GCN代码

代码可以在github上fork,本文主要是加了一些注释，并且搭配本人所作笔记【GCN代码笔记】
layers.py

import mathimport torchfrom torch.nn.parameter import Parameter
from torch.nn.modules.module import Moduleclass GraphConvolution(Module):"""Simple GCN layer, similar to https://arxiv.org/abs/1609.02907"""def __init__(self, in_features, out_features, bias=True):super(GraphConvolution, self).__init__()self.in_features = in_featuresself.out_features = out_featuresself.weight = Parameter(torch.FloatTensor(in_features, out_features))if bias:self.bias = Parameter(torch.FloatTensor(out_features))else:self.register_parameter('bias', None)#Parameter()与register_parameter()都是将一个不可训练的类型Tensor转换成可以训练的类型parameter，并将这个parameter绑定到这个module里面，相当于变成了模型的一部分，成为了模型中可以根据训练进行变化的参数。self.reset_parameters()#构建好特征之后对特征进行随机初始化的过程def reset_parameters(self):stdv = 1. / math.sqrt(self.weight.size(1))self.weight.data.uniform_(-stdv, stdv)if self.bias is not None:self.bias.data.uniform_(-stdv, stdv)def forward(self, input, adj):support = torch.mm(input, self.weight)#将input函数和weight函数做了一个乘积；torch.mm(a, b)是矩阵a和b矩阵相乘output = torch.spmm(adj, support)#稀疏矩阵的相乘，是之前归一化之后的结果和上一步相乘if self.bias is not None:return output + self.biaselse:return outputdef __repr__(self):return self.__class__.__name__ + ' (' \+ str(self.in_features) + ' -> ' \+ str(self.out_features) + ')'

models.py

import torch.nn as nn
import torch.nn.functional as F
from pygcn.layers import GraphConvolutionclass GCN(nn.Module):def __init__(self, nfeat, nhid, nclass, dropout):super(GCN, self).__init__()#构建第一层GCN，第一个参数是初始特征，第二个参数是隐藏层特征self.gc1 = GraphConvolution(nfeat, nhid)#构建第二层GCN，传入的nhid维度是16维，输出的nclass与最后要判定的类别数是一致的，是7个类别self.gc2 = GraphConvolution(nhid, nclass)self.dropout = dropoutdef forward(self, x, adj):x = F.relu(self.gc1(x, adj))x = F.dropout(x, self.dropout, training=self.training)#完成第一层GCN训练之后，x.shape是2708*16维的。16是hidden layer的维度，16是传参传进去的x = self.gc2(x, adj)return F.log_softmax(x, dim=1)

train.py

from __future__ import division
from __future__ import print_functionimport time
import argparse
import numpy as npimport torch
import torch.nn.functional as F
import torch.optim as optimfrom pygcn.utils import load_data, accuracy
from pygcn.models import GCN# Training settings添加train set的一些参数
parser = argparse.ArgumentParser()
parser.add_argument('--no-cuda', action='store_true', default=False,help='Disables CUDA training.')
parser.add_argument('--fastmode', action='store_true', default=False,help='Validate during training pass.')
parser.add_argument('--seed', type=int, default=42, help='Random seed.')
parser.add_argument('--epochs', type=int, default=200,help='Number of epochs to train.')
parser.add_argument('--lr', type=float, default=0.01,help='Initial learning rate.')
parser.add_argument('--weight_decay', type=float, default=5e-4,help='Weight decay (L2 loss on parameters).')
parser.add_argument('--hidden', type=int, default=16,help='Number of hidden units.')
parser.add_argument('--dropout', type=float, default=0.5,help='Dropout rate (1 - keep probability).')args = parser.parse_args()
args.cuda = not args.no_cuda and torch.cuda.is_available()np.random.seed(args.seed)
torch.manual_seed(args.seed)
if args.cuda:torch.cuda.manual_seed(args.seed)# Load data
adj, features, labels, idx_train, idx_val, idx_test = load_data()# Model and optimizer 构建GCN模型，初始化参数。两层GCN
model = GCN(nfeat=features.shape[1],nhid=args.hidden,nclass=labels.max().item() + 1,dropout=args.dropout)
optimizer = optim.Adam(model.parameters(),lr=args.lr, weight_decay=args.weight_decay)if args.cuda:model.cuda()features = features.cuda()adj = adj.cuda()labels = labels.cuda()idx_train = idx_train.cuda()idx_val = idx_val.cuda()idx_test = idx_test.cuda()#定义训练的一个函数
def train(epoch):t = time.time()# 如果模型中有BN层(Batch Normalization）和Dropout，需要在训练时添加model.train()。# model.train()是保证BN层能够用到每一批数据的均值和方差。对于Dropout，model.train()是随机取一部分网络连接来训练更新参数。model.train()optimizer.zero_grad()#把模型中参数的梯度设为0output = model(features, adj)#运行模型，输入参数（features,adj）即特征和邻接矩阵#output输出结果是2708*7的维度，值代表每一位可能输出的概率loss_train = F.nll_loss(output[idx_train], labels[idx_train])#等同于cross_entropyacc_train = accuracy(output[idx_train], labels[idx_train])loss_train.backward()optimizer.step()if not args.fastmode:# Evaluate validation set performance separately,# deactivates dropout during validation run.model.eval()output = model(features, adj)loss_val = F.nll_loss(output[idx_val], labels[idx_val])acc_val = accuracy(output[idx_val], labels[idx_val])print('Epoch: {:04d}'.format(epoch+1),'loss_train: {:.4f}'.format(loss_train.item()),'acc_train: {:.4f}'.format(acc_train.item()),'loss_val: {:.4f}'.format(loss_val.item()),'acc_val: {:.4f}'.format(acc_val.item()),'time: {:.4f}s'.format(time.time() - t))#定义测试的一个函数
def test():# 如果模型中有BN层(BatchNormalization）和Dropout，在测试时添加model.eval()。# model.eval()是保证BN层能够用全部训练数据的均值和方差，即测试过程中要保证BN层的均值和方差不变。# 对于Dropout，model.eval()是利用到了所有网络连接，即不进行随机舍弃神经元。# 训练完train样本后，生成的模型model要用来测试样本。# 在model(test)之前，需要加上model.eval()，否则的话，有输入数据，即使不训练，它也会改变权值。这是model中含有BN层和Dropout所带来的的性质。model.eval()output = model(features, adj)loss_test = F.nll_loss(output[idx_test], labels[idx_test])acc_test = accuracy(output[idx_test], labels[idx_test])print("Test set results:","loss= {:.4f}".format(loss_test.item()),"accuracy= {:.4f}".format(acc_test.item()))# Train model
t_total = time.time()
#定义了训练多少回合
for epoch in range(args.epochs):train(epoch)
print("Optimization Finished!")
print("Total time elapsed: {:.4f}s".format(time.time() - t_total))# Testing
test()

utils.py

import numpy as np
import scipy.sparse as sp
import torchdef encode_onehot(labels):classes = set(labels)classes_dict = {c: np.identity(len(classes))[i, :] for i, c inenumerate(classes)}labels_onehot = np.array(list(map(classes_dict.get, labels)),dtype=np.int32)return labels_onehotdef load_data(path="../data/cora/", dataset="cora"):"""Load citation network dataset (cora only for now)"""print('Loading {} dataset...'.format(dataset))#得到cora.content中的数据idx_features_labels = np.genfromtxt("{}{}.content".format(path, dataset),dtype=np.dtype(str))#取特征，从第一个到倒数第二个#sp.csr_matrix就是按照列来压缩features = sp.csr_matrix(idx_features_labels[:, 1:-1], dtype=np.float32)#labels即最后一列 文章分类labels = encode_onehot(idx_features_labels[:, -1])# build graph#取索引，取的是cora.content中的第0列的数据idx = np.array(idx_features_labels[:, 0], dtype=np.int32)#构建节点的索引字典，就是转换为从0到索引长度-1的编号，例如将第一个元素31336转换为0，{31336：0，等等}idx_map = {j: i for i, j in enumerate(idx)}#导入边的一个数据 cora.cites文件中是边之间的联系edges_unordered = np.genfromtxt("{}{}.cites".format(path, dataset),dtype=np.int32)#将之前的转换成字典编号后的边，即假设23：9，23与31336之间有联系，现在转化为9与0之间有联系。edges = np.array(list(map(idx_map.get, edges_unordered.flatten())),dtype=np.int32).reshape(edges_unordered.shape)#构建边的邻接矩阵（sp.coo_matrix构建的是稀疏矩阵）adj = sp.coo_matrix((np.ones(edges.shape[0]), (edges[:, 0], edges[:, 1])),shape=(labels.shape[0], labels.shape[0]),dtype=np.float32)   #2708*2708 因为有2708个节点# build symmetric adjacency matrix，计算转置矩阵，将有向图转化为无向图adj = adj + adj.T.multiply(adj.T > adj) - adj.multiply(adj.T > adj)features = normalize(features) #对特征做了归一化的操作，但不是一个必要的操作adj = normalize(adj + sp.eye(adj.shape[0])) #对（邻接矩阵+单位矩阵）做归一化的操作，这是必要的一个操作，这一步完成之后就是完成公式中的D的负1次*A#划分训练，验证，测试的样本idx_train = range(140)idx_val = range(200, 500)idx_test = range(500, 1500)#将numpy的数据转换成torch格式features = torch.FloatTensor(np.array(features.todense()))labels = torch.LongTensor(np.where(labels)[1])adj = sparse_mx_to_torch_sparse_tensor(adj)idx_train = torch.LongTensor(idx_train)idx_val = torch.LongTensor(idx_val)idx_test = torch.LongTensor(idx_test)
#返回经过归一化后的邻接矩阵，特征，标签，训练集、验证集、测试集所对应到的索引return adj, features, labels, idx_train, idx_val, idx_test#对节点做归一化
def normalize(mx):"""Row-normalize sparse matrix"""rowsum = np.array(mx.sum(1))#矩阵按行求和r_inv = np.power(rowsum, -1).flatten() #求倒数，有的可能是0 然后求倒数之后出现无穷大，，此时是2708维r_inv[np.isinf(r_inv)] = 0. #如果是inf（无穷大），转换成0r_mat_inv = sp.diags(r_inv) #构建对角矩阵 ，转换成2708*2708维的矩阵#用转换之后的矩阵乘以之前的特征，即将特征按行做了归一化的一个操作。mx = r_mat_inv.dot(mx)return mxdef accuracy(output, labels):preds = output.max(1)[1].type_as(labels)#求output矩阵每一行的最大值，最大值即要输出的labelcorrect = preds.eq(labels).double()correct = correct.sum()return correct / len(labels)def sparse_mx_to_torch_sparse_tensor(sparse_mx):"""Convert a scipy sparse matrix to a torch sparse tensor."""sparse_mx = sparse_mx.tocoo().astype(np.float32)indices = torch.from_numpy(np.vstack((sparse_mx.row, sparse_mx.col)).astype(np.int64))values = torch.from_numpy(sparse_mx.data)shape = torch.Size(sparse_mx.shape)return torch.sparse.FloatTensor(indices, values, shape)

本文来自互联网用户投稿，文章观点仅代表作者本人，不代表本站立场，不承担相关法律责任。如若转载，请注明出处。 如若内容造成侵权/违法违规/事实不符，请点击【内容举报】进行投诉反馈！