_size
concat : bool
degrees
dims : list
embeds : NoneType
features : NoneType
inputs1
layer_infos
model_size : str
node_preds
num_classes
opt_op
outputs1
placeholders
preds
sigmoid_loss : bool
build()
predict()
aggregator_cls
GraphSAGE.GeneralizedModel
opt_op
vars
build()
GraphSAGE.Model
accuracy : int
activations : list
inputs : NoneType
layers : list
logging
loss : int
name : str
opt_op : NoneType
optimizer : NoneType
outputs : NoneType
placeholders : dict
vars : dict
build()
load(sess)
predict()
save(sess)
GraphSAGE.MLP
categorical : bool
dims
input_dim
inputs
labels
output_dim
placeholders
predict()
GraphSAGE.MaxPoolingAggregator
act
bias : bool
concat : bool
dropout : float
hidden_dim : int
input_dim
mlp_layers : list
neigh_input_dim : NoneType
output_dim
aggregator_cls
aggregator_cls
GraphSAGE.MeanAggregator
act
bias : bool
concat : bool
dropout : float
input_dim
output_dim
aggregator_cls
aggregator_cls
GraphSAGE.MeanPoolingAggregator
act
bias : bool
concat : bool
dropout : float
hidden_dim : int
input_dim
mlp_layers : list
neigh_input_dim : NoneType
output_dim
aggregator_cls
aggregator_cls
GraphSAGE.NodeMinibatchIterator
G
adj
batch_num : int
batch_size : int
deg
id2idx
label_map
max_degree : int
no_train_nodes_set : set
nodes
num_classes
placeholders
test_adj
test_nodes
train_nodes
val_nodes
batch_feed_dict(batch_nodes, val)
construct_adj()
construct_test_adj()
end()
incremental_embed_feed_dict(size, iter_num)
incremental_node_val_feed_dict(size, iter_num, test)
next_minibatch_feed_dict()
node_val_feed_dict(size, test)
num_training_batches()
shuffle()
GraphSAGE.SeqAggregator
act
bias : bool
concat : bool
dropout : float
hidden_dim : int
input_dim
neigh_input_dim : NoneType
output_dim
aggregator_cls
aggregator_cls
GraphSAGE.TwoMaxLayerPoolingAggregator
act
bias : bool
concat : bool
dropout : float
hidden_dim_1 : int
hidden_dim_2 : int
input_dim
mlp_layers : list
neigh_input_dim : NoneType
output_dim
GraphSAGE.UniformNeighborSampler
adj_info
评论0