mlp_citation2_sign.py

import argparse
import os

os.system("pip install ogb==1.3.2")

#os.system("pip install ./torch-1.7.1+cu101-cp36-cp36m-linux_x86_64.whl")
#os.system("pip install ./torchvision-0.8.2+cu101-cp36-cp36m-linux_x86_64.whl")

#os.system("pip install ./torch_cluster-1.5.9-cp36-cp36m-linux_x86_64.whl")
#os.system("pip install ./torch_scatter-2.0.6-cp36-cp36m-linux_x86_64.whl")
#os.system("pip install ./torch_sparse-0.6.8-cp36-cp36m-linux_x86_64.whl")
#os.system("pip install ./torch_spline_conv-1.2.1-cp36-cp36m-linux_x86_64.whl")
#os.system("pip install torch-scatter torch-sparse torch-cluster torch-spline-conv torch-geometric -f https://data.pyg.org/whl/torch-1.7.1+cu101.html")
#os.system("pip install torch-geometric==2.0.1")

import torch
import torch.nn.functional as F
from torch.utils.data import DataLoader

from ogb.linkproppred import PygLinkPropPredDataset, Evaluator

from logger import Logger
from torch_geometric.transforms import SIGN


class LinkPredictor(torch.nn.Module):
    def __init__(self, in_channels, hidden_channels, out_channels, num_layers,
                 dropout):
        super(LinkPredictor, self).__init__()

        self.lins = torch.nn.ModuleList()
        self.lins.append(torch.nn.Linear(in_channels, hidden_channels))
        for _ in range(num_layers - 2):
            self.lins.append(torch.nn.Linear(hidden_channels, hidden_channels))
        self.lins.append(torch.nn.Linear(hidden_channels, out_channels))

        self.dropout = dropout

    def reset_parameters(self):
        for lin in self.lins:
            lin.reset_parameters()

    def forward(self, x_i, x_j):
        x = x_i * x_j
        for lin in self.lins[:-1]:
            x = lin(x)
            x = F.relu(x)
            x = F.dropout(x, p=self.dropout, training=self.training)
        x = self.lins[-1](x)
        return torch.sigmoid(x)


def train(predictor, x, split_edge, optimizer, batch_size):
    predictor.train()

    source_edge = split_edge['train']['source_node'].to(x.device)
    target_edge = split_edge['train']['target_node'].to(x.device)

    total_loss = total_examples = 0
    for perm in DataLoader(range(source_edge.size(0)), batch_size,
                           shuffle=True):
        optimizer.zero_grad()

        src, dst = source_edge[perm], target_edge[perm]

        pos_out = predictor(x[src], x[dst])
        pos_loss = -torch.log(pos_out + 1e-15).mean()

        # Just do some trivial random sampling.
        dst_neg = torch.randint(0, x.size(0), src.size(), dtype=torch.long,
                                device=x.device)
        neg_out = predictor(x[src], x[dst_neg])
        neg_loss = -torch.log(1 - neg_out + 1e-15).mean()

        loss = pos_loss + neg_loss
        loss.backward()
        optimizer.step()

        num_examples = pos_out.size(0)
        total_loss += loss.item() * num_examples
        total_examples += num_examples

    return total_loss / total_examples


@torch.no_grad()
def test(predictor, x, split_edge, evaluator, batch_size):
    predictor.eval()

    def test_split(split):
        source = split_edge[split]['source_node'].to(x.device)
        target = split_edge[split]['target_node'].to(x.device)
        target_neg = split_edge[split]['target_node_neg'].to(x.device)

        pos_preds = []
        for perm in DataLoader(range(source.size(0)), batch_size):
            src, dst = source[perm], target[perm]
            pos_preds += [predictor(x[src], x[dst]).squeeze().cpu()]
        pos_pred = torch.cat(pos_preds, dim=0)

        neg_preds = []
        source = source.view(-1, 1).repeat(1, 1000).view(-1)
        target_neg = target_neg.view(-1)
        for perm in DataLoader(range(source.size(0)), batch_size):
            src, dst_neg = source[perm], target_neg[perm]
            neg_preds += [predictor(x[src], x[dst_neg]).squeeze().cpu()]
        neg_pred = torch.cat(neg_preds, dim=0).view(-1, 1000)

        return evaluator.eval({
            'y_pred_pos': pos_pred,
            'y_pred_neg': neg_pred,
        })['mrr_list'].mean().item()

    train_mrr = test_split('eval_train')
    valid_mrr = test_split('valid')
    test_mrr = test_split('test')

    return train_mrr, valid_mrr, test_mrr


def main():
    parser = argparse.ArgumentParser(description='OGBL-Citation2 (MLP)')
    parser.add_argument('--device', type=int, default=0)
    parser.add_argument('--log_steps', type=int, default=1)
    parser.add_argument('--use_node_embedding', action='store_true')
    parser.add_argument('--num_layers', type=int, default=3)
    parser.add_argument('--hidden_channels', type=int, default=256)
    parser.add_argument('--dropout', type=float, default=0.0)
    parser.add_argument('--batch_size', type=int, default=64 * 1024)
    parser.add_argument('--lr', type=float, default=0.01)
    parser.add_argument('--epochs', type=int, default=100)
    parser.add_argument('--eval_steps', type=int, default=10)
    parser.add_argument('--runs', type=int, default=10)
    args = parser.parse_args()
    print(args)

    device = f'cuda:{args.device}' if torch.cuda.is_available() else 'cpu'
    device = torch.device(device)

    dataset = PygLinkPropPredDataset(name='ogbl-citation2') #,  root = "./dataset"
    split_edge = dataset.get_edge_split()
    data = dataset[0]

    # We randomly pick some training samples that we want to evaluate on:
    torch.manual_seed(12345)
    idx = torch.randperm(split_edge['train']['source_node'].numel())[:86596]
    split_edge['eval_train'] = {
        'source_node': split_edge['train']['source_node'][idx],
        'target_node': split_edge['train']['target_node'][idx],
        'target_node_neg': split_edge['valid']['target_node_neg'],
    }
    
    data = SIGN(args.num_layers)(dataset[0])  # This might take a while.
    xs = [data.x] + [data[f'x{i}'] for i in range(1, 3 + 1)]
    x = torch.cat(xs, dim=-1)
    
    #x = data.x
    if args.use_node_embedding:
        embedding = torch.load('embedding.pt', map_location='cpu')
        x = torch.cat([x, embedding], dim=-1)
    x = x.to(device)

    predictor = LinkPredictor(x.size(-1), args.hidden_channels, 1,
                              args.num_layers, args.dropout).to(device)

    evaluator = Evaluator(name='ogbl-citation2')
    logger = Logger(args.runs, args)

    for run in range(args.runs):
        predictor.reset_parameters()
        optimizer = torch.optim.Adam(predictor.parameters(), lr=args.lr)

        for epoch in range(1, 1 + args.epochs):
            loss = train(predictor, x, split_edge, optimizer,
                         args.batch_size)
            print(f'Run: {run + 1:02d}, Epoch: {epoch:02d}, Loss: {loss:.4f}')

            if epoch % args.eval_steps == 0:
                result = test(predictor, x, split_edge, evaluator,
                              args.batch_size)
                logger.add_result(run, result)

                if epoch % args.log_steps == 0:
                    train_mrr, valid_mrr, test_mrr = result
                    print(f'Run: {run + 1:02d}, '
                          f'Epoch: {epoch:02d}, '
                          f'Loss: {loss:.4f}, '
                          f'Train: {train_mrr:.4f}, '
                          f'Valid: {valid_mrr:.4f}, '
                          f'Test: {test_mrr:.4f}')

        print('Node2vec' if args.use_node_embedding else 'MLP')
        logger.print_statistics(run)
    print('Node2vec' if args.use_node_embedding else 'MLP')
    logger.print_statistics()


if __name__ == "__main__":
    main()