file_address = r"../Data/PB.gml" if data_set_name == 'bio-CE-GT': file_address = r"../Data/bio-CE-GT.gml" if data_set_name == 'hamster': file_address = r"../Data/hamster.gml" if data_set_name == 'USAir': file_address = r"../Data/USAir.gml" if data_set_name == 'Yeast': file_address = r"../Data/Yeast.gml" if torch.cuda.is_available(): GPU = True else: GPU = False G, A, nodes, all_neighbors, As = process_gml_file(file_address) train_loader, test_loader, A_test, weight = get_data_loader( A, radio=radio, batch_size=batchSize, sample_method=sample_method, GPU=GPU) if module_name == 'DeepWalk': module = Deep_Walk(G=G, A=A_test, walk_length=walk_length, embed_size=embed_size, window_size=window_size, workers=workers) if module_name == 'Node2Vec': module = Node2vec(G=G, A=A_test, walk_length=walk_length, p=p, q=q,
import torch from torch import nn from torch.nn import Parameter import numpy from torch.utils import data as Data from Tools import process_gml_file import torch.optim as optim G, A, nodes, all_neighbors, As = process_gml_file( r"C:\Users\mihao\Desktop\米昊的东西\input.gml") def getDataLoader(A, radio): data = [] label = [] N = len(A) for i in range(N): for j in range(N): if i != j: data.append([i, j]) label.append(A[i][j]) train_data = torch.tensor(data[:int(N * (N - 1) * radio)]) train_label = torch.tensor(label[:int(N * (N - 1) * radio)]) train_dataset = Data.TensorDataset(train_data, train_label) train_loader = Data.DataLoader( dataset=train_dataset, batch_size=16, shuffle=True, )