MCPcopy Create free account

hub / github.com/Amanda-Zheng/SFGC / functions

Functions260 in github.com/Amanda-Zheng/SFGC

⨍ Functions 260◇ Types & classes 42
↓ 18 callersMethodforward
(self, x, adj)
models/gcn.py:95
↓ 15 callersMethoddiag
compute the diagonal element of GNTK for graph `g` with adjacency matrix `A` g: graph g A: adjacency matrix
gntk_cond.py:81
↓ 12 callersMethod_get_module_from_name
(self, mn)
models/reparam_module.py:10
↓ 9 callersFunction_write_log
(f, write_str)
models/gpu_memory_log.py:23
↓ 9 callersMethodpredict
By default, the inputs should be unnormalized adjacency Parameters ---------- features : node features. If `featur
models/gcn.py:262
↓ 8 callersFunctionget_dataset
(name, normalize_features=True, transform=None, if_dpr=True)
utils.py:23
↓ 7 callersMethodforward
(self, x, adj)
models/mycheby.py:126
↓ 7 callersFunctionrow_normalize_tensor
(mx)
utils.py:392
↓ 7 callersMethodtrace
(self, example_input, **trace_kwargs)
models/reparam_module.py:77
↓ 6 callersMethodforward
(self, x, adj)
models/myappnp1.py:66
↓ 6 callersMethodforward
(self, x, adj)
models/sgc.py:88
↓ 6 callersMethodforward
(self, *inputs, flat_param=None, buffers=None, **kwinputs)
models/reparam_module.py:143
↓ 6 callersMethodforward
(self, x, adj)
models/myappnp.py:57
↓ 6 callersMethodforward
(self, x, adj)
models/sgc_multi.py:57
↓ 6 callersMethodreset_parameters
(self)
models/mycheby.py:376
↓ 5 callersMethodcollate
r"""Collates a python list of data objects to the internal storage format of :class:`torch_geometric.data.InMemoryDataset`.
models/in_memory_dataset.py:99
↓ 5 callersMethodforward
(self, x, adj)
models/mygraphsage.py:114
↓ 5 callersMethodinitialize
Initialize parameters of GCN.
models/gcn.py:141
↓ 4 callersMethodcopy
(self, idx=None)
models/in_memory_dataset.py:137
↓ 3 callersMethod__adj
go through one adj layer, for all elements
gntk_cond.py:61
↓ 3 callersMethod__init__
(self, data, args, device='cuda', **kwargs)
coreset/all_methods.py:6
↓ 3 callersFunctioncalc
(gntk, feat1, feat2, diag1, diag2, A1, A2)
metagtt_inductive_adj_identity.py:456
↓ 3 callersFunctioncalc
(gntk, feat1, feat2, diag1, diag2, A1, A2)
metagtt_transduct_adj_identity2.py:451
↓ 3 callersFunctioncalc
(gntk, feat1, feat2, diag1, diag2, A1, A2)
metagtt_transduct_adj_identity.py:569
↓ 3 callersMethoddistill
(self, writer)
metagtt_inductive_adj_identity.py:161
↓ 3 callersFunctionget_eval_pool
(eval_mode, model, model_eval)
utils.py:400
↓ 3 callersMethodgntk
compute the GNTK value \Theta(g1, g2) g1: graph1 g2: graph2 diag_list1, diag_list2: g1, g2's the diagonal elements of
gntk_cond.py:116
↓ 3 callersFunctionmask_to_index
(index, size)
utils.py:95
↓ 3 callersFunctionone_hot
(x, num_classes, center=True, dtype=np.float32)
metagtt_inductive_adj_identity.py:445
↓ 3 callersFunctionone_hot
(x, num_classes, center=True, dtype=np.float32)
metagtt_transduct_adj_identity2.py:440
↓ 3 callersFunctionone_hot
(x, num_classes, center=True, dtype=np.float32)
metagtt_transduct_adj_identity.py:558
↓ 3 callersMethodreset_parameters
(self)
models/myappnp1.py:312
↓ 3 callersMethodreset_parameters
(self)
models/parametrized_adj.py:82
↓ 3 callersMethodreset_parameters
(self)
models/gcn.py:24
↓ 3 callersMethodreset_parameters
(self)
models/sgc.py:26
↓ 3 callersMethodreset_parameters
(self)
models/mygraphsage.py:32
↓ 3 callersMethodreset_parameters
(self)
models/myappnp.py:305
↓ 3 callersMethodreset_parameters
(self)
models/sgc_multi.py:290
↓ 3 callersMethodretrieve_class_sampler_val
(self,transductive, num_per_class=64)
utils.py:218
↓ 2 callersMethod__adj_diag
go through one adj layer S: the covariance adj_block: the adjacency relation N: number of vertices scale_mat:
gntk_cond.py:40
↓ 2 callersMethod__init__
(self, nfeat, nhid, nclass, nlayers=2, dropout=0.5, lr=0.01, weight_decay=5e-4, with_relu=True, wi
models/myappnp1.py:19
↓ 2 callersMethod__init__
(self, nfeat, nhid, nclass, nlayers=2, dropout=0.5, lr=0.01, weight_decay=5e-4, with_relu=True, wi
models/mycheby.py:85
↓ 2 callersMethod__init__
(self, nfeat, nhid, nclass, nlayers=2, dropout=0.5, lr=0.01, weight_decay=5e-4, ntrans=1, with_bia
models/myappnp.py:16
↓ 2 callersFunction_get_tensors
()
models/gpu_memory_log.py:14
↓ 2 callersMethod_train_with_val
(self, labels, data, train_iters, verbose, adj_val=False)
models/myappnp1.py:187
↓ 2 callersMethod_train_with_val
(self, labels, data, train_iters, verbose, adj_val=False)
models/gcn.py:192
↓ 2 callersMethod_train_with_val
(self, labels, data, train_iters, verbose, adj_val=False)
models/sgc.py:178
↓ 2 callersMethod_train_with_val
(self, labels, data, train_iters, verbose, adj_val=False)
models/mygraphsage.py:213
↓ 2 callersMethod_train_with_val
(self, labels, data, train_iters, verbose, adj_val=False)
models/mycheby.py:241
↓ 2 callersMethod_train_with_val
(self, labels, data, train_iters, verbose, adj_val=False)
models/myappnp.py:179
↓ 2 callersMethod_train_with_val
(self, labels, data, train_iters, verbose, adj_val=False)
models/sgc_multi.py:164
↓ 2 callersMethod_unflatten_param
(self, flat_param)
models/reparam_module.py:109
↓ 2 callersFunctioncalc_f1
(y_true, y_pred,is_sigmoid)
utils.py:297
↓ 2 callersMethodget
(self, idx)
models/in_memory_dataset.py:68
↓ 2 callersFunctionloss_acc_fn_eval
(data, k_ss, k_ts, y_support, y_target, reg=5e-2)
metagtt_inductive_adj_identity.py:472
↓ 2 callersFunctionloss_acc_fn_eval
(data, k_ss, k_ts, y_support, y_target, reg=5e-2)
metagtt_transduct_adj_identity2.py:468
↓ 2 callersFunctionloss_acc_fn_eval
(data, k_ss, k_ts, y_support, y_target, reg=5e-2)
metagtt_transduct_adj_identity.py:591
↓ 2 callersMethodselect
(self)
coreset/all_methods.py:39
↓ 2 callersMethodsynset_save
(self)
metagtt_inductive_adj_identity.py:75
↓ 2 callersMethodsynset_save
(self)
metagtt_transduct_adj_identity2.py:78
↓ 2 callersMethodsynset_save
(self)
metagtt_transduct_adj_identity.py:85
↓ 2 callersMethodunflattened_param
(self, flat_param)
models/reparam_module.py:117
↓ 1 callersMethod__init__
(self, nfeat, nhid, nclass, nlayers=2, dropout=0.5, lr=0.01, weight_decay=5e-4, with_relu=True, wi
models/gcn.py:54
↓ 1 callersMethod__init__
(self, nfeat, nhid, nclass, nlayers=2, dropout=0.5, lr=0.01, weight_decay=5e-4, with_relu=True, wi
models/sgc.py:57
↓ 1 callersMethod__init__
(self, nfeat, nhid, nclass, heads=8, output_heads=1, dropout=0.5, with_bias=True, device=None, **kwargs)
models/gat.py:9
↓ 1 callersMethod__init__
(self, nfeat, nhid, nclass, nlayers=2, dropout=0.5, lr=0.01, weight_decay=5e-4, with_relu=True, wi
models/mygraphsage.py:73
↓ 1 callersMethod__init__
nlayers indicates the number of propagations
models/sgc_multi.py:18
↓ 1 callersMethod__next
go through one normal layer, for all elements
gntk_cond.py:50
↓ 1 callersMethod__next_diag
go through one normal layer, for diagonal element S: covariance of last layer
gntk_cond.py:25
↓ 1 callersMethod_forward_with_param
(self, flat_param, *inputs, **kwinputs)
models/reparam_module.py:139
↓ 1 callersMethod_forward_with_param_and_buffers
(self, flat_param, buffers, *inputs, **kwinputs)
models/reparam_module.py:134
↓ 1 callersFunctiondistance_wb
(gwr, gws)
utils.py:270
↓ 1 callersMethodevaluate_synset_ntk
(self)
metagtt_inductive_adj_identity.py:88
↓ 1 callersMethodevaluate_synset_ntk
(self)
metagtt_transduct_adj_identity2.py:91
↓ 1 callersMethodevaluate_synset_ntk
(self)
metagtt_transduct_adj_identity.py:103
↓ 1 callersMethodexpert_load
(self)
metagtt_inductive_adj_identity.py:39
↓ 1 callersMethodexpert_load
(self)
metagtt_transduct_adj_identity2.py:42
↓ 1 callersMethodexpert_load
(self)
metagtt_transduct_adj_identity.py:49
↓ 1 callersFunctionfeature_smoothing
(adj, X)
utils.py:373
↓ 1 callersMethodforward
(self, x, inference=False)
models/parametrized_adj.py:42
↓ 1 callersMethodforward
(self, data)
models/gat.py:37
↓ 1 callersMethodforward_sampler
(self, x, adjs)
models/mygraphsage.py:128
↓ 1 callersMethodgenerate_labels_syn
(self, data)
coreset/all_methods.py:15
↓ 1 callersMethodget_coreset_init
(self, features, adj, labels)
metagtt_inductive_adj_identity.py:435
↓ 1 callersMethodget_coreset_init
(self, features, adj, labels)
metagtt_transduct_adj_identity2.py:430
↓ 1 callersMethodget_coreset_init
(self, features, adj, labels)
metagtt_transduct_adj_identity.py:549
↓ 1 callersFunctionget_syn_data
(args, data, device, model_type=None)
test_condg.py:218
↓ 1 callersMethodinitialize
Initialize parameters of GCN.
models/myappnp1.py:134
↓ 1 callersMethodinitialize
Initialize parameters of GCN.
models/sgc.py:127
↓ 1 callersMethodinitialize
Initialize parameters of GAT.
models/gat.py:46
↓ 1 callersMethodinitialize
Initialize parameters of GCN.
models/mygraphsage.py:161
↓ 1 callersMethodinitialize
Initialize parameters of GCN.
models/mycheby.py:175
↓ 1 callersMethodinitialize
Initialize parameters of GCN.
models/myappnp.py:125
↓ 1 callersMethodinitialize
Initialize parameters of GCN.
models/sgc_multi.py:112
↓ 1 callersMethodlen
(self)
models/in_memory_dataset.py:63
↓ 1 callersFunctionmain
(args)
buffer_inductive.py:21
↓ 1 callersFunctionmain
(args)
buffer_transduct.py:22
↓ 1 callersFunctionmain
(args)
test_condg.py:31
↓ 1 callersFunctionnormalize_adj
Normalize sparse adjacency matrix, A' = (D + I)^-1/2 * ( A + I ) * (D + I)^-1/2 Row-normalize sparse matrix Parameters ----------
models/mycheby.py:401
↓ 1 callersMethodpredict
By default, the inputs should be unnormalized adjacency Parameters ---------- features : node features. If `featur
models/myappnp1.py:256
next →1–100 of 260, ranked by callers