├── Graph ├── config │ ├── hiv.json │ ├── pcba.json │ ├── pcqm.json │ ├── pcqms.json │ └── zinc.json ├── dgl_main.py ├── dgldataclass.py ├── get_dataset.py ├── large_model.py ├── master.csv ├── medium_model.py ├── small_model.py └── zinc_model.py ├── LICENSE ├── Node ├── config.yaml ├── data │ └── chameleon.pt.zip ├── main_node.py ├── master.csv ├── model_node.py ├── node_raw_data │ ├── 2Dgrid.mat │ ├── Penn94.mat │ ├── actor │ │ ├── out1_graph_edges.txt │ │ └── out1_node_feature_label.txt │ ├── amazon_electronics_photo.npz │ ├── chameleon │ │ ├── out1_graph_edges.txt │ │ └── out1_node_feature_label.txt │ ├── citeseer │ │ ├── ind.citeseer.allx │ │ ├── ind.citeseer.ally │ │ ├── ind.citeseer.graph │ │ ├── ind.citeseer.test.index │ │ ├── ind.citeseer.tx │ │ ├── ind.citeseer.ty │ │ ├── ind.citeseer.x │ │ └── ind.citeseer.y │ ├── cora │ │ ├── ind.cora.allx │ │ ├── ind.cora.ally │ │ ├── ind.cora.graph │ │ ├── ind.cora.test.index │ │ ├── ind.cora.tx │ │ ├── ind.cora.ty │ │ ├── ind.cora.x │ │ └── ind.cora.y │ ├── fb100-Penn94-splits.npy │ └── squirrel │ │ ├── out1_graph_edges.txt │ │ └── out1_node_feature_label.txt ├── preprocess_node_data.py └── utils.py ├── README.md └── requirements.txt /Graph/config/hiv.json: -------------------------------------------------------------------------------- 1 | { 2 | "nlayer": 8, 3 | "nheads": 4, 4 | "hidden_dim": 80, 5 | "trans_dropout": 0.1, 6 | "feat_dropout": 0.1, 7 | "adj_dropout": 0.3, 8 | "lr": 1e-4, 9 | "weight_decay": 1e-4, 10 | "epochs": 50, 11 | "warm_up_epoch": 5, 12 | "batch_size": 64 13 | } 14 | -------------------------------------------------------------------------------- /Graph/config/pcba.json: -------------------------------------------------------------------------------- 1 | { 2 | "nlayer": 8, 3 | "nheads": 8, 4 | "hidden_dim": 272, 5 | "trans_dropout": 0.3, 6 | "feat_dropout": 0.1, 7 | "adj_dropout": 0.1, 8 | "lr": 5e-4, 9 | "weight_decay": 5e-3, 10 | "epochs": 30, 11 | "warm_up_epoch": 5, 12 | "batch_size": 64 13 | } 14 | -------------------------------------------------------------------------------- /Graph/config/pcqm.json: -------------------------------------------------------------------------------- 1 | { 2 | "nlayer": 10, 3 | "nheads": 16, 4 | "hidden_dim": 400, 5 | "trans_dropout": 0.05, 6 | "feat_dropout": 0.05, 7 | "adj_dropout": 0.05, 8 | "lr": 2e-4, 9 | "weight_decay": 0.0, 10 | "epochs": 150, 11 | "warm_up_epoch": 10, 12 | "batch_size": 64 13 | } 14 | -------------------------------------------------------------------------------- /Graph/config/pcqms.json: -------------------------------------------------------------------------------- 1 | { 2 | "nlayer": 6, 3 | "nheads": 8, 4 | "hidden_dim": 240, 5 | "filter_poly": 12, 6 | "trans_dropout": 0.3, 7 | "feat_dropout": 0.1, 8 | "adj_dropout": 0.1, 9 | "lr": 5e-4, 10 | "weight_decay": 5e-4, 11 | "epochs": 100, 12 | "warm_up_epoch": 5, 13 | "batch_size": 256 14 | } 15 | -------------------------------------------------------------------------------- /Graph/config/zinc.json: -------------------------------------------------------------------------------- 1 | { 2 | "nlayer": 4, 3 | "nheads": 8, 4 | "hidden_dim": 160, 5 | "trans_dropout": 0.1, 6 | "feat_dropout": 0.05, 7 | "adj_dropout": 0.0, 8 | "lr": 1e-3, 9 | "weight_decay": 5e-4, 10 | "epochs": 1000, 11 | "warm_up_epoch": 50, 12 | "batch_size": 32 13 | } 14 | -------------------------------------------------------------------------------- /Graph/dgl_main.py: -------------------------------------------------------------------------------- 1 | import time 2 | import math 3 | import copy 4 | import wandb 5 | import argparse 6 | import datetime 7 | import random, os 8 | import numpy as np 9 | import torch 10 | import torch.nn as nn 11 | import torch.nn.functional as F 12 | from torch.utils.data import DataLoader, Dataset, TensorDataset 13 | from torch.optim.lr_scheduler import LambdaLR 14 | import json5 15 | from easydict import EasyDict 16 | 17 | from ema_pytorch import EMA 18 | from zinc_model import SpecformerZINC 19 | from large_model import SpecformerLarge 20 | from medium_model import SpecformerMedium 21 | from small_model import SpecformerSmall 22 | from get_dataset import DynamicBatchSampler, RandomSampler, collate_pad, collate_dgl, get_dataset 23 | 24 | 25 | def init_params(module): 26 | if isinstance(module, nn.Linear): 27 | module.weight.data.normal_(mean=0.0, std=0.02) 28 | if module.bias is not None: 29 | module.bias.data.zero_() 30 | if isinstance(module, nn.Embedding): 31 | module.weight.data.normal_(mean=0.0, std=0.02) 32 | 33 | 34 | def get_config_from_json(json_file): 35 | with open('config/' + json_file + '.json', 'r') as config_file: 36 | config_dict = json5.load(config_file) 37 | config = EasyDict(config_dict) 38 | 39 | return config 40 | 41 | 42 | def seed_everything(seed): 43 | random.seed(seed) 44 | os.environ['PYTHONHASHSEED'] = str(seed) 45 | np.random.seed(seed) 46 | torch.manual_seed(seed) 47 | torch.cuda.manual_seed(seed) 48 | torch.backends.cudnn.deterministic = True 49 | torch.backends.cudnn.benchmark = True 50 | torch.backends.cudnn.allow_tf32 = False 51 | 52 | 53 | def count_parameters(model): 54 | return sum(p.numel() for p in model.parameters() if p.requires_grad) 55 | 56 | 57 | def train_epoch(dataset, model, device, dataloader, loss_fn, optimizer, wandb=None, wandb_item=None): 58 | model.train() 59 | 60 | for i, data in enumerate(dataloader): 61 | e, u, g, length, y = data 62 | e, u, g, length, y = e.to(device), u.to(device), g.to(device), length.to(device), y.to(device) 63 | 64 | logits = model(e, u, g, length) 65 | optimizer.zero_grad() 66 | 67 | y_idx = y == y 68 | loss = loss_fn(logits.to(torch.float32)[y_idx], y.to(torch.float32)[y_idx]) 69 | 70 | loss.backward() 71 | optimizer.step() 72 | 73 | if wandb: 74 | wandb.log({wandb_item: loss.item()}) 75 | 76 | 77 | def eval_epoch(dataset, model, device, dataloader, evaluator, metric): 78 | model.eval() 79 | 80 | y_true = [] 81 | y_pred = [] 82 | 83 | with torch.no_grad(): 84 | for i, data in enumerate(dataloader): 85 | e, u, g, length, y = data 86 | e, u, g, length, y = e.to(device), u.to(device), g.to(device), length.to(device), y.to(device) 87 | 88 | logits = model(e, u, g, length) 89 | 90 | y_true.append(y.view(logits.shape).detach().cpu()) 91 | y_pred.append(logits.detach().cpu()) 92 | 93 | y_true = torch.cat(y_true, dim=0).numpy() 94 | y_pred = torch.cat(y_pred, dim=0).numpy() 95 | 96 | return evaluator.eval({'y_true': y_true, 'y_pred': y_pred})[metric] 97 | 98 | 99 | def main_worker(args): 100 | seed_everything(args.seed) 101 | rank = 'cuda:{}'.format(args.cuda) 102 | print(args) 103 | 104 | datainfo = get_dataset(args.dataset) 105 | nclass = datainfo['num_class'] 106 | loss_fn = datainfo['loss_fn'] 107 | evaluator = datainfo['evaluator'] 108 | train = datainfo['train_dataset'] 109 | valid = datainfo['valid_dataset'] 110 | test = datainfo['test_dataset'] 111 | metric = datainfo['metric'] 112 | metric_mode = datainfo['metric_mode'] 113 | 114 | # dataloader 115 | ''' 116 | train_batch_sampler = DynamicBatchSampler(RandomSampler(train), [data.num_nodes for data in train], 117 | batch_size=32, max_nodes=50, drop_last=False) 118 | valid_batch_sampler = DynamicBatchSampler(RandomSampler(valid), [data.num_nodes for data in valid], 119 | batch_size=32, max_nodes=50, drop_last=False) 120 | test_batch_sampler = DynamicBatchSampler(RandomSampler(test), [data.num_nodes for data in test], 121 | batch_size=32, max_nodes=50, drop_last=False) 122 | train_dataloader = DataLoader(train, batch_sampler=train_batch_sampler, collate_fn=collate_pad) 123 | valid_dataloader = DataLoader(valid, batch_sampler=valid_batch_sampler, collate_fn=collate_pad) 124 | test_dataloader = DataLoader(test, batch_sampler=test_batch_sampler, collate_fn=collate_pad) 125 | ''' 126 | 127 | train_dataloader = DataLoader(train, batch_size = args.batch_size, num_workers=4, collate_fn=collate_dgl, shuffle = True) 128 | valid_dataloader = DataLoader(valid, batch_size = args.batch_size // 2, num_workers=4, collate_fn=collate_dgl, shuffle = False) 129 | test_dataloader = DataLoader(test, batch_size = args.batch_size // 2, num_workers=4, collate_fn=collate_dgl, shuffle = False) 130 | 131 | if args.dataset == 'zinc': 132 | print('zinc') 133 | model = SpecformerZINC(nclass, args.nlayer, args.hidden_dim, args.nheads, 134 | args.feat_dropout, args.trans_dropout, args.adj_dropout).to(rank) 135 | 136 | elif args.dataset == 'pcqm' or args.dataset == 'pcqms': 137 | print('pcqm') 138 | model = SpecformerLarge(nclass, args.nlayer, args.hidden_dim, args.nheads, 139 | args.feat_dropout, args.trans_dropout, args.adj_dropout).to(rank) 140 | print('init') 141 | model.apply(init_params) 142 | 143 | elif args.dataset == 'pcba': 144 | print('pcba') 145 | model = SpecformerMedium(nclass, args.nlayer, args.hidden_dim, args.nheads, 146 | args.feat_dropout, args.trans_dropout, args.adj_dropout).to(rank) 147 | model.apply(init_params) 148 | 149 | else: 150 | print('hiv') 151 | model = SpecformerSmall(nclass, args.nlayer, args.hidden_dim, args.nheads, 152 | args.feat_dropout, args.trans_dropout, args.adj_dropout).to(rank) 153 | 154 | print(count_parameters(model)) 155 | 156 | optimizer = torch.optim.AdamW(model.parameters(), lr=args.lr, betas=(0.9, 0.999), eps=1e-8, weight_decay=args.weight_decay) 157 | # warm_up + cosine weight decay 158 | lr_plan = lambda cur_epoch: (cur_epoch+1) / args.warm_up_epoch if cur_epoch < args.warm_up_epoch else \ 159 | (0.5 * (1.0 + math.cos(math.pi * (cur_epoch - args.warm_up_epoch) / (args.epochs - args.warm_up_epoch)))) 160 | scheduler = LambdaLR(optimizer, lr_lambda=lr_plan) 161 | 162 | results = [] 163 | for epoch in range(args.epochs): 164 | 165 | train_epoch(args.dataset, model, rank, train_dataloader, loss_fn, optimizer, wandb=None, wandb_item='loss') 166 | scheduler.step() 167 | 168 | torch.save(model.state_dict(), 'checkpoint/{}_{}.pth'.format(args.project_name, epoch)) 169 | 170 | if epoch % 1 == 0: 171 | 172 | val_res = eval_epoch(args.dataset, model, rank, valid_dataloader, evaluator, metric) 173 | test_res = eval_epoch(args.dataset, model, rank, test_dataloader, evaluator, metric) 174 | 175 | results.append([val_res, test_res]) 176 | 177 | if metric_mode == 'min': 178 | best_res = sorted(results, key = lambda x: x[0], reverse=False)[0][1] 179 | else: 180 | best_res = sorted(results, key = lambda x: x[0], reverse=True)[0][1] 181 | 182 | print(epoch, 'valid: {:.4f}'.format(val_res), 'test: {:.4f}'.format(test_res), 'best: {:.4f}'.format(best_res)) 183 | 184 | # wandb.log({'val': val_res, 'test': test_res}) 185 | 186 | torch.save(model.state_dict(), 'checkpoint/{}.pth'.format(args.project_name)) 187 | 188 | 189 | if __name__ == '__main__': 190 | parser = argparse.ArgumentParser() 191 | parser.add_argument('--seed', type=int, default=0) 192 | parser.add_argument('--cuda', type=int, default=0) 193 | parser.add_argument('--dataset', default='zinc') 194 | 195 | args = parser.parse_args() 196 | args.project_name = datetime.datetime.now().strftime('%m-%d-%X') 197 | 198 | config = get_config_from_json(args.dataset) 199 | 200 | for key in config.keys(): 201 | setattr(args, key, config[key]) 202 | 203 | main_worker(args) 204 | 205 | -------------------------------------------------------------------------------- /Graph/dgldataclass.py: -------------------------------------------------------------------------------- 1 | import pickle 2 | import pandas as pd 3 | import shutil, os 4 | import os.path as osp 5 | import torch 6 | import numpy as np 7 | import dgl 8 | from dgl.data.utils import load_graphs, save_graphs, Subset 9 | from ogb.utils import smiles2graph 10 | from ogb.utils.torch_util import replace_numpy_with_torchtensor 11 | from ogb.utils.url import decide_download, download_url, extract_zip 12 | from ogb.io.read_graph_raw import read_csv_graph_raw, read_csv_heterograph_raw, read_binary_graph_raw, read_binary_heterograph_raw 13 | from tqdm import tqdm 14 | from torch_geometric.utils import to_dense_adj, remove_isolated_nodes 15 | from torch_geometric.data import InMemoryDataset 16 | 17 | 18 | def read_graph_dgl(raw_dir, add_inverse_edge = False, additional_node_files = [], additional_edge_files = [], binary=False): 19 | 20 | if binary: 21 | # npz 22 | graph_list = read_binary_graph_raw(raw_dir, add_inverse_edge) 23 | else: 24 | # csv 25 | graph_list = read_csv_graph_raw(raw_dir, add_inverse_edge, additional_node_files = additional_node_files, additional_edge_files = additional_edge_files) 26 | 27 | dgl_graph_list = [] 28 | 29 | print('Converting graphs into DGL objects...') 30 | 31 | for graph in tqdm(graph_list): 32 | 33 | src, dst = torch.from_numpy(graph['edge_index']) 34 | num_nodes = graph['num_nodes'] 35 | 36 | if num_nodes == 1: # some graphs have one node 37 | A_ = torch.tensor(1.).view(1, 1) 38 | else: 39 | A = torch.zeros([num_nodes, num_nodes], dtype=torch.float) 40 | A[src, dst] = 1.0 41 | for i in range(num_nodes): 42 | A[i, i] = 1.0 43 | deg = torch.sum(A, axis=0).squeeze() ** -0.5 44 | D = torch.diag(deg) 45 | A_ = D @ A @ D 46 | e, u = torch.linalg.eigh(A_) 47 | 48 | fully_connected = torch.ones([num_nodes, num_nodes], dtype=torch.float).nonzero(as_tuple=True) 49 | g = dgl.graph(fully_connected, num_nodes = num_nodes) 50 | 51 | g.ndata['e'] = e 52 | g.ndata['u'] = u 53 | 54 | if graph['node_feat'] is not None: 55 | g.ndata['feat'] = torch.from_numpy(graph['node_feat']) 56 | 57 | if graph['edge_feat'] is not None: 58 | edge_idx = torch.stack([src, dst], dim=0) 59 | edge_attr = torch.from_numpy(graph['edge_feat']) + 1 # for padding 60 | 61 | if len(edge_attr.shape) == 1: 62 | edge_attr_dense = to_dense_adj(edge_idx, edge_attr=edge_attr.unsqueeze(-1)).squeeze(0).squeeze(-1).view(-1) 63 | else: 64 | if edge_attr.size(0) == 0: 65 | edge_attr_dense = torch.zeros([num_nodes ** 2, edge_attr.size(1)]).long() # for graphs without edge 66 | else: 67 | edge_attr_dense = to_dense_adj(edge_idx, edge_attr=edge_attr, max_num_nodes=num_nodes).squeeze(0).view(-1, edge_attr.shape[-1]) 68 | 69 | g.edata['feat'] = edge_attr_dense 70 | 71 | dgl_graph_list.append(g) 72 | 73 | return dgl_graph_list 74 | 75 | 76 | class DglGraphPropPredDataset(object): 77 | '''Adapted from https://docs.dgl.ai/en/latest/_modules/dgl/data/chem/csv_dataset.html#CSVDataset''' 78 | def __init__(self, name, root = 'dataset', meta_dict = None): 79 | ''' 80 | - name (str): name of the dataset 81 | - root (str): root directory to store the dataset folder 82 | - meta_dict: dictionary that stores all the meta-information about data. Default is None, 83 | but when something is passed, it uses its information. Useful for debugging for external contributers. 84 | ''' 85 | 86 | self.name = name ## original name, e.g., ogbg-molhiv 87 | 88 | if meta_dict is None: 89 | self.dir_name = '_'.join(name.split('-')) 90 | 91 | # check if previously-downloaded folder exists. 92 | # If so, use that one. 93 | if osp.exists(osp.join(root, self.dir_name + '_dgl')): 94 | self.dir_name = self.dir_name + '_dgl' 95 | 96 | self.original_root = root 97 | self.root = osp.join(root, self.dir_name) 98 | 99 | master = pd.read_csv(os.path.join(os.path.dirname(__file__), 'master.csv'), index_col = 0) 100 | if not self.name in master: 101 | error_mssg = 'Invalid dataset name {}.\n'.format(self.name) 102 | error_mssg += 'Available datasets are as follows:\n' 103 | error_mssg += '\n'.join(master.keys()) 104 | raise ValueError(error_mssg) 105 | self.meta_info = master[self.name] 106 | 107 | else: 108 | self.dir_name = meta_dict['dir_path'] 109 | self.original_root = '' 110 | self.root = meta_dict['dir_path'] 111 | self.meta_info = meta_dict 112 | 113 | # check version 114 | # First check whether the dataset has been already downloaded or not. 115 | # If so, check whether the dataset version is the newest or not. 116 | # If the dataset is not the newest version, notify this to the user. 117 | if osp.isdir(self.root) and (not osp.exists(osp.join(self.root, 'RELEASE_v' + str(self.meta_info['version']) + '.txt'))): 118 | print(self.name + ' has been updated.') 119 | if input('Will you update the dataset now? (y/N)\n').lower() == 'y': 120 | shutil.rmtree(self.root) 121 | 122 | self.download_name = self.meta_info['download_name'] ## name of downloaded file, e.g., tox21 123 | 124 | self.num_tasks = int(self.meta_info['num tasks']) 125 | self.eval_metric = self.meta_info['eval metric'] 126 | self.task_type = self.meta_info['task type'] 127 | self.num_classes = self.meta_info['num classes'] 128 | self.binary = self.meta_info['binary'] == 'True' 129 | 130 | super(DglGraphPropPredDataset, self).__init__() 131 | 132 | self.pre_process() 133 | 134 | def pre_process(self): 135 | processed_dir = osp.join(self.root, 'processed') 136 | raw_dir = osp.join(self.root, 'raw') 137 | pre_processed_file_path = osp.join(processed_dir, 'dgl_data_processed') 138 | 139 | if self.task_type == 'subtoken prediction': 140 | target_sequence_file_path = osp.join(processed_dir, 'target_sequence') 141 | 142 | if os.path.exists(pre_processed_file_path): 143 | 144 | if self.task_type == 'subtoken prediction': 145 | self.graphs, _ = load_graphs(pre_processed_file_path) 146 | self.labels = torch.load(target_sequence_file_path) 147 | 148 | else: 149 | self.graphs, label_dict = load_graphs(pre_processed_file_path) 150 | self.labels = label_dict['labels'] 151 | 152 | else: 153 | ### check download 154 | if self.binary: 155 | # npz format 156 | has_necessary_file = osp.exists(osp.join(self.root, 'raw', 'data.npz')) 157 | else: 158 | # csv file 159 | has_necessary_file = osp.exists(osp.join(self.root, 'raw', 'edge.csv.gz')) 160 | 161 | ### download 162 | if not has_necessary_file: 163 | url = self.meta_info['url'] 164 | if decide_download(url): 165 | path = download_url(url, self.original_root) 166 | extract_zip(path, self.original_root) 167 | os.unlink(path) 168 | # delete folder if there exists 169 | try: 170 | shutil.rmtree(self.root) 171 | except: 172 | pass 173 | shutil.move(osp.join(self.original_root, self.download_name), self.root) 174 | else: 175 | print('Stop download.') 176 | exit(-1) 177 | 178 | ### preprocess 179 | add_inverse_edge = self.meta_info['add_inverse_edge'] == 'True' 180 | 181 | if self.meta_info['additional node files'] == 'None': 182 | additional_node_files = [] 183 | else: 184 | additional_node_files = self.meta_info['additional node files'].split(',') 185 | 186 | if self.meta_info['additional edge files'] == 'None': 187 | additional_edge_files = [] 188 | else: 189 | additional_edge_files = self.meta_info['additional edge files'].split(',') 190 | 191 | graphs = read_graph_dgl(raw_dir, add_inverse_edge = add_inverse_edge, additional_node_files = additional_node_files, additional_edge_files = additional_edge_files, binary=self.binary) 192 | 193 | if self.task_type == 'subtoken prediction': 194 | # the downloaded labels are initially joined by ' ' 195 | labels_joined = pd.read_csv(osp.join(raw_dir, 'graph-label.csv.gz'), compression='gzip', header = None).values 196 | # need to split each element into subtokens 197 | labels = [str(labels_joined[i][0]).split(' ') for i in range(len(labels_joined))] 198 | 199 | print('Saving...') 200 | save_graphs(pre_processed_file_path, graphs) 201 | torch.save(labels, target_sequence_file_path) 202 | 203 | ### load preprocessed files 204 | self.graphs, _ = load_graphs(pre_processed_file_path) 205 | self.labels = torch.load(target_sequence_file_path) 206 | 207 | else: 208 | if self.binary: 209 | labels = np.load(osp.join(raw_dir, 'graph-label.npz'))['graph_label'] 210 | else: 211 | labels = pd.read_csv(osp.join(raw_dir, 'graph-label.csv.gz'), compression='gzip', header = None).values 212 | 213 | has_nan = np.isnan(labels).any() 214 | 215 | if 'classification' in self.task_type: 216 | if has_nan: 217 | labels = torch.from_numpy(labels).to(torch.float32) 218 | else: 219 | labels = torch.from_numpy(labels).to(torch.long) 220 | else: 221 | labels = torch.from_numpy(labels).to(torch.float32) 222 | 223 | 224 | print('Saving...') 225 | save_graphs(pre_processed_file_path, graphs, labels={'labels': labels}) 226 | 227 | ### load preprocessed files 228 | self.graphs, label_dict = load_graphs(pre_processed_file_path) 229 | self.labels = label_dict['labels'] 230 | 231 | 232 | def get_idx_split(self, split_type = None): 233 | if split_type is None: 234 | split_type = self.meta_info['split'] 235 | 236 | path = osp.join(self.root, 'split', split_type) 237 | 238 | # short-cut if split_dict.pt exists 239 | if os.path.isfile(os.path.join(path, 'split_dict.pt')): 240 | return torch.load(os.path.join(path, 'split_dict.pt')) 241 | 242 | train_idx = pd.read_csv(osp.join(path, 'train.csv.gz'), compression='gzip', header = None).values.T[0] 243 | valid_idx = pd.read_csv(osp.join(path, 'valid.csv.gz'), compression='gzip', header = None).values.T[0] 244 | test_idx = pd.read_csv(osp.join(path, 'test.csv.gz'), compression='gzip', header = None).values.T[0] 245 | 246 | return {'train': torch.tensor(train_idx, dtype = torch.long), 'valid': torch.tensor(valid_idx, dtype = torch.long), 'test': torch.tensor(test_idx, dtype = torch.long)} 247 | 248 | def __getitem__(self, idx): 249 | '''Get datapoint with index''' 250 | 251 | if isinstance(idx, int): 252 | return self.graphs[idx], self.labels[idx] 253 | elif torch.is_tensor(idx) and idx.dtype == torch.long: 254 | if idx.dim() == 0: 255 | return self.graphs[idx], self.labels[idx] 256 | elif idx.dim() == 1: 257 | return Subset(self, idx.cpu()) 258 | 259 | raise IndexError( 260 | 'Only integers and long are valid ' 261 | 'indices (got {}).'.format(type(idx).__name__)) 262 | 263 | def __len__(self): 264 | '''Length of the dataset 265 | Returns 266 | ------- 267 | int 268 | Length of Dataset 269 | ''' 270 | return len(self.graphs) 271 | 272 | def __repr__(self): # pragma: no cover 273 | return '{}({})'.format(self.__class__.__name__, len(self)) 274 | 275 | 276 | class DglPCQM4Mv2Dataset(object): 277 | def __init__(self, root = 'dataset', smiles2graph = smiles2graph): 278 | ''' 279 | DGL PCQM4Mv2 dataset object 280 | - root (str): the dataset folder will be located at root/pcqm4m_kddcup2021 281 | - smiles2graph (callable): A callable function that converts a SMILES string into a graph object 282 | * The default smiles2graph requires rdkit to be installed 283 | ''' 284 | 285 | self.original_root = root 286 | self.smiles2graph = smiles2graph 287 | self.folder = osp.join(root, 'pcqm4m-v2') 288 | self.version = 1 289 | 290 | # Old url hosted at Stanford 291 | # md5sum: 65b742bafca5670be4497499db7d361b 292 | # self.url = f'http://ogb-data.stanford.edu/data/lsc/pcqm4m-v2.zip' 293 | # New url hosted by DGL team at AWS--much faster to download 294 | self.url = 'https://dgl-data.s3-accelerate.amazonaws.com/dataset/OGB-LSC/pcqm4m-v2.zip' 295 | 296 | # check version and update if necessary 297 | if osp.isdir(self.folder) and (not osp.exists(osp.join(self.folder, f'RELEASE_v{self.version}.txt'))): 298 | print('PCQM4Mv2 dataset has been updated.') 299 | if input('Will you update the dataset now? (y/N)\n').lower() == 'y': 300 | shutil.rmtree(self.folder) 301 | 302 | super(DglPCQM4Mv2Dataset, self).__init__() 303 | 304 | # Prepare everything. 305 | # download if there is no raw file 306 | # preprocess if there is no processed file 307 | # load data if processed file is found. 308 | self.prepare_graph() 309 | 310 | def download(self): 311 | if decide_download(self.url): 312 | path = download_url(self.url, self.original_root) 313 | extract_zip(path, self.original_root) 314 | os.unlink(path) 315 | else: 316 | print('Stop download.') 317 | exit(-1) 318 | 319 | def prepare_graph(self): 320 | processed_dir = osp.join(self.folder, 'processed') 321 | raw_dir = osp.join(self.folder, 'raw') 322 | pre_processed_file_path = osp.join(processed_dir, 'dgl_data_processed') 323 | 324 | if osp.exists(pre_processed_file_path): 325 | # if pre-processed file already exists 326 | self.graphs, label_dict = load_graphs(pre_processed_file_path) 327 | self.labels = label_dict['labels'] 328 | else: 329 | # if pre-processed file does not exist 330 | 331 | if not osp.exists(osp.join(raw_dir, 'data.csv.gz')): 332 | # if the raw file does not exist, then download it. 333 | self.download() 334 | 335 | data_df = pd.read_csv(osp.join(raw_dir, 'data.csv.gz')) 336 | smiles_list = data_df['smiles'] 337 | homolumogap_list = data_df['homolumogap'] 338 | 339 | print('Converting SMILES strings into graphs...') 340 | self.graphs = [] 341 | self.labels = [] 342 | for i in tqdm(range(len(smiles_list))): 343 | 344 | smiles = smiles_list[i] 345 | homolumogap = homolumogap_list[i] 346 | graph = self.smiles2graph(smiles) 347 | 348 | assert(len(graph['edge_feat']) == graph['edge_index'].shape[1]) 349 | assert(len(graph['node_feat']) == graph['num_nodes']) 350 | 351 | src, dst = torch.from_numpy(graph['edge_index']) 352 | num_nodes = graph['num_nodes'] 353 | 354 | if num_nodes == 1: # some graphs have one node 355 | A_ = torch.tensor(1.).view(1, 1) 356 | else: 357 | A = torch.zeros([num_nodes, num_nodes], dtype=torch.float) 358 | A[src, dst] = 1.0 359 | for i in range(num_nodes): 360 | A[i, i] = 1.0 361 | deg = torch.sum(A, axis=0).squeeze() ** -0.5 362 | D = torch.diag(deg) 363 | A_ = D @ A @ D 364 | e, u = torch.linalg.eigh(A_) 365 | 366 | fully_connected = torch.ones([num_nodes, num_nodes], dtype=torch.float).nonzero(as_tuple=True) 367 | g = dgl.graph(fully_connected, num_nodes = num_nodes) 368 | 369 | g.ndata['e'] = e 370 | g.ndata['u'] = u 371 | 372 | if graph['node_feat'] is not None: 373 | g.ndata['feat'] = torch.from_numpy(graph['node_feat']).long() 374 | 375 | if graph['edge_feat'] is not None: 376 | edge_idx = torch.stack([src, dst], dim=0) 377 | edge_attr = torch.from_numpy(graph['edge_feat']).long() + 1 # for padding 378 | 379 | if len(edge_attr.shape) == 1: 380 | edge_attr_dense = to_dense_adj(edge_idx, edge_attr=edge_attr.unsqueeze(-1)).squeeze(0).squeeze(-1).view(-1) 381 | else: 382 | if edge_attr.size(0) == 0: # for graphs without edge 383 | edge_attr_dense = torch.zeros([num_nodes ** 2, edge_attr.size(1)]).long() 384 | else: 385 | edge_attr_dense = to_dense_adj(edge_idx, edge_attr=edge_attr, max_num_nodes=num_nodes).squeeze(0).view(-1, edge_attr.shape[-1]) 386 | 387 | g.edata['feat'] = edge_attr_dense 388 | 389 | self.graphs.append(g) 390 | self.labels.append(homolumogap) 391 | 392 | self.labels = torch.tensor(self.labels, dtype=torch.float32) 393 | 394 | # double-check prediction target 395 | split_dict = self.get_idx_split() 396 | assert(all([not torch.isnan(self.labels[i]) for i in split_dict['train']])) 397 | assert(all([not torch.isnan(self.labels[i]) for i in split_dict['valid']])) 398 | assert(all([torch.isnan(self.labels[i]) for i in split_dict['test-dev']])) 399 | assert(all([torch.isnan(self.labels[i]) for i in split_dict['test-challenge']])) 400 | 401 | print('Saving...') 402 | save_graphs(pre_processed_file_path, self.graphs, labels={'labels': self.labels}) 403 | 404 | 405 | def get_idx_split(self): 406 | split_dict = replace_numpy_with_torchtensor(torch.load(osp.join(self.folder, 'split_dict.pt'))) 407 | return split_dict 408 | 409 | def __getitem__(self, idx): 410 | '''Get datapoint with index''' 411 | 412 | if isinstance(idx, int): 413 | return self.graphs[idx], self.labels[idx] 414 | elif torch.is_tensor(idx) and idx.dtype == torch.long: 415 | if idx.dim() == 0: 416 | return self.graphs[idx], self.labels[idx] 417 | elif idx.dim() == 1: 418 | return Subset(self, idx.cpu()) 419 | 420 | raise IndexError( 421 | 'Only integers and long are valid ' 422 | 'indices (got {}).'.format(type(idx).__name__)) 423 | 424 | def __len__(self): 425 | '''Length of the dataset 426 | Returns 427 | ------- 428 | int 429 | Length of Dataset 430 | ''' 431 | return len(self.graphs) 432 | 433 | def __repr__(self): # pragma: no cover 434 | return '{}({})'.format(self.__class__.__name__, len(self)) 435 | 436 | 437 | class DglZincDataset(InMemoryDataset): 438 | 439 | url = 'https://www.dropbox.com/s/feo9qle74kg48gy/molecules.zip?dl=1' 440 | split_url = ('https://raw.githubusercontent.com/graphdeeplearning/' 441 | 'benchmarking-gnns/master/data/molecules/{}.index') 442 | 443 | def __init__(self, root, subset=False, split='train', transform=None, 444 | pre_transform=None, pre_filter=None): 445 | self.subset = subset 446 | assert split in ['train', 'val', 'test'] 447 | super().__init__(root, transform, pre_transform, pre_filter) 448 | path = osp.join(self.processed_dir, f'dgl_{split}.pt') 449 | print(path) 450 | self.graphs, label_dict = load_graphs(path) 451 | self.labels = label_dict['labels'] 452 | 453 | @property 454 | def raw_file_names(self): 455 | return [ 456 | 'train.pickle', 'val.pickle', 'test.pickle', 'train.index', 457 | 'val.index', 'test.index' 458 | ] 459 | 460 | @property 461 | def processed_dir(self): 462 | name = 'subset' if self.subset else 'full' 463 | return osp.join(self.root, name, 'processed') 464 | 465 | @property 466 | def processed_file_names(self): 467 | return ['dgl_train.pt', 'dgl_val.pt', 'dgl_test.pt'] 468 | 469 | def download(self): 470 | shutil.rmtree(self.raw_dir) 471 | path = download_url(self.url, self.root) 472 | extract_zip(path, self.root) 473 | os.rename(osp.join(self.root, 'molecules'), self.raw_dir) 474 | os.unlink(path) 475 | 476 | for split in ['train', 'val', 'test']: 477 | download_url(self.split_url.format(split), self.raw_dir) 478 | 479 | def process(self): 480 | for split in ['train', 'val', 'test']: 481 | with open(osp.join(self.raw_dir, f'{split}.pickle'), 'rb') as f: 482 | mols = pickle.load(f) 483 | 484 | indices = range(len(mols)) 485 | 486 | if self.subset: 487 | with open(osp.join(self.raw_dir, f'{split}.index'), 'r') as f: 488 | indices = [int(x) for x in f.read()[:-1].split(',')] 489 | 490 | pbar = tqdm(total=len(indices)) 491 | pbar.set_description(f'Processing {split} dataset') 492 | 493 | graphs = [] 494 | labels = [] 495 | for idx in indices: 496 | mol = mols[idx] 497 | 498 | x = mol['atom_type'].to(torch.long).view(-1, 1) 499 | y = mol['logP_SA_cycle_normalized'].to(torch.float) 500 | 501 | adj = mol['bond_type'] 502 | edge_idx = adj.nonzero(as_tuple=False).t().contiguous() 503 | edge_attr = adj[edge_idx[0], edge_idx[1]].to(torch.long) + 1 # for padding 504 | 505 | src, dst = edge_idx 506 | num_nodes = x.size(0) 507 | 508 | if num_nodes == 1: # some graphs have one node 509 | A_ = torch.tensor(1.).view(1, 1) 510 | else: 511 | A = torch.zeros([num_nodes, num_nodes], dtype=torch.float) 512 | A[src, dst] = 1.0 513 | for i in range(num_nodes): 514 | A[i, i] = 1.0 515 | deg = torch.sum(A, axis=0).squeeze() ** -0.5 516 | D = torch.diag(deg) 517 | A_ = D @ A @ D 518 | e, u = torch.linalg.eigh(A_) 519 | 520 | fully_connected = torch.ones([num_nodes, num_nodes], dtype=torch.float).nonzero(as_tuple=True) 521 | g = dgl.graph(fully_connected, num_nodes = num_nodes) 522 | 523 | g.ndata['e'] = e 524 | g.ndata['u'] = u 525 | 526 | g.ndata['feat'] = x 527 | g.edata['feat'] = to_dense_adj(edge_idx, edge_attr=edge_attr.unsqueeze(-1)).squeeze(0).squeeze(-1).view(-1) 528 | 529 | if self.pre_filter is not None and not self.pre_filter(data): 530 | continue 531 | 532 | if self.pre_transform is not None: 533 | data = self.pre_transform(data) 534 | 535 | graphs.append(g) 536 | labels.append(y) 537 | 538 | pbar.update(1) 539 | 540 | pbar.close() 541 | 542 | labels = torch.tensor(labels, dtype=torch.float32) 543 | save_graphs(osp.join(self.processed_dir, f'dgl_{split}.pt'), graphs, labels={'labels': labels}) 544 | 545 | def __getitem__(self, idx): 546 | '''Get datapoint with index''' 547 | 548 | if isinstance(idx, int): 549 | return self.graphs[idx], self.labels[idx] 550 | elif torch.is_tensor(idx) and idx.dtype == torch.long: 551 | if idx.dim() == 0: 552 | return self.graphs[idx], self.labels[idx] 553 | elif idx.dim() == 1: 554 | return Subset(self, idx.cpu()) 555 | 556 | raise IndexError( 557 | 'Only integers and long are valid ' 558 | 'indices (got {}).'.format(type(idx).__name__)) 559 | 560 | def __len__(self): 561 | '''Length of the dataset 562 | Returns 563 | ------- 564 | int 565 | Length of Dataset 566 | ''' 567 | return len(self.graphs) 568 | 569 | -------------------------------------------------------------------------------- /Graph/get_dataset.py: -------------------------------------------------------------------------------- 1 | from dataclass import * 2 | from dgldataclass import DglGraphPropPredDataset, DglPCQM4Mv2Dataset, DglZincDataset 3 | from pygdataclass import PygGraphPropPredDataset 4 | import dgl 5 | from dgl.data.utils import load_graphs, save_graphs, Subset 6 | import torch 7 | from torch.nn import functional as F 8 | from torch.utils.data import DataLoader, Sampler, RandomSampler 9 | from torch_geometric.data import InMemoryDataset, Data 10 | from ogb.graphproppred import Evaluator 11 | from torch_geometric.utils import to_dense_adj 12 | 13 | 14 | class PCQM4Mv2Evaluator: 15 | def __init__(self): 16 | ''' 17 | Evaluator for the PCQM4Mv2 dataset 18 | Metric is Mean Absolute Error 19 | ''' 20 | pass 21 | 22 | def eval(self, input_dict): 23 | ''' 24 | y_true: numpy.ndarray or torch.Tensor of shape (num_graphs,) 25 | y_pred: numpy.ndarray or torch.Tensor of shape (num_graphs,) 26 | y_true and y_pred need to be of the same type (either numpy.ndarray or torch.Tensor) 27 | ''' 28 | assert('y_pred' in input_dict) 29 | assert('y_true' in input_dict) 30 | 31 | y_pred, y_true = input_dict['y_pred'].reshape(-1), input_dict['y_true'].reshape(-1) 32 | 33 | assert((isinstance(y_true, np.ndarray) and isinstance(y_pred, np.ndarray)) 34 | or 35 | (isinstance(y_true, torch.Tensor) and isinstance(y_pred, torch.Tensor))) 36 | assert(y_true.shape == y_pred.shape) 37 | assert(len(y_true.shape) == 1) 38 | 39 | if isinstance(y_true, torch.Tensor): 40 | return {'mae': torch.mean(torch.abs(y_pred - y_true)).cpu().item()} 41 | else: 42 | return {'mae': float(np.mean(np.absolute(y_pred - y_true)))} 43 | 44 | 45 | class DynamicBatchSampler(Sampler): 46 | def __init__(self, sampler, num_nodes_list, batch_size=32, max_nodes=200, drop_last=False): 47 | 48 | super(DynamicBatchSampler, self).__init__(sampler) 49 | self.sampler = sampler 50 | self.num_nodes_list = num_nodes_list 51 | self.batch_size = batch_size 52 | self.max_nodes = max_nodes 53 | self.drop_last = drop_last 54 | 55 | def __iter__(self): 56 | 57 | batch = [] 58 | total_nodes = 0 59 | memory = self.max_nodes * self.max_nodes * self.batch_size 60 | 61 | for idx in self.sampler: 62 | cur_nodes = self.num_nodes_list[idx] 63 | 64 | # beyond memory, truncate batch 65 | # squre for Transformer 66 | if total_nodes + cur_nodes ** 2 > memory: 67 | yield batch 68 | batch = [idx] 69 | total_nodes = cur_nodes ** 2 70 | else: 71 | batch.append(idx) 72 | total_nodes += cur_nodes ** 2 73 | 74 | if len(batch) == self.batch_size: 75 | yield batch 76 | batch = [] 77 | total_nodes = 0 78 | 79 | if len(batch) > 0 and not self.drop_last: 80 | yield batch 81 | 82 | def __len__(self): 83 | # we do not know the exactly batch size, so do not call len(dataloader) 84 | pass 85 | 86 | 87 | def collate_dgl(samples): 88 | graphs, labels = map(list, zip(*samples)) 89 | 90 | graph_list = [] 91 | length = [] 92 | E = [] 93 | U = [] 94 | 95 | max_nodes = max([g.num_nodes() for g in graphs]) 96 | 97 | for i, g in enumerate(graphs): 98 | num_nodes = g.num_nodes() 99 | 100 | e = g.ndata['e'] 101 | u = g.ndata['u'] 102 | 103 | pad_e = e.new_zeros([max_nodes]) 104 | pad_e[:num_nodes] = e 105 | 106 | pad_u = u.new_zeros([max_nodes, max_nodes]) 107 | pad_u[:num_nodes, :num_nodes] = u 108 | 109 | E.append(pad_e) 110 | U.append(pad_u) 111 | graph_list.append(g) 112 | length.append(num_nodes) 113 | 114 | E = torch.stack(E, 0) 115 | U = torch.stack(U, 0) 116 | length = torch.LongTensor(length) 117 | batched_graph = dgl.batch(graphs, ndata=['feat'], edata=['feat']) 118 | 119 | if isinstance(labels[0], torch.Tensor): 120 | return E, U, batched_graph, length, torch.stack(labels) 121 | else: 122 | return E, U, batched_graph, length, labels 123 | 124 | 125 | def collate_pad(batch): 126 | E = [] 127 | U = [] 128 | X = [] 129 | F = [] 130 | Y = [] 131 | 132 | max_nodes = min(max([data.num_nodes for data in batch]), 128) 133 | 134 | for data in batch: 135 | length = data.num_nodes 136 | e = data.e 137 | u = data.u.view(length, length) 138 | x = data.x 139 | f = data.edge_attr 140 | 141 | if length > max_nodes: 142 | src, dst = data.edge_index 143 | A = torch.zeros([length, length], dtype=torch.float) 144 | A[src, dst] = 1.0 145 | A = A[:max_nodes, :max_nodes] 146 | deg = torch.sum(A, axis=0).squeeze() 147 | deg = torch.clamp(deg, min=1.0) ** -0.5 148 | D = torch.diag(deg) 149 | A_ = D @ A @ D 150 | 151 | pad_e, pad_u = torch.linalg.eigh(A_) 152 | pad_x = x[:max_nodes, :] + 1 153 | 154 | fdim = f.size(-1) 155 | pad_f = torch.zeros([length, length, fdim], dtype=torch.long) 156 | pad_f[src, dst] = f + 1 157 | pad_f = pad_f[:max_nodes, :max_nodes] 158 | else: 159 | pad_e = e.new_zeros([max_nodes]) 160 | pad_e[:length] = e 161 | 162 | pad_u = u.new_zeros([max_nodes, max_nodes]) 163 | pad_u[:length, :length] = u 164 | 165 | xdim = x.size(-1) 166 | pad_x = x.new_zeros([max_nodes, xdim]) 167 | pad_x[:length, :] = x + 1 168 | 169 | fdim = f.size(-1) 170 | src, dst = data.edge_index 171 | pad_f = f.new_zeros([max_nodes, max_nodes, fdim]) 172 | pad_f[src, dst, :] = f + 1 173 | 174 | E.append(pad_e) 175 | U.append(pad_u) 176 | X.append(pad_x) 177 | F.append(pad_f) 178 | Y.append(data.y.squeeze()) 179 | 180 | return torch.stack(E, 0), torch.stack(U, 0), torch.stack(X, 0), torch.stack(F, 0), torch.stack(Y, 0) 181 | 182 | 183 | def get_dataset(dataset_name='abaaba'): 184 | 185 | if dataset_name == 'zinc': 186 | data_info = { 187 | 'num_class': 1, 188 | 'loss_fn': F.l1_loss, 189 | 'metric': 'mae', 190 | 'metric_mode': 'min', 191 | 'evaluator': PCQM4Mv2Evaluator(), 192 | 'train_dataset': DglZincDataset('dataset/zinc', subset=True, split='train'), 193 | 'valid_dataset': DglZincDataset('dataset/zinc', subset=True, split='val'), 194 | 'test_dataset': DglZincDataset('dataset/zinc', subset=True, split='test'), 195 | } 196 | elif dataset_name == 'pcqm': 197 | dataset = DglPCQM4Mv2Dataset() 198 | split_idx = dataset.get_idx_split() 199 | idx = split_idx['train'] 200 | rand_idx = torch.randperm(idx.size(0)) 201 | train_idx = idx[rand_idx[150000:]] 202 | valid_idx = idx[rand_idx[:150000]] 203 | test_idx = split_idx['valid'] 204 | 205 | data_info = { 206 | 'num_class': 1, 207 | 'loss_fn': F.l1_loss, 208 | 'metric': 'mae', 209 | 'metric_mode': 'min', 210 | 'evaluator': PCQM4Mv2Evaluator(), 211 | 'train_dataset': dataset[train_idx], 212 | 'valid_dataset': dataset[valid_idx], 213 | 'test_dataset': dataset[test_idx], 214 | } 215 | elif dataset_name == 'pcqms': 216 | train_g, train_dict = load_graphs('dataset/pcqm_subset_train.pt') 217 | valid_g, valid_dict = load_graphs('dataset/pcqm_subset_valid.pt') 218 | test_g, test_dict = load_graphs('dataset/pcqm_subset_test.pt') 219 | 220 | data_info = { 221 | 'num_class': 1, 222 | 'loss_fn': F.l1_loss, 223 | 'metric': 'mae', 224 | 'metric_mode': 'min', 225 | 'evaluator': PCQM4Mv2Evaluator(), 226 | 'train_dataset': list(zip(train_g, train_dict['labels'])), 227 | 'valid_dataset': list(zip(valid_g, valid_dict['labels'])), 228 | 'test_dataset': list(zip(test_g, test_dict['labels'])), 229 | } 230 | elif dataset_name == 'hiv': 231 | dataset = DglGraphPropPredDataset('ogbg-molhiv') 232 | split_idx = dataset.get_idx_split() 233 | data_info = { 234 | 'num_class': 1, 235 | 'loss_fn': F.binary_cross_entropy_with_logits, 236 | 'metric': 'rocauc', 237 | 'metric_mode': 'max', 238 | 'evaluator': Evaluator('ogbg-molhiv'), 239 | 'train_dataset': dataset[split_idx['train']], 240 | 'valid_dataset': dataset[split_idx['valid']], 241 | 'test_dataset': dataset[split_idx['test']], 242 | } 243 | elif dataset_name == 'pcba': 244 | dataset = DglGraphPropPredDataset(name = 'ogbg-molpcba') 245 | split_idx = dataset.get_idx_split() 246 | data_info = { 247 | 'num_class': 128, 248 | 'loss_fn': F.binary_cross_entropy_with_logits, 249 | 'metric': 'ap', 250 | 'metric_mode': 'max', 251 | 'evaluator': Evaluator('ogbg-molpcba'), 252 | 'train_dataset': dataset[split_idx['train']], 253 | 'valid_dataset': dataset[split_idx['valid']], 254 | 'test_dataset': dataset[split_idx['test']], 255 | } 256 | elif dataset_name == 'ppa': 257 | dataset = PygGraphPropPredDataset(name = 'ogbg-ppa') 258 | split_idx = dataset.get_idx_split() 259 | data_info = { 260 | 'num_class': 37, 261 | 'loss_fn': F.cross_entropy, 262 | 'metric': 'acc', 263 | 'metric_mode': 'max', 264 | 'evaluator': Evaluator('ogbg-ppa'), 265 | 'train_dataset': dataset[split_idx['train']], 266 | 'valid_dataset': dataset[split_idx['valid']], 267 | 'test_dataset': dataset[split_idx['test']], 268 | } 269 | else: 270 | raise NotImplementedError 271 | 272 | return data_info 273 | 274 | -------------------------------------------------------------------------------- /Graph/large_model.py: -------------------------------------------------------------------------------- 1 | import time 2 | import math 3 | import torch 4 | import torch.nn as nn 5 | import torch.nn.functional as F 6 | import numpy as np 7 | 8 | from ogb.utils.features import get_atom_feature_dims, get_bond_feature_dims 9 | from dgl.nn.pytorch.glob import AvgPooling 10 | from dgl import function as fn 11 | from dgl.ops.edge_softmax import edge_softmax 12 | from torch.nn.init import xavier_uniform_, xavier_normal_, constant_ 13 | 14 | 15 | class AtomEncoder(torch.nn.Module): 16 | 17 | def __init__(self, emb_dim): 18 | super(AtomEncoder, self).__init__() 19 | 20 | self.atom_embedding_list = torch.nn.ModuleList() 21 | 22 | for _, dim in enumerate(get_atom_feature_dims()): 23 | emb = torch.nn.Embedding(dim, emb_dim) 24 | torch.nn.init.xavier_uniform_(emb.weight.data) 25 | self.atom_embedding_list.append(emb) 26 | 27 | def forward(self, x): 28 | x_embedding = 0 29 | for i in range(x.shape[1]): 30 | x_embedding += self.atom_embedding_list[i](x[:, i]) 31 | 32 | return x_embedding 33 | 34 | 35 | class BondEncoder(torch.nn.Module): 36 | 37 | def __init__(self, emb_dim): 38 | super(BondEncoder, self).__init__() 39 | 40 | self.bond_embedding_list = torch.nn.ModuleList() 41 | 42 | for _, dim in enumerate(get_bond_feature_dims()): 43 | emb = torch.nn.Embedding(dim + 1, emb_dim, padding_idx=0) # for padding 44 | torch.nn.init.xavier_uniform_(emb.weight.data) 45 | self.bond_embedding_list.append(emb) 46 | 47 | def forward(self, edge_attr): 48 | bond_embedding = 0 49 | for i in range(edge_attr.shape[1]): 50 | bond_embedding += self.bond_embedding_list[i](edge_attr[:, i]) 51 | 52 | return bond_embedding 53 | 54 | 55 | class SineEncoding(nn.Module): 56 | def __init__(self, hidden_dim=128): 57 | super(SineEncoding, self).__init__() 58 | self.constant = 100 59 | self.hidden_dim = hidden_dim 60 | self.eig_w = nn.Linear(hidden_dim + 1, hidden_dim) 61 | 62 | def forward(self, e): 63 | # input: [B, N] 64 | # output: [B, N, d] 65 | 66 | ee = e * self.constant 67 | div = torch.exp(torch.arange(0, self.hidden_dim, 2) * (-math.log(10000)/self.hidden_dim)).to(e.device) 68 | pe = ee.unsqueeze(2) * div 69 | eeig = torch.cat((e.unsqueeze(2), torch.sin(pe), torch.cos(pe)), dim=2) 70 | 71 | return self.eig_w(eeig) 72 | 73 | 74 | class FeedForwardNetwork(nn.Module): 75 | 76 | def __init__(self, input_dim, hidden_dim, output_dim): 77 | super(FeedForwardNetwork, self).__init__() 78 | self.layer1 = nn.Linear(input_dim, hidden_dim) 79 | self.gelu = nn.GELU() 80 | self.layer2 = nn.Linear(hidden_dim, output_dim) 81 | 82 | def forward(self, x): 83 | x = self.layer1(x) 84 | x = self.gelu(x) 85 | x = self.layer2(x) 86 | return x 87 | 88 | 89 | class Conv(nn.Module): 90 | def __init__(self, hidden_dim, nheads, trans_dropout, feat_dropout, adj_dropout): 91 | super(Conv, self).__init__() 92 | self.nheads = nheads 93 | 94 | self.mha_norm = nn.LayerNorm(hidden_dim) 95 | self.ffn_norm = nn.LayerNorm(hidden_dim) 96 | self.mha_dropout = nn.Dropout(trans_dropout) 97 | self.ffn_dropout = nn.Dropout(trans_dropout) 98 | self.mha = nn.MultiheadAttention(hidden_dim, nheads, trans_dropout, batch_first=True) 99 | self.ffn = FeedForwardNetwork(hidden_dim, hidden_dim, hidden_dim) 100 | self.decoder = nn.Linear(hidden_dim, nheads) 101 | 102 | self.adj_dropout = nn.Dropout(adj_dropout) 103 | self.filter_encoder = nn.Sequential( 104 | nn.Linear(nheads + 1, hidden_dim), 105 | nn.BatchNorm1d(hidden_dim), 106 | nn.GELU(), 107 | nn.Linear(hidden_dim, hidden_dim), 108 | nn.BatchNorm1d(hidden_dim), 109 | nn.GELU(), 110 | ) 111 | 112 | self.pre_ffn = nn.Sequential( 113 | nn.Linear(hidden_dim, hidden_dim), 114 | nn.GELU() 115 | ) 116 | 117 | self.preffn_dropout = nn.Dropout(feat_dropout) 118 | self.x_ffn_dropout = nn.Dropout(feat_dropout) 119 | 120 | self.x_ffn = nn.Sequential( 121 | nn.Linear(hidden_dim, hidden_dim), 122 | nn.BatchNorm1d(hidden_dim), 123 | nn.ReLU(), 124 | nn.Linear(hidden_dim, hidden_dim), 125 | nn.BatchNorm1d(hidden_dim), 126 | nn.ReLU() 127 | ) 128 | 129 | def forward(self, eig, u, ut, graph, x_feat, edge_attr, eig_mask, edge_idx): 130 | B, N = eig.size()[:2] 131 | 132 | mha_eig = self.mha_norm(eig) 133 | mha_eig, attn = self.mha(mha_eig, mha_eig, mha_eig, key_padding_mask=eig_mask) 134 | eig = eig + self.mha_dropout(mha_eig) 135 | 136 | ffn_eig = self.ffn_norm(eig) 137 | ffn_eig = self.ffn(ffn_eig) 138 | eig = eig + self.ffn_dropout(ffn_eig) # [B, N, d] 139 | 140 | new_e = self.decoder(eig).transpose(1, 2) # [B, m, N] 141 | diag_e = torch.diag_embed(new_e) # [B, m, N, N] 142 | 143 | identity = torch.diag_embed(torch.ones(B, N)).to(u.device) 144 | bases = [identity] 145 | for i in range(self.nheads): 146 | filters = u @ diag_e[:, i, :, :] @ ut 147 | bases.append(filters) 148 | 149 | bases = torch.stack(bases, axis=-1) # [B, N, N, H] 150 | bases = bases[edge_idx] 151 | bases = self.adj_dropout(self.filter_encoder(bases)) 152 | bases = edge_softmax(graph, bases) 153 | 154 | with graph.local_scope(): 155 | graph.ndata['x'] = x_feat 156 | graph.apply_edges(fn.copy_u('x', '_x')) 157 | xee = self.pre_ffn(graph.edata['_x'] + edge_attr) * bases 158 | graph.edata['v'] = xee 159 | graph.update_all(fn.copy_e('v', '_aggr_e'), fn.sum('_aggr_e', 'aggr_e')) 160 | y = graph.ndata['aggr_e'] 161 | y = self.preffn_dropout(y) 162 | x = x_feat + y 163 | y = self.x_ffn(x) 164 | y = self.x_ffn_dropout(y) 165 | x = x + y 166 | 167 | return eig, x 168 | 169 | 170 | class SpecformerLarge(nn.Module): 171 | 172 | def __init__(self, nclass, nlayer, hidden_dim=128, nheads=4, feat_dropout=0.1, trans_dropout=0.1, adj_dropout=0.1): 173 | super(SpecformerLarge, self).__init__() 174 | 175 | self.nlayer = nlayer 176 | self.nclass = nclass 177 | self.hidden_dim = hidden_dim 178 | self.nheads = nheads 179 | 180 | self.atom_encoder = AtomEncoder(hidden_dim) 181 | self.bond_encoder = BondEncoder(hidden_dim) 182 | self.eig_encoder = SineEncoding(hidden_dim) 183 | 184 | self.convs = nn.ModuleList([Conv(hidden_dim, nheads, trans_dropout, feat_dropout, adj_dropout) for _ in range(nlayer)]) 185 | self.pool = AvgPooling() 186 | self.linear = nn.Linear(hidden_dim, nclass) 187 | 188 | def forward(self, e, u, g, length): 189 | 190 | # e: [B, N] eigenvalues 191 | # u: [B, N, N] eigenvectors 192 | # x: [B, N, d] node features 193 | # f: [B, N, N, d] edge features 194 | 195 | B, N = e.size() 196 | ut = u.transpose(1, 2) 197 | 198 | node_feat = g.ndata['feat'] 199 | edge_feat = g.edata['feat'] 200 | 201 | # do not use u to generate edge_idx because of the connected components 202 | e_mask, edge_idx = self.length_to_mask(length) 203 | 204 | node_feat = self.atom_encoder(node_feat) 205 | edge_feat = self.bond_encoder(edge_feat) 206 | eig = self.eig_encoder(e) 207 | 208 | for conv in self.convs: 209 | eig, node_feat = conv(eig, u, ut, g, node_feat, edge_feat, e_mask, edge_idx) 210 | 211 | h = self.pool(g, node_feat) 212 | h = self.linear(h) 213 | 214 | return h 215 | 216 | 217 | def length_to_mask(self, length): 218 | ''' 219 | length: [B] 220 | return: [B, max_len]. 221 | ''' 222 | B = len(length) 223 | N = length.max().item() 224 | mask1d = torch.arange(N, device=length.device).expand(B, N) >= length.unsqueeze(1) 225 | mask2d = (~mask1d).float().unsqueeze(2) @ (~mask1d).float().unsqueeze(1) 226 | mask2d = mask2d.bool() 227 | 228 | # Example 229 | # length=[1, 2, 3], B=3, N=3, 230 | 231 | # mask1d for key_padding_mask of MultiheadAttention [B, N] 232 | # [False, True, True ] 233 | # [False, False, True ] 234 | # [False, False, False] 235 | 236 | # mask2d for edge indexing [B, N, N] 237 | # [[1, 0, 0], | [1, 1, 0], | [1, 1, 1], 238 | # [0, 0, 0], | [1, 1, 0], | [1, 1, 1], 239 | # [0, 0, 0], | [0, 0, 0], | [1, 1, 1],] 240 | 241 | return mask1d, mask2d 242 | 243 | 244 | ''' 245 | def length_to_mask(self, length): 246 | ''' 247 | length: [B] 248 | return: [B, max_len]. 249 | ''' 250 | B = len(length) 251 | N = length.max().item() 252 | mask1d = torch.arange(N, device=length.device).expand(B, N) >= length.unsqueeze(1) 253 | 254 | mask2d = torch.zeros(B, N, N, device=length.device) 255 | for i in range(B): 256 | mask2d[i, :length[i], :length[i]] = 1.0 257 | 258 | # mask1d for key_padding_mask, a True value indicates that the corresponding key value will be ignored for the purpose of attention 259 | # mask2d for edge selection from padding 260 | return mask1d, mask2d.bool() 261 | ''' 262 | -------------------------------------------------------------------------------- /Graph/master.csv: -------------------------------------------------------------------------------- 1 | ,ogbg-molbace,ogbg-molbbbp,ogbg-molclintox,ogbg-molmuv,ogbg-molpcba,ogbg-molsider,ogbg-moltox21,ogbg-moltoxcast,ogbg-molhiv,ogbg-molesol,ogbg-molfreesolv,ogbg-mollipo,ogbg-molchembl,ogbg-ppa,ogbg-code2 2 | num tasks,1,1,2,17,128,27,12,617,1,1,1,1,1310,1,1 3 | eval metric,rocauc,rocauc,rocauc,ap,ap,rocauc,rocauc,rocauc,rocauc,rmse,rmse,rmse,rocauc,acc,F1 4 | download_name,bace,bbbp,clintox,muv,pcba,sider,tox21,toxcast,hiv,esol,freesolv,lipophilicity,chembl,ogbg_ppi_medium,code2 5 | version,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1 6 | url,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/bace.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/bbbp.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/clintox.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/muv.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/pcba.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/sider.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/tox21.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/toxcast.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/hiv.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/esol.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/freesolv.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/lipophilicity.zip,http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/chembl.zip,http://snap.stanford.edu/ogb/data/graphproppred/ogbg_ppi_medium.zip,http://snap.stanford.edu/ogb/data/graphproppred/code2.zip 7 | add_inverse_edge,True,True,True,True,True,True,True,True,True,True,True,True,True,True,False 8 | data type,mol,mol,mol,mol,mol,mol,mol,mol,mol,mol,mol,mol,mol,, 9 | has_node_attr,True,True,True,True,True,True,True,True,True,True,True,True,True,False,True 10 | has_edge_attr,True,True,True,True,True,True,True,True,True,True,True,True,True,True,False 11 | task type,binary classification,binary classification,binary classification,binary classification,binary classification,binary classification,binary classification,binary classification,binary classification,regression,regression,regression,binary classification,multiclass classification,subtoken prediction 12 | num classes,2,2,2,2,2,2,2,2,2,-1,-1,-1,2,37,-1 13 | split,scaffold,scaffold,scaffold,scaffold,scaffold,scaffold,scaffold,scaffold,scaffold,scaffold,scaffold,scaffold,scaffold,species,project 14 | additional node files,None,None,None,None,None,None,None,None,None,None,None,None,None,None,"node_is_attributed,node_dfs_order,node_depth" 15 | additional edge files,None,None,None,None,None,None,None,None,None,None,None,None,None,None,None 16 | binary,False,False,False,False,False,False,False,False,False,False,False,False,False,False,False 17 | -------------------------------------------------------------------------------- /Graph/medium_model.py: -------------------------------------------------------------------------------- 1 | import time 2 | import math 3 | import torch 4 | import torch.nn as nn 5 | import torch.nn.functional as F 6 | import numpy as np 7 | 8 | from ogb.utils.features import get_atom_feature_dims, get_bond_feature_dims 9 | from dgl.nn.pytorch.glob import AvgPooling 10 | from dgl import function as fn 11 | from dgl.ops.edge_softmax import edge_softmax 12 | from torch.nn.init import xavier_uniform_, xavier_normal_, constant_ 13 | 14 | 15 | class AtomEncoder(torch.nn.Module): 16 | 17 | def __init__(self, emb_dim): 18 | super(AtomEncoder, self).__init__() 19 | 20 | self.atom_embedding_list = torch.nn.ModuleList() 21 | 22 | for _, dim in enumerate(get_atom_feature_dims()): 23 | emb = torch.nn.Embedding(dim, emb_dim) 24 | torch.nn.init.xavier_uniform_(emb.weight.data) 25 | self.atom_embedding_list.append(emb) 26 | 27 | def forward(self, x): 28 | x_embedding = 0 29 | for i in range(x.shape[1]): 30 | x_embedding += self.atom_embedding_list[i](x[:, i]) 31 | 32 | return x_embedding 33 | 34 | 35 | class BondEncoder(torch.nn.Module): 36 | 37 | def __init__(self, emb_dim): 38 | super(BondEncoder, self).__init__() 39 | 40 | self.bond_embedding_list = torch.nn.ModuleList() 41 | 42 | for _, dim in enumerate(get_bond_feature_dims()): 43 | emb = torch.nn.Embedding(dim + 1, emb_dim, padding_idx=0) # for padding 44 | torch.nn.init.xavier_uniform_(emb.weight.data) 45 | self.bond_embedding_list.append(emb) 46 | 47 | def forward(self, edge_attr): 48 | bond_embedding = 0 49 | for i in range(edge_attr.shape[1]): 50 | bond_embedding += self.bond_embedding_list[i](edge_attr[:, i]) 51 | 52 | return bond_embedding 53 | 54 | 55 | class SineEncoding(nn.Module): 56 | def __init__(self, hidden_dim=128): 57 | super(SineEncoding, self).__init__() 58 | self.constant = 100 59 | self.hidden_dim = hidden_dim 60 | self.eig_w = nn.Linear(hidden_dim + 1, hidden_dim) 61 | 62 | def forward(self, e): 63 | # input: [B, N] 64 | # output: [B, N, d] 65 | 66 | ee = e * self.constant 67 | div = torch.exp(torch.arange(0, self.hidden_dim, 2) * (-math.log(10000)/self.hidden_dim)).to(e.device) 68 | pe = ee.unsqueeze(2) * div 69 | eeig = torch.cat((e.unsqueeze(2), torch.sin(pe), torch.cos(pe)), dim=2) 70 | 71 | return self.eig_w(eeig) 72 | 73 | 74 | class FeedForwardNetwork(nn.Module): 75 | 76 | def __init__(self, input_dim, hidden_dim, output_dim): 77 | super(FeedForwardNetwork, self).__init__() 78 | self.layer1 = nn.Linear(input_dim, hidden_dim) 79 | self.gelu = nn.GELU() 80 | self.layer2 = nn.Linear(hidden_dim, output_dim) 81 | 82 | def forward(self, x): 83 | x = self.layer1(x) 84 | x = self.gelu(x) 85 | x = self.layer2(x) 86 | return x 87 | 88 | 89 | class Conv(nn.Module): 90 | def __init__(self, nheads, hidden_size, feat_dropout, adj_dropout): 91 | super(Conv, self).__init__() 92 | 93 | self.adj_dropout = nn.Dropout(adj_dropout) 94 | self.filter_encoder = nn.Sequential( 95 | nn.Linear(nheads + 1, hidden_size), 96 | nn.BatchNorm1d(hidden_size), 97 | nn.GELU(), 98 | nn.Linear(hidden_size, hidden_size), 99 | nn.BatchNorm1d(hidden_size), 100 | nn.GELU(), 101 | ) 102 | 103 | self.pre_ffn = nn.Sequential( 104 | nn.Linear(hidden_size, hidden_size), 105 | # nn.BatchNorm1d(hidden_size), 106 | nn.GELU() 107 | ) 108 | 109 | self.preffn_dropout = nn.Dropout(feat_dropout) 110 | self.ffn_dropout = nn.Dropout(feat_dropout) 111 | 112 | self.ffn = nn.Sequential( 113 | nn.Linear(hidden_size, hidden_size), 114 | nn.BatchNorm1d(hidden_size), 115 | nn.ReLU(), 116 | nn.Linear(hidden_size, hidden_size), 117 | nn.BatchNorm1d(hidden_size), 118 | nn.ReLU() 119 | ) 120 | 121 | def forward(self, graph, x_feat, edge_attr, bases): 122 | bases = self.adj_dropout(self.filter_encoder(bases)) 123 | bases = edge_softmax(graph, bases) 124 | 125 | with graph.local_scope(): 126 | graph.ndata['x'] = x_feat 127 | graph.apply_edges(fn.copy_u('x', '_x')) 128 | xee = self.pre_ffn(graph.edata['_x'] + edge_attr) * bases 129 | graph.edata['v'] = xee 130 | graph.update_all(fn.copy_e('v', '_aggr_e'), fn.sum('_aggr_e', 'aggr_e')) 131 | y = graph.ndata['aggr_e'] 132 | y = self.preffn_dropout(y) 133 | x = x_feat + y 134 | y = self.ffn(x) 135 | y = self.ffn_dropout(y) 136 | x = x + y 137 | return x 138 | 139 | 140 | class SpecformerMedium(nn.Module): 141 | 142 | def __init__(self, nclass, nlayer, hidden_dim=128, nheads=4, feat_dropout=0.1, trans_dropout=0.1, adj_dropout=0.1): 143 | super(SpecformerMedium, self).__init__() 144 | 145 | print('medium model') 146 | self.nlayer = nlayer 147 | self.nclass = nclass 148 | self.hidden_dim = hidden_dim 149 | self.nheads = nheads 150 | 151 | self.atom_encoder = AtomEncoder(hidden_dim) 152 | self.bond_encoder = BondEncoder(hidden_dim) 153 | 154 | self.eig_encoder = SineEncoding(hidden_dim) 155 | #self.eig_encoder = nn.Linear(1, hidden_dim) # ablation 156 | self.decoder = nn.Linear(hidden_dim, nheads) 157 | 158 | self.mha_norm = nn.LayerNorm(hidden_dim) 159 | self.ffn_norm = nn.LayerNorm(hidden_dim) 160 | self.mha_dropout = nn.Dropout(trans_dropout) 161 | self.ffn_dropout = nn.Dropout(trans_dropout) 162 | self.mha = nn.MultiheadAttention(hidden_dim, nheads, trans_dropout, batch_first=True) 163 | self.ffn = FeedForwardNetwork(hidden_dim, hidden_dim, hidden_dim) 164 | 165 | self.convs = nn.ModuleList([Conv(nheads, hidden_dim, feat_dropout, adj_dropout) for _ in range(nlayer)]) 166 | self.pool = AvgPooling() 167 | self.linear = nn.Linear(hidden_dim, nclass) 168 | 169 | 170 | def forward(self, e, u, g, length): 171 | 172 | # e: [B, N] eigenvalues 173 | # u: [B, N, N] eigenvectors 174 | # x: [B, N, d] node features 175 | # f: [B, N, N, d] edge features 176 | 177 | B, N = e.size() 178 | ut = u.transpose(1, 2) 179 | 180 | node_feat = g.ndata['feat'] 181 | edge_feat = g.edata['feat'] 182 | 183 | # do not use u to generate edge_idx because of the connected components 184 | e_mask, edge_idx = self.length_to_mask(length) 185 | 186 | node_feat = self.atom_encoder(node_feat) 187 | edge_feat = self.bond_encoder(edge_feat) 188 | 189 | eig = self.eig_encoder(e) 190 | 191 | mha_eig = self.mha_norm(eig) 192 | mha_eig, attn = self.mha(mha_eig, mha_eig, mha_eig, key_padding_mask=e_mask) 193 | eig = eig + self.mha_dropout(mha_eig) 194 | 195 | ffn_eig = self.ffn_norm(eig) 196 | ffn_eig = self.ffn(ffn_eig) 197 | eig = eig + self.ffn_dropout(ffn_eig) 198 | 199 | new_e = self.decoder(eig).transpose(2, 1) # [B, m, N] 200 | diag_e = torch.diag_embed(new_e) # [B, m, N, N] 201 | 202 | identity = torch.diag_embed(torch.ones_like(e)) 203 | bases = [identity] 204 | for i in range(self.nheads): 205 | filters = u @ diag_e[:, i, :, :] @ ut 206 | bases.append(filters) 207 | 208 | bases = torch.stack(bases, axis=-1) # [B, N, N, H] 209 | bases = bases[edge_idx] 210 | 211 | for conv in self.convs: 212 | node_feat = conv(g, node_feat, edge_feat, bases) 213 | 214 | h = self.pool(g, node_feat) 215 | h = self.linear(h) 216 | 217 | return h 218 | 219 | 220 | def length_to_mask(self, length): 221 | ''' 222 | length: [B] 223 | return: [B, max_len]. 224 | ''' 225 | B = len(length) 226 | N = length.max().item() 227 | mask1d = torch.arange(N, device=length.device).expand(B, N) >= length.unsqueeze(1) 228 | mask2d = (~mask1d).float().unsqueeze(2) @ (~mask1d).float().unsqueeze(1) 229 | mask2d = mask2d.bool() 230 | 231 | # Example 232 | # length=[1, 2, 3], B=3, N=3, 233 | 234 | # mask1d for key_padding_mask of MultiheadAttention [B, N] 235 | # [False, True, True ] 236 | # [False, False, True ] 237 | # [False, False, False] 238 | 239 | # mask2d for edge indexing [B, N, N] 240 | # [[1, 0, 0], | [1, 1, 0], | [1, 1, 1], 241 | # [0, 0, 0], | [1, 1, 0], | [1, 1, 1], 242 | # [0, 0, 0], | [0, 0, 0], | [1, 1, 1],] 243 | 244 | return mask1d, mask2d 245 | 246 | 247 | ''' 248 | def length_to_mask(self, length): 249 | ''' 250 | length: [B] 251 | return: [B, max_len]. 252 | ''' 253 | B = len(length) 254 | N = length.max().item() 255 | mask1d = torch.arange(N, device=length.device).expand(B, N) >= length.unsqueeze(1) 256 | 257 | mask2d = torch.zeros(B, N, N, device=length.device) 258 | for i in range(B): 259 | mask2d[i, :length[i], :length[i]] = 1.0 260 | 261 | # mask1d for key_padding_mask, a True value indicates that the corresponding key value will be ignored for the purpose of attention 262 | # mask2d for edge selection from padding 263 | return mask1d, mask2d.bool() 264 | ''' 265 | -------------------------------------------------------------------------------- /Graph/small_model.py: -------------------------------------------------------------------------------- 1 | import time 2 | import math 3 | import torch 4 | import torch.nn as nn 5 | import torch.nn.functional as F 6 | import numpy as np 7 | 8 | from ogb.utils.features import get_atom_feature_dims, get_bond_feature_dims 9 | from dgl.nn.pytorch.glob import AvgPooling 10 | from dgl import function as fn 11 | from dgl.ops.edge_softmax import edge_softmax 12 | from torch.nn.init import xavier_uniform_, xavier_normal_, constant_ 13 | 14 | 15 | class AtomEncoder(torch.nn.Module): 16 | 17 | def __init__(self, emb_dim): 18 | super(AtomEncoder, self).__init__() 19 | 20 | self.atom_embedding_list = torch.nn.ModuleList() 21 | 22 | for _, dim in enumerate(get_atom_feature_dims()): 23 | emb = torch.nn.Embedding(dim, emb_dim) 24 | torch.nn.init.xavier_uniform_(emb.weight.data) 25 | self.atom_embedding_list.append(emb) 26 | 27 | def forward(self, x): 28 | x_embedding = 0 29 | for i in range(x.shape[1]): 30 | x_embedding += self.atom_embedding_list[i](x[:, i]) 31 | 32 | return x_embedding 33 | 34 | 35 | class BondEncoder(torch.nn.Module): 36 | 37 | def __init__(self, emb_dim): 38 | super(BondEncoder, self).__init__() 39 | 40 | self.bond_embedding_list = torch.nn.ModuleList() 41 | 42 | for _, dim in enumerate(get_bond_feature_dims()): 43 | emb = torch.nn.Embedding(dim + 1, emb_dim, padding_idx=0) # for padding 44 | torch.nn.init.xavier_uniform_(emb.weight.data) 45 | self.bond_embedding_list.append(emb) 46 | 47 | def forward(self, edge_attr): 48 | bond_embedding = 0 49 | for i in range(edge_attr.shape[1]): 50 | bond_embedding += self.bond_embedding_list[i](edge_attr[:, i]) 51 | 52 | return bond_embedding 53 | 54 | 55 | class SineEncoding(nn.Module): 56 | def __init__(self, hidden_dim=128): 57 | super(SineEncoding, self).__init__() 58 | self.constant = 100 59 | self.hidden_dim = hidden_dim 60 | self.eig_w = nn.Linear(hidden_dim + 1, hidden_dim) 61 | 62 | def forward(self, e): 63 | # input: [B, N] 64 | # output: [B, N, d] 65 | 66 | ee = e * self.constant 67 | div = torch.exp(torch.arange(0, self.hidden_dim, 2) * (-math.log(10000)/self.hidden_dim)).to(e.device) 68 | pe = ee.unsqueeze(2) * div 69 | eeig = torch.cat((e.unsqueeze(2), torch.sin(pe), torch.cos(pe)), dim=2) 70 | 71 | return self.eig_w(eeig) 72 | 73 | 74 | class FeedForwardNetwork(nn.Module): 75 | 76 | def __init__(self, input_dim, hidden_dim, output_dim): 77 | super(FeedForwardNetwork, self).__init__() 78 | self.layer1 = nn.Linear(input_dim, hidden_dim) 79 | self.gelu = nn.GELU() 80 | self.layer2 = nn.Linear(hidden_dim, output_dim) 81 | 82 | def forward(self, x): 83 | x = self.layer1(x) 84 | x = self.gelu(x) 85 | x = self.layer2(x) 86 | return x 87 | 88 | 89 | class Conv(nn.Module): 90 | def __init__(self, hidden_size, dropout_rate): 91 | super(Conv, self).__init__() 92 | 93 | self.pre_ffn = nn.Sequential( 94 | nn.Linear(hidden_size, hidden_size), 95 | # nn.BatchNorm1d(hidden_size), 96 | nn.GELU() 97 | ) 98 | 99 | self.preffn_dropout = nn.Dropout(dropout_rate) 100 | self.ffn_dropout = nn.Dropout(dropout_rate) 101 | 102 | self.ffn = nn.Sequential( 103 | nn.Linear(hidden_size, hidden_size), 104 | nn.BatchNorm1d(hidden_size), 105 | nn.ReLU(), 106 | nn.Linear(hidden_size, hidden_size), 107 | nn.BatchNorm1d(hidden_size), 108 | nn.ReLU() 109 | ) 110 | 111 | def forward(self, graph, x_feat, edge_attr, bases): 112 | with graph.local_scope(): 113 | graph.ndata['x'] = x_feat 114 | graph.apply_edges(fn.copy_u('x', '_x')) 115 | xee = self.pre_ffn(graph.edata['_x'] + edge_attr) * bases 116 | graph.edata['v'] = xee 117 | graph.update_all(fn.copy_e('v', '_aggr_e'), fn.sum('_aggr_e', 'aggr_e')) 118 | y = graph.ndata['aggr_e'] 119 | y = self.preffn_dropout(y) 120 | x = x_feat + y 121 | y = self.ffn(x) 122 | y = self.ffn_dropout(y) 123 | x = x + y 124 | return x 125 | 126 | 127 | class SpecformerSmall(nn.Module): 128 | 129 | def __init__(self, nclass, nlayer, hidden_dim=128, nheads=4, feat_dropout=0.1, trans_dropout=0.1, adj_dropout=0.1): 130 | super(SpecformerSmall, self).__init__() 131 | 132 | print('small model') 133 | self.nlayer = nlayer 134 | self.nclass = nclass 135 | self.hidden_dim = hidden_dim 136 | self.nheads = nheads 137 | 138 | self.atom_encoder = AtomEncoder(hidden_dim) 139 | self.bond_encoder = BondEncoder(hidden_dim) 140 | 141 | self.eig_encoder = SineEncoding(hidden_dim) 142 | self.decoder = nn.Linear(hidden_dim, nheads) 143 | 144 | self.mha_norm = nn.LayerNorm(hidden_dim) 145 | self.ffn_norm = nn.LayerNorm(hidden_dim) 146 | self.mha_dropout = nn.Dropout(trans_dropout) 147 | self.ffn_dropout = nn.Dropout(trans_dropout) 148 | self.mha = nn.MultiheadAttention(hidden_dim, nheads, trans_dropout, batch_first=True) 149 | self.ffn = FeedForwardNetwork(hidden_dim, hidden_dim, hidden_dim) 150 | 151 | self.adj_dropout = nn.Dropout(adj_dropout) 152 | self.filter_encoder = nn.Sequential( 153 | nn.Linear(nheads + 1, hidden_dim), 154 | nn.BatchNorm1d(hidden_dim), 155 | nn.GELU(), 156 | nn.Linear(hidden_dim, hidden_dim), 157 | nn.BatchNorm1d(hidden_dim), 158 | nn.GELU(), 159 | ) 160 | 161 | self.convs = nn.ModuleList([Conv(hidden_dim, feat_dropout) for _ in range(nlayer)]) 162 | self.pool = AvgPooling() 163 | self.linear = nn.Linear(hidden_dim, nclass) 164 | 165 | 166 | def forward(self, e, u, g, length): 167 | 168 | # e: [B, N] eigenvalues 169 | # u: [B, N, N] eigenvectors 170 | # x: [B, N, d] node features 171 | # f: [B, N, N, d] edge features 172 | 173 | B, N = e.size() 174 | ut = u.transpose(1, 2) 175 | 176 | node_feat = g.ndata['feat'] 177 | edge_feat = g.edata['feat'] 178 | 179 | # do not use u to generate edge_idx because of the connected components 180 | e_mask, edge_idx = self.length_to_mask(length) 181 | 182 | node_feat = self.atom_encoder(node_feat) 183 | edge_feat = self.bond_encoder(edge_feat) 184 | eig = self.eig_encoder(e) 185 | 186 | mha_eig = self.mha_norm(eig) 187 | mha_eig, attn = self.mha(mha_eig, mha_eig, mha_eig, key_padding_mask=e_mask) 188 | eig = eig + self.mha_dropout(mha_eig) 189 | 190 | ffn_eig = self.ffn_norm(eig) 191 | ffn_eig = self.ffn(ffn_eig) 192 | eig = eig + self.ffn_dropout(ffn_eig) 193 | 194 | new_e = self.decoder(eig).transpose(2, 1) # [B, m, N] 195 | diag_e = torch.diag_embed(new_e) # [B, m, N, N] 196 | 197 | identity = torch.diag_embed(torch.ones_like(e)) 198 | bases = [identity] 199 | for i in range(self.nheads): 200 | filters = u @ diag_e[:, i, :, :] @ ut 201 | bases.append(filters) 202 | 203 | bases = torch.stack(bases, axis=-1) # [B, N, N, H] 204 | bases = bases[edge_idx] 205 | bases = self.adj_dropout(self.filter_encoder(bases)) 206 | bases = edge_softmax(g, bases) 207 | 208 | for conv in self.convs: 209 | node_feat = conv(g, node_feat, edge_feat, bases) 210 | 211 | h = self.pool(g, node_feat) 212 | h = self.linear(h) 213 | 214 | return h 215 | 216 | 217 | def length_to_mask(self, length): 218 | ''' 219 | length: [B] 220 | return: [B, max_len]. 221 | ''' 222 | B = len(length) 223 | N = length.max().item() 224 | mask1d = torch.arange(N, device=length.device).expand(B, N) >= length.unsqueeze(1) 225 | mask2d = (~mask1d).float().unsqueeze(2) @ (~mask1d).float().unsqueeze(1) 226 | mask2d = mask2d.bool() 227 | 228 | # Example 229 | # length=[1, 2, 3], B=3, N=3, 230 | 231 | # mask1d for key_padding_mask of MultiheadAttention [B, N] 232 | # [False, True, True ] 233 | # [False, False, True ] 234 | # [False, False, False] 235 | 236 | # mask2d for edge indexing [B, N, N] 237 | # [[1, 0, 0], | [1, 1, 0], | [1, 1, 1], 238 | # [0, 0, 0], | [1, 1, 0], | [1, 1, 1], 239 | # [0, 0, 0], | [0, 0, 0], | [1, 1, 1],] 240 | 241 | return mask1d, mask2d 242 | 243 | 244 | ''' 245 | def length_to_mask(self, length): 246 | ''' 247 | length: [B] 248 | return: [B, max_len]. 249 | ''' 250 | B = len(length) 251 | N = length.max().item() 252 | mask1d = torch.arange(N, device=length.device).expand(B, N) >= length.unsqueeze(1) 253 | 254 | mask2d = torch.zeros(B, N, N, device=length.device) 255 | for i in range(B): 256 | mask2d[i, :length[i], :length[i]] = 1.0 257 | 258 | # mask1d for key_padding_mask, a True value indicates that the corresponding key value will be ignored for the purpose of attention 259 | # mask2d for edge selection from padding 260 | return mask1d, mask2d.bool() 261 | ''' 262 | -------------------------------------------------------------------------------- /Graph/zinc_model.py: -------------------------------------------------------------------------------- 1 | import time 2 | import math 3 | import torch 4 | import torch.nn as nn 5 | import torch.nn.functional as F 6 | import numpy as np 7 | 8 | from dgl.ops.edge_softmax import edge_softmax 9 | from dgl.nn.pytorch.glob import AvgPooling 10 | from dgl import function as fn 11 | from torch.nn.init import xavier_uniform_, xavier_normal_, constant_ 12 | 13 | 14 | class SineEncoding(nn.Module): 15 | def __init__(self, hidden_dim=128): 16 | super(SineEncoding, self).__init__() 17 | self.constant = 100 18 | self.hidden_dim = hidden_dim 19 | self.eig_w = nn.Linear(hidden_dim + 1, hidden_dim) 20 | 21 | def forward(self, e): 22 | # input: [B, N] 23 | # output: [B, N, d] 24 | 25 | ee = e * self.constant 26 | div = torch.exp(torch.arange(0, self.hidden_dim, 2) * (-math.log(10000)/self.hidden_dim)).to(e.device) 27 | pe = ee.unsqueeze(2) * div 28 | eeig = torch.cat((e.unsqueeze(2), torch.sin(pe), torch.cos(pe)), dim=2) 29 | 30 | return self.eig_w(eeig) 31 | 32 | 33 | class FeedForwardNetwork(nn.Module): 34 | 35 | def __init__(self, input_dim, hidden_dim, output_dim): 36 | super(FeedForwardNetwork, self).__init__() 37 | self.layer1 = nn.Linear(input_dim, hidden_dim) 38 | self.gelu = nn.GELU() 39 | self.layer2 = nn.Linear(hidden_dim, output_dim) 40 | 41 | def forward(self, x): 42 | x = self.layer1(x) 43 | x = self.gelu(x) 44 | x = self.layer2(x) 45 | return x 46 | 47 | 48 | class Conv(nn.Module): 49 | def __init__(self, hidden_size, dropout_rate): 50 | super(Conv, self).__init__() 51 | 52 | self.pre_ffn = nn.Sequential( 53 | nn.Linear(hidden_size, hidden_size), 54 | # nn.BatchNorm1d(hidden_size), 55 | nn.GELU() 56 | ) 57 | 58 | self.preffn_dropout = nn.Dropout(dropout_rate) 59 | self.ffn_dropout = nn.Dropout(dropout_rate) 60 | 61 | self.ffn = nn.Sequential( 62 | nn.Linear(hidden_size, hidden_size), 63 | nn.BatchNorm1d(hidden_size), 64 | nn.ReLU(), 65 | nn.Linear(hidden_size, hidden_size), 66 | nn.BatchNorm1d(hidden_size), 67 | nn.ReLU() 68 | ) 69 | 70 | def forward(self, graph, x_feat, edge_attr, bases): 71 | with graph.local_scope(): 72 | graph.ndata['x'] = x_feat 73 | graph.apply_edges(fn.copy_u('x', '_x')) 74 | xee = self.pre_ffn(graph.edata['_x'] + edge_attr) * bases 75 | graph.edata['v'] = xee 76 | graph.update_all(fn.copy_e('v', '_aggr_e'), fn.sum('_aggr_e', 'aggr_e')) 77 | y = graph.ndata['aggr_e'] 78 | y = self.preffn_dropout(y) 79 | x = x_feat + y 80 | y = self.ffn(x) 81 | y = self.ffn_dropout(y) 82 | x = x + y 83 | return x 84 | 85 | 86 | class SpecformerZINC(nn.Module): 87 | 88 | def __init__(self, nclass, nlayer, hidden_dim=128, nheads=4, feat_dropout=0.1, trans_dropout=0.1, adj_dropout=0.1): 89 | super(SpecformerZINC, self).__init__() 90 | 91 | self.nlayer = nlayer 92 | self.nclass = nclass 93 | self.hidden_dim = hidden_dim 94 | self.nheads = nheads 95 | 96 | self.atom_encoder = nn.Embedding(40, hidden_dim) 97 | self.bond_encoder = nn.Embedding(10, hidden_dim, padding_idx=0) 98 | 99 | self.eig_encoder = SineEncoding(hidden_dim) 100 | self.decoder = nn.Linear(hidden_dim, nheads) 101 | 102 | self.mha_norm = nn.LayerNorm(hidden_dim) 103 | self.ffn_norm = nn.LayerNorm(hidden_dim) 104 | self.mha_dropout = nn.Dropout(trans_dropout) 105 | self.ffn_dropout = nn.Dropout(trans_dropout) 106 | self.mha = nn.MultiheadAttention(hidden_dim, nheads, trans_dropout, batch_first=True) 107 | self.ffn = FeedForwardNetwork(hidden_dim, hidden_dim, hidden_dim) 108 | 109 | self.adj_dropout = nn.Dropout(adj_dropout) 110 | self.filter_encoder = nn.Sequential( 111 | nn.Linear(nheads + 1, hidden_dim), 112 | nn.GELU(), 113 | nn.Linear(hidden_dim, hidden_dim), 114 | nn.GELU(), 115 | ) 116 | 117 | self.convs = nn.ModuleList([Conv(hidden_dim, feat_dropout) for _ in range(nlayer)]) 118 | self.pool = AvgPooling() 119 | self.linear = nn.Linear(hidden_dim, nclass) 120 | 121 | def forward(self, e, u, g, length): 122 | 123 | # e: [B, N] eigenvalues 124 | # u: [B, N, N] eigenvectors 125 | # x: [B, N, d] node features 126 | # f: [B, N, N, d] edge features 127 | # do not use u to generate edge_idx because of the existing of connected components 128 | 129 | B, N = e.size() 130 | ut = u.transpose(1, 2) 131 | 132 | node_feat = g.ndata['feat'] 133 | edge_feat = g.edata['feat'] 134 | 135 | eig_mask, edge_idx = self.length_to_mask(length) 136 | 137 | node_feat = self.atom_encoder(node_feat).squeeze(-2) 138 | edge_feat = self.bond_encoder(edge_feat).squeeze(-2) 139 | 140 | eig = self.eig_encoder(e) 141 | mha_eig = self.mha_norm(eig) 142 | mha_eig, attn = self.mha(mha_eig, mha_eig, mha_eig, key_padding_mask=eig_mask, average_attn_weights=False) 143 | eig = eig + self.mha_dropout(mha_eig) 144 | 145 | ffn_eig = self.ffn_norm(eig) 146 | ffn_eig = self.ffn(ffn_eig) 147 | eig = eig + self.ffn_dropout(ffn_eig) 148 | 149 | new_e = self.decoder(eig).transpose(2, 1) # [B, m, N] 150 | diag_e = torch.diag_embed(new_e) # [B, m, N, N] 151 | 152 | bases = [torch.diag_embed(torch.ones_like(e))] 153 | for i in range(self.nheads): 154 | filters = u @ diag_e[:, i, :, :] @ ut 155 | bases.append(filters) 156 | 157 | bases = torch.stack(bases, axis=-1) # [B, N, N, H] 158 | bases = bases[edge_idx] 159 | bases = self.adj_dropout(self.filter_encoder(bases)) 160 | 161 | for conv in self.convs: 162 | node_feat = conv(g, node_feat, edge_feat, bases) 163 | 164 | h = self.pool(g, node_feat) 165 | h = self.linear(h) 166 | 167 | return h, new_e, attn 168 | 169 | 170 | def length_to_mask(self, length): 171 | ''' 172 | length: [B] 173 | return: [B, max_len]. 174 | ''' 175 | B = len(length) 176 | N = length.max().item() 177 | mask1d = torch.arange(N, device=length.device).expand(B, N) >= length.unsqueeze(1) 178 | 179 | mask2d = torch.zeros(B, N, N, device=length.device) 180 | for i in range(B): 181 | mask2d[i, :length[i], :length[i]] = 1.0 182 | 183 | # mask1d for key_padding_mask, a True value indicates that the corresponding key value will be ignored for the purpose of attention 184 | # mask2d for edge selection from padding 185 | return mask1d, mask2d.bool() 186 | 187 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | GNU GENERAL PUBLIC LICENSE 2 | Version 3, 29 June 2007 3 | 4 | Copyright (C) 2007 Free Software Foundation, Inc. 5 | Everyone is permitted to copy and distribute verbatim copies 6 | of this license document, but changing it is not allowed. 7 | 8 | Preamble 9 | 10 | The GNU General Public License is a free, copyleft license for 11 | software and other kinds of works. 12 | 13 | The licenses for most software and other practical works are designed 14 | to take away your freedom to share and change the works. By contrast, 15 | the GNU General Public License is intended to guarantee your freedom to 16 | share and change all versions of a program--to make sure it remains free 17 | software for all its users. We, the Free Software Foundation, use the 18 | GNU General Public License for most of our software; it applies also to 19 | any other work released this way by its authors. You can apply it to 20 | your programs, too. 21 | 22 | When we speak of free software, we are referring to freedom, not 23 | price. Our General Public Licenses are designed to make sure that you 24 | have the freedom to distribute copies of free software (and charge for 25 | them if you wish), that you receive source code or can get it if you 26 | want it, that you can change the software or use pieces of it in new 27 | free programs, and that you know you can do these things. 28 | 29 | To protect your rights, we need to prevent others from denying you 30 | these rights or asking you to surrender the rights. Therefore, you have 31 | certain responsibilities if you distribute copies of the software, or if 32 | you modify it: responsibilities to respect the freedom of others. 33 | 34 | For example, if you distribute copies of such a program, whether 35 | gratis or for a fee, you must pass on to the recipients the same 36 | freedoms that you received. You must make sure that they, too, receive 37 | or can get the source code. And you must show them these terms so they 38 | know their rights. 39 | 40 | Developers that use the GNU GPL protect your rights with two steps: 41 | (1) assert copyright on the software, and (2) offer you this License 42 | giving you legal permission to copy, distribute and/or modify it. 43 | 44 | For the developers' and authors' protection, the GPL clearly explains 45 | that there is no warranty for this free software. For both users' and 46 | authors' sake, the GPL requires that modified versions be marked as 47 | changed, so that their problems will not be attributed erroneously to 48 | authors of previous versions. 49 | 50 | Some devices are designed to deny users access to install or run 51 | modified versions of the software inside them, although the manufacturer 52 | can do so. This is fundamentally incompatible with the aim of 53 | protecting users' freedom to change the software. The systematic 54 | pattern of such abuse occurs in the area of products for individuals to 55 | use, which is precisely where it is most unacceptable. Therefore, we 56 | have designed this version of the GPL to prohibit the practice for those 57 | products. If such problems arise substantially in other domains, we 58 | stand ready to extend this provision to those domains in future versions 59 | of the GPL, as needed to protect the freedom of users. 60 | 61 | Finally, every program is threatened constantly by software patents. 62 | States should not allow patents to restrict development and use of 63 | software on general-purpose computers, but in those that do, we wish to 64 | avoid the special danger that patents applied to a free program could 65 | make it effectively proprietary. To prevent this, the GPL assures that 66 | patents cannot be used to render the program non-free. 67 | 68 | The precise terms and conditions for copying, distribution and 69 | modification follow. 70 | 71 | TERMS AND CONDITIONS 72 | 73 | 0. Definitions. 74 | 75 | "This License" refers to version 3 of the GNU General Public License. 76 | 77 | "Copyright" also means copyright-like laws that apply to other kinds of 78 | works, such as semiconductor masks. 79 | 80 | "The Program" refers to any copyrightable work licensed under this 81 | License. Each licensee is addressed as "you". "Licensees" and 82 | "recipients" may be individuals or organizations. 83 | 84 | To "modify" a work means to copy from or adapt all or part of the work 85 | in a fashion requiring copyright permission, other than the making of an 86 | exact copy. The resulting work is called a "modified version" of the 87 | earlier work or a work "based on" the earlier work. 88 | 89 | A "covered work" means either the unmodified Program or a work based 90 | on the Program. 91 | 92 | To "propagate" a work means to do anything with it that, without 93 | permission, would make you directly or secondarily liable for 94 | infringement under applicable copyright law, except executing it on a 95 | computer or modifying a private copy. Propagation includes copying, 96 | distribution (with or without modification), making available to the 97 | public, and in some countries other activities as well. 98 | 99 | To "convey" a work means any kind of propagation that enables other 100 | parties to make or receive copies. Mere interaction with a user through 101 | a computer network, with no transfer of a copy, is not conveying. 102 | 103 | An interactive user interface displays "Appropriate Legal Notices" 104 | to the extent that it includes a convenient and prominently visible 105 | feature that (1) displays an appropriate copyright notice, and (2) 106 | tells the user that there is no warranty for the work (except to the 107 | extent that warranties are provided), that licensees may convey the 108 | work under this License, and how to view a copy of this License. If 109 | the interface presents a list of user commands or options, such as a 110 | menu, a prominent item in the list meets this criterion. 111 | 112 | 1. Source Code. 113 | 114 | The "source code" for a work means the preferred form of the work 115 | for making modifications to it. "Object code" means any non-source 116 | form of a work. 117 | 118 | A "Standard Interface" means an interface that either is an official 119 | standard defined by a recognized standards body, or, in the case of 120 | interfaces specified for a particular programming language, one that 121 | is widely used among developers working in that language. 122 | 123 | The "System Libraries" of an executable work include anything, other 124 | than the work as a whole, that (a) is included in the normal form of 125 | packaging a Major Component, but which is not part of that Major 126 | Component, and (b) serves only to enable use of the work with that 127 | Major Component, or to implement a Standard Interface for which an 128 | implementation is available to the public in source code form. A 129 | "Major Component", in this context, means a major essential component 130 | (kernel, window system, and so on) of the specific operating system 131 | (if any) on which the executable work runs, or a compiler used to 132 | produce the work, or an object code interpreter used to run it. 133 | 134 | The "Corresponding Source" for a work in object code form means all 135 | the source code needed to generate, install, and (for an executable 136 | work) run the object code and to modify the work, including scripts to 137 | control those activities. However, it does not include the work's 138 | System Libraries, or general-purpose tools or generally available free 139 | programs which are used unmodified in performing those activities but 140 | which are not part of the work. For example, Corresponding Source 141 | includes interface definition files associated with source files for 142 | the work, and the source code for shared libraries and dynamically 143 | linked subprograms that the work is specifically designed to require, 144 | such as by intimate data communication or control flow between those 145 | subprograms and other parts of the work. 146 | 147 | The Corresponding Source need not include anything that users 148 | can regenerate automatically from other parts of the Corresponding 149 | Source. 150 | 151 | The Corresponding Source for a work in source code form is that 152 | same work. 153 | 154 | 2. Basic Permissions. 155 | 156 | All rights granted under this License are granted for the term of 157 | copyright on the Program, and are irrevocable provided the stated 158 | conditions are met. This License explicitly affirms your unlimited 159 | permission to run the unmodified Program. The output from running a 160 | covered work is covered by this License only if the output, given its 161 | content, constitutes a covered work. This License acknowledges your 162 | rights of fair use or other equivalent, as provided by copyright law. 163 | 164 | You may make, run and propagate covered works that you do not 165 | convey, without conditions so long as your license otherwise remains 166 | in force. You may convey covered works to others for the sole purpose 167 | of having them make modifications exclusively for you, or provide you 168 | with facilities for running those works, provided that you comply with 169 | the terms of this License in conveying all material for which you do 170 | not control copyright. Those thus making or running the covered works 171 | for you must do so exclusively on your behalf, under your direction 172 | and control, on terms that prohibit them from making any copies of 173 | your copyrighted material outside their relationship with you. 174 | 175 | Conveying under any other circumstances is permitted solely under 176 | the conditions stated below. Sublicensing is not allowed; section 10 177 | makes it unnecessary. 178 | 179 | 3. Protecting Users' Legal Rights From Anti-Circumvention Law. 180 | 181 | No covered work shall be deemed part of an effective technological 182 | measure under any applicable law fulfilling obligations under article 183 | 11 of the WIPO copyright treaty adopted on 20 December 1996, or 184 | similar laws prohibiting or restricting circumvention of such 185 | measures. 186 | 187 | When you convey a covered work, you waive any legal power to forbid 188 | circumvention of technological measures to the extent such circumvention 189 | is effected by exercising rights under this License with respect to 190 | the covered work, and you disclaim any intention to limit operation or 191 | modification of the work as a means of enforcing, against the work's 192 | users, your or third parties' legal rights to forbid circumvention of 193 | technological measures. 194 | 195 | 4. Conveying Verbatim Copies. 196 | 197 | You may convey verbatim copies of the Program's source code as you 198 | receive it, in any medium, provided that you conspicuously and 199 | appropriately publish on each copy an appropriate copyright notice; 200 | keep intact all notices stating that this License and any 201 | non-permissive terms added in accord with section 7 apply to the code; 202 | keep intact all notices of the absence of any warranty; and give all 203 | recipients a copy of this License along with the Program. 204 | 205 | You may charge any price or no price for each copy that you convey, 206 | and you may offer support or warranty protection for a fee. 207 | 208 | 5. Conveying Modified Source Versions. 209 | 210 | You may convey a work based on the Program, or the modifications to 211 | produce it from the Program, in the form of source code under the 212 | terms of section 4, provided that you also meet all of these conditions: 213 | 214 | a) The work must carry prominent notices stating that you modified 215 | it, and giving a relevant date. 216 | 217 | b) The work must carry prominent notices stating that it is 218 | released under this License and any conditions added under section 219 | 7. This requirement modifies the requirement in section 4 to 220 | "keep intact all notices". 221 | 222 | c) You must license the entire work, as a whole, under this 223 | License to anyone who comes into possession of a copy. This 224 | License will therefore apply, along with any applicable section 7 225 | additional terms, to the whole of the work, and all its parts, 226 | regardless of how they are packaged. This License gives no 227 | permission to license the work in any other way, but it does not 228 | invalidate such permission if you have separately received it. 229 | 230 | d) If the work has interactive user interfaces, each must display 231 | Appropriate Legal Notices; however, if the Program has interactive 232 | interfaces that do not display Appropriate Legal Notices, your 233 | work need not make them do so. 234 | 235 | A compilation of a covered work with other separate and independent 236 | works, which are not by their nature extensions of the covered work, 237 | and which are not combined with it such as to form a larger program, 238 | in or on a volume of a storage or distribution medium, is called an 239 | "aggregate" if the compilation and its resulting copyright are not 240 | used to limit the access or legal rights of the compilation's users 241 | beyond what the individual works permit. Inclusion of a covered work 242 | in an aggregate does not cause this License to apply to the other 243 | parts of the aggregate. 244 | 245 | 6. Conveying Non-Source Forms. 246 | 247 | You may convey a covered work in object code form under the terms 248 | of sections 4 and 5, provided that you also convey the 249 | machine-readable Corresponding Source under the terms of this License, 250 | in one of these ways: 251 | 252 | a) Convey the object code in, or embodied in, a physical product 253 | (including a physical distribution medium), accompanied by the 254 | Corresponding Source fixed on a durable physical medium 255 | customarily used for software interchange. 256 | 257 | b) Convey the object code in, or embodied in, a physical product 258 | (including a physical distribution medium), accompanied by a 259 | written offer, valid for at least three years and valid for as 260 | long as you offer spare parts or customer support for that product 261 | model, to give anyone who possesses the object code either (1) a 262 | copy of the Corresponding Source for all the software in the 263 | product that is covered by this License, on a durable physical 264 | medium customarily used for software interchange, for a price no 265 | more than your reasonable cost of physically performing this 266 | conveying of source, or (2) access to copy the 267 | Corresponding Source from a network server at no charge. 268 | 269 | c) Convey individual copies of the object code with a copy of the 270 | written offer to provide the Corresponding Source. This 271 | alternative is allowed only occasionally and noncommercially, and 272 | only if you received the object code with such an offer, in accord 273 | with subsection 6b. 274 | 275 | d) Convey the object code by offering access from a designated 276 | place (gratis or for a charge), and offer equivalent access to the 277 | Corresponding Source in the same way through the same place at no 278 | further charge. You need not require recipients to copy the 279 | Corresponding Source along with the object code. If the place to 280 | copy the object code is a network server, the Corresponding Source 281 | may be on a different server (operated by you or a third party) 282 | that supports equivalent copying facilities, provided you maintain 283 | clear directions next to the object code saying where to find the 284 | Corresponding Source. Regardless of what server hosts the 285 | Corresponding Source, you remain obligated to ensure that it is 286 | available for as long as needed to satisfy these requirements. 287 | 288 | e) Convey the object code using peer-to-peer transmission, provided 289 | you inform other peers where the object code and Corresponding 290 | Source of the work are being offered to the general public at no 291 | charge under subsection 6d. 292 | 293 | A separable portion of the object code, whose source code is excluded 294 | from the Corresponding Source as a System Library, need not be 295 | included in conveying the object code work. 296 | 297 | A "User Product" is either (1) a "consumer product", which means any 298 | tangible personal property which is normally used for personal, family, 299 | or household purposes, or (2) anything designed or sold for incorporation 300 | into a dwelling. In determining whether a product is a consumer product, 301 | doubtful cases shall be resolved in favor of coverage. For a particular 302 | product received by a particular user, "normally used" refers to a 303 | typical or common use of that class of product, regardless of the status 304 | of the particular user or of the way in which the particular user 305 | actually uses, or expects or is expected to use, the product. A product 306 | is a consumer product regardless of whether the product has substantial 307 | commercial, industrial or non-consumer uses, unless such uses represent 308 | the only significant mode of use of the product. 309 | 310 | "Installation Information" for a User Product means any methods, 311 | procedures, authorization keys, or other information required to install 312 | and execute modified versions of a covered work in that User Product from 313 | a modified version of its Corresponding Source. The information must 314 | suffice to ensure that the continued functioning of the modified object 315 | code is in no case prevented or interfered with solely because 316 | modification has been made. 317 | 318 | If you convey an object code work under this section in, or with, or 319 | specifically for use in, a User Product, and the conveying occurs as 320 | part of a transaction in which the right of possession and use of the 321 | User Product is transferred to the recipient in perpetuity or for a 322 | fixed term (regardless of how the transaction is characterized), the 323 | Corresponding Source conveyed under this section must be accompanied 324 | by the Installation Information. But this requirement does not apply 325 | if neither you nor any third party retains the ability to install 326 | modified object code on the User Product (for example, the work has 327 | been installed in ROM). 328 | 329 | The requirement to provide Installation Information does not include a 330 | requirement to continue to provide support service, warranty, or updates 331 | for a work that has been modified or installed by the recipient, or for 332 | the User Product in which it has been modified or installed. Access to a 333 | network may be denied when the modification itself materially and 334 | adversely affects the operation of the network or violates the rules and 335 | protocols for communication across the network. 336 | 337 | Corresponding Source conveyed, and Installation Information provided, 338 | in accord with this section must be in a format that is publicly 339 | documented (and with an implementation available to the public in 340 | source code form), and must require no special password or key for 341 | unpacking, reading or copying. 342 | 343 | 7. Additional Terms. 344 | 345 | "Additional permissions" are terms that supplement the terms of this 346 | License by making exceptions from one or more of its conditions. 347 | Additional permissions that are applicable to the entire Program shall 348 | be treated as though they were included in this License, to the extent 349 | that they are valid under applicable law. If additional permissions 350 | apply only to part of the Program, that part may be used separately 351 | under those permissions, but the entire Program remains governed by 352 | this License without regard to the additional permissions. 353 | 354 | When you convey a copy of a covered work, you may at your option 355 | remove any additional permissions from that copy, or from any part of 356 | it. (Additional permissions may be written to require their own 357 | removal in certain cases when you modify the work.) You may place 358 | additional permissions on material, added by you to a covered work, 359 | for which you have or can give appropriate copyright permission. 360 | 361 | Notwithstanding any other provision of this License, for material you 362 | add to a covered work, you may (if authorized by the copyright holders of 363 | that material) supplement the terms of this License with terms: 364 | 365 | a) Disclaiming warranty or limiting liability differently from the 366 | terms of sections 15 and 16 of this License; or 367 | 368 | b) Requiring preservation of specified reasonable legal notices or 369 | author attributions in that material or in the Appropriate Legal 370 | Notices displayed by works containing it; or 371 | 372 | c) Prohibiting misrepresentation of the origin of that material, or 373 | requiring that modified versions of such material be marked in 374 | reasonable ways as different from the original version; or 375 | 376 | d) Limiting the use for publicity purposes of names of licensors or 377 | authors of the material; or 378 | 379 | e) Declining to grant rights under trademark law for use of some 380 | trade names, trademarks, or service marks; or 381 | 382 | f) Requiring indemnification of licensors and authors of that 383 | material by anyone who conveys the material (or modified versions of 384 | it) with contractual assumptions of liability to the recipient, for 385 | any liability that these contractual assumptions directly impose on 386 | those licensors and authors. 387 | 388 | All other non-permissive additional terms are considered "further 389 | restrictions" within the meaning of section 10. If the Program as you 390 | received it, or any part of it, contains a notice stating that it is 391 | governed by this License along with a term that is a further 392 | restriction, you may remove that term. If a license document contains 393 | a further restriction but permits relicensing or conveying under this 394 | License, you may add to a covered work material governed by the terms 395 | of that license document, provided that the further restriction does 396 | not survive such relicensing or conveying. 397 | 398 | If you add terms to a covered work in accord with this section, you 399 | must place, in the relevant source files, a statement of the 400 | additional terms that apply to those files, or a notice indicating 401 | where to find the applicable terms. 402 | 403 | Additional terms, permissive or non-permissive, may be stated in the 404 | form of a separately written license, or stated as exceptions; 405 | the above requirements apply either way. 406 | 407 | 8. Termination. 408 | 409 | You may not propagate or modify a covered work except as expressly 410 | provided under this License. Any attempt otherwise to propagate or 411 | modify it is void, and will automatically terminate your rights under 412 | this License (including any patent licenses granted under the third 413 | paragraph of section 11). 414 | 415 | However, if you cease all violation of this License, then your 416 | license from a particular copyright holder is reinstated (a) 417 | provisionally, unless and until the copyright holder explicitly and 418 | finally terminates your license, and (b) permanently, if the copyright 419 | holder fails to notify you of the violation by some reasonable means 420 | prior to 60 days after the cessation. 421 | 422 | Moreover, your license from a particular copyright holder is 423 | reinstated permanently if the copyright holder notifies you of the 424 | violation by some reasonable means, this is the first time you have 425 | received notice of violation of this License (for any work) from that 426 | copyright holder, and you cure the violation prior to 30 days after 427 | your receipt of the notice. 428 | 429 | Termination of your rights under this section does not terminate the 430 | licenses of parties who have received copies or rights from you under 431 | this License. If your rights have been terminated and not permanently 432 | reinstated, you do not qualify to receive new licenses for the same 433 | material under section 10. 434 | 435 | 9. Acceptance Not Required for Having Copies. 436 | 437 | You are not required to accept this License in order to receive or 438 | run a copy of the Program. Ancillary propagation of a covered work 439 | occurring solely as a consequence of using peer-to-peer transmission 440 | to receive a copy likewise does not require acceptance. However, 441 | nothing other than this License grants you permission to propagate or 442 | modify any covered work. These actions infringe copyright if you do 443 | not accept this License. Therefore, by modifying or propagating a 444 | covered work, you indicate your acceptance of this License to do so. 445 | 446 | 10. Automatic Licensing of Downstream Recipients. 447 | 448 | Each time you convey a covered work, the recipient automatically 449 | receives a license from the original licensors, to run, modify and 450 | propagate that work, subject to this License. You are not responsible 451 | for enforcing compliance by third parties with this License. 452 | 453 | An "entity transaction" is a transaction transferring control of an 454 | organization, or substantially all assets of one, or subdividing an 455 | organization, or merging organizations. If propagation of a covered 456 | work results from an entity transaction, each party to that 457 | transaction who receives a copy of the work also receives whatever 458 | licenses to the work the party's predecessor in interest had or could 459 | give under the previous paragraph, plus a right to possession of the 460 | Corresponding Source of the work from the predecessor in interest, if 461 | the predecessor has it or can get it with reasonable efforts. 462 | 463 | You may not impose any further restrictions on the exercise of the 464 | rights granted or affirmed under this License. For example, you may 465 | not impose a license fee, royalty, or other charge for exercise of 466 | rights granted under this License, and you may not initiate litigation 467 | (including a cross-claim or counterclaim in a lawsuit) alleging that 468 | any patent claim is infringed by making, using, selling, offering for 469 | sale, or importing the Program or any portion of it. 470 | 471 | 11. Patents. 472 | 473 | A "contributor" is a copyright holder who authorizes use under this 474 | License of the Program or a work on which the Program is based. The 475 | work thus licensed is called the contributor's "contributor version". 476 | 477 | A contributor's "essential patent claims" are all patent claims 478 | owned or controlled by the contributor, whether already acquired or 479 | hereafter acquired, that would be infringed by some manner, permitted 480 | by this License, of making, using, or selling its contributor version, 481 | but do not include claims that would be infringed only as a 482 | consequence of further modification of the contributor version. For 483 | purposes of this definition, "control" includes the right to grant 484 | patent sublicenses in a manner consistent with the requirements of 485 | this License. 486 | 487 | Each contributor grants you a non-exclusive, worldwide, royalty-free 488 | patent license under the contributor's essential patent claims, to 489 | make, use, sell, offer for sale, import and otherwise run, modify and 490 | propagate the contents of its contributor version. 491 | 492 | In the following three paragraphs, a "patent license" is any express 493 | agreement or commitment, however denominated, not to enforce a patent 494 | (such as an express permission to practice a patent or covenant not to 495 | sue for patent infringement). To "grant" such a patent license to a 496 | party means to make such an agreement or commitment not to enforce a 497 | patent against the party. 498 | 499 | If you convey a covered work, knowingly relying on a patent license, 500 | and the Corresponding Source of the work is not available for anyone 501 | to copy, free of charge and under the terms of this License, through a 502 | publicly available network server or other readily accessible means, 503 | then you must either (1) cause the Corresponding Source to be so 504 | available, or (2) arrange to deprive yourself of the benefit of the 505 | patent license for this particular work, or (3) arrange, in a manner 506 | consistent with the requirements of this License, to extend the patent 507 | license to downstream recipients. "Knowingly relying" means you have 508 | actual knowledge that, but for the patent license, your conveying the 509 | covered work in a country, or your recipient's use of the covered work 510 | in a country, would infringe one or more identifiable patents in that 511 | country that you have reason to believe are valid. 512 | 513 | If, pursuant to or in connection with a single transaction or 514 | arrangement, you convey, or propagate by procuring conveyance of, a 515 | covered work, and grant a patent license to some of the parties 516 | receiving the covered work authorizing them to use, propagate, modify 517 | or convey a specific copy of the covered work, then the patent license 518 | you grant is automatically extended to all recipients of the covered 519 | work and works based on it. 520 | 521 | A patent license is "discriminatory" if it does not include within 522 | the scope of its coverage, prohibits the exercise of, or is 523 | conditioned on the non-exercise of one or more of the rights that are 524 | specifically granted under this License. You may not convey a covered 525 | work if you are a party to an arrangement with a third party that is 526 | in the business of distributing software, under which you make payment 527 | to the third party based on the extent of your activity of conveying 528 | the work, and under which the third party grants, to any of the 529 | parties who would receive the covered work from you, a discriminatory 530 | patent license (a) in connection with copies of the covered work 531 | conveyed by you (or copies made from those copies), or (b) primarily 532 | for and in connection with specific products or compilations that 533 | contain the covered work, unless you entered into that arrangement, 534 | or that patent license was granted, prior to 28 March 2007. 535 | 536 | Nothing in this License shall be construed as excluding or limiting 537 | any implied license or other defenses to infringement that may 538 | otherwise be available to you under applicable patent law. 539 | 540 | 12. No Surrender of Others' Freedom. 541 | 542 | If conditions are imposed on you (whether by court order, agreement or 543 | otherwise) that contradict the conditions of this License, they do not 544 | excuse you from the conditions of this License. If you cannot convey a 545 | covered work so as to satisfy simultaneously your obligations under this 546 | License and any other pertinent obligations, then as a consequence you may 547 | not convey it at all. For example, if you agree to terms that obligate you 548 | to collect a royalty for further conveying from those to whom you convey 549 | the Program, the only way you could satisfy both those terms and this 550 | License would be to refrain entirely from conveying the Program. 551 | 552 | 13. Use with the GNU Affero General Public License. 553 | 554 | Notwithstanding any other provision of this License, you have 555 | permission to link or combine any covered work with a work licensed 556 | under version 3 of the GNU Affero General Public License into a single 557 | combined work, and to convey the resulting work. The terms of this 558 | License will continue to apply to the part which is the covered work, 559 | but the special requirements of the GNU Affero General Public License, 560 | section 13, concerning interaction through a network will apply to the 561 | combination as such. 562 | 563 | 14. Revised Versions of this License. 564 | 565 | The Free Software Foundation may publish revised and/or new versions of 566 | the GNU General Public License from time to time. Such new versions will 567 | be similar in spirit to the present version, but may differ in detail to 568 | address new problems or concerns. 569 | 570 | Each version is given a distinguishing version number. If the 571 | Program specifies that a certain numbered version of the GNU General 572 | Public License "or any later version" applies to it, you have the 573 | option of following the terms and conditions either of that numbered 574 | version or of any later version published by the Free Software 575 | Foundation. If the Program does not specify a version number of the 576 | GNU General Public License, you may choose any version ever published 577 | by the Free Software Foundation. 578 | 579 | If the Program specifies that a proxy can decide which future 580 | versions of the GNU General Public License can be used, that proxy's 581 | public statement of acceptance of a version permanently authorizes you 582 | to choose that version for the Program. 583 | 584 | Later license versions may give you additional or different 585 | permissions. However, no additional obligations are imposed on any 586 | author or copyright holder as a result of your choosing to follow a 587 | later version. 588 | 589 | 15. Disclaimer of Warranty. 590 | 591 | THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY 592 | APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT 593 | HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY 594 | OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, 595 | THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 596 | PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM 597 | IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF 598 | ALL NECESSARY SERVICING, REPAIR OR CORRECTION. 599 | 600 | 16. Limitation of Liability. 601 | 602 | IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING 603 | WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS 604 | THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY 605 | GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE 606 | USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF 607 | DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD 608 | PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), 609 | EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF 610 | SUCH DAMAGES. 611 | 612 | 17. Interpretation of Sections 15 and 16. 613 | 614 | If the disclaimer of warranty and limitation of liability provided 615 | above cannot be given local legal effect according to their terms, 616 | reviewing courts shall apply local law that most closely approximates 617 | an absolute waiver of all civil liability in connection with the 618 | Program, unless a warranty or assumption of liability accompanies a 619 | copy of the Program in return for a fee. 620 | 621 | END OF TERMS AND CONDITIONS 622 | 623 | How to Apply These Terms to Your New Programs 624 | 625 | If you develop a new program, and you want it to be of the greatest 626 | possible use to the public, the best way to achieve this is to make it 627 | free software which everyone can redistribute and change under these terms. 628 | 629 | To do so, attach the following notices to the program. It is safest 630 | to attach them to the start of each source file to most effectively 631 | state the exclusion of warranty; and each file should have at least 632 | the "copyright" line and a pointer to where the full notice is found. 633 | 634 | 635 | Copyright (C) 636 | 637 | This program is free software: you can redistribute it and/or modify 638 | it under the terms of the GNU General Public License as published by 639 | the Free Software Foundation, either version 3 of the License, or 640 | (at your option) any later version. 641 | 642 | This program is distributed in the hope that it will be useful, 643 | but WITHOUT ANY WARRANTY; without even the implied warranty of 644 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 645 | GNU General Public License for more details. 646 | 647 | You should have received a copy of the GNU General Public License 648 | along with this program. If not, see . 649 | 650 | Also add information on how to contact you by electronic and paper mail. 651 | 652 | If the program does terminal interaction, make it output a short 653 | notice like this when it starts in an interactive mode: 654 | 655 | Copyright (C) 656 | This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'. 657 | This is free software, and you are welcome to redistribute it 658 | under certain conditions; type `show c' for details. 659 | 660 | The hypothetical commands `show w' and `show c' should show the appropriate 661 | parts of the General Public License. Of course, your program's commands 662 | might be different; for a GUI interface, you would use an "about box". 663 | 664 | You should also get your employer (if you work as a programmer) or school, 665 | if any, to sign a "copyright disclaimer" for the program, if necessary. 666 | For more information on this, and how to apply and follow the GNU GPL, see 667 | . 668 | 669 | The GNU General Public License does not permit incorporating your program 670 | into proprietary programs. If your program is a subroutine library, you 671 | may consider it more useful to permit linking proprietary applications with 672 | the library. If this is what you want to do, use the GNU Lesser General 673 | Public License instead of this License. But first, please read 674 | . 675 | -------------------------------------------------------------------------------- /Node/config.yaml: -------------------------------------------------------------------------------- 1 | signal: 2 | nclass: 1 3 | nlayer: 1 4 | num_heads: 1 5 | hidden_dim: 16 6 | epoch: 2000 7 | lr: 0.01 8 | weight_decay: 0.0 9 | tran_dropout: 0.0 10 | feat_dropout: 0.0 11 | prop_dropout: 0.0 12 | norm: 'none' 13 | cora: 14 | nclass: 7 15 | nlayer: 2 16 | num_heads: 2 17 | hidden_dim: 32 18 | epoch: 2000 19 | lr: 0.0002 20 | weight_decay: 0.0001 21 | tran_dropout: 0.2 22 | feat_dropout: 0.6 23 | prop_dropout: 0.2 24 | norm: 'none' 25 | citeseer: 26 | nclass: 6 27 | nlayer: 2 28 | num_heads: 2 29 | hidden_dim: 32 30 | epoch: 2000 31 | lr: 0.0002 32 | weight_decay: 0.001 33 | tran_dropout: 0.0 34 | feat_dropout: 0.7 35 | prop_dropout: 0.5 36 | norm: 'none' 37 | photo: 38 | nclass: 8 39 | nlayer: 2 40 | num_heads: 4 41 | hidden_dim: 32 42 | epoch: 2000 43 | lr: 0.0002 44 | weight_decay: 0.0001 45 | tran_dropout: 0.2 46 | feat_dropout: 0.3 47 | prop_dropout: 0.2 48 | norm: 'none' 49 | arxiv: 50 | nclass: 40 51 | nlayer: 1 52 | num_heads: 1 53 | hidden_dim: 512 54 | epoch: 2000 55 | lr: 0.001 56 | weight_decay: 0.0 57 | tran_dropout: 0.1 58 | feat_dropout: 0.1 59 | prop_dropout: 0.1 60 | norm: 'layer' 61 | chameleon: 62 | nclass: 5 63 | nlayer: 2 64 | num_heads: 4 65 | hidden_dim: 32 66 | epoch: 2000 67 | lr: 0.001 68 | weight_decay: 0.0005 69 | tran_dropout: 0.2 70 | feat_dropout: 0.4 71 | prop_dropout: 0.5 72 | norm: 'none' 73 | squirrel: 74 | nclass: 5 75 | nlayer: 2 76 | num_heads: 2 77 | hidden_dim: 32 78 | epoch: 2000 79 | lr: 0.001 80 | weight_decay: 0.001 81 | tran_dropout: 0.1 82 | feat_dropout: 0.4 83 | prop_dropout: 0.4 84 | norm: 'none' 85 | actor: 86 | nclass: 5 87 | nlayer: 2 88 | num_heads: 1 89 | hidden_dim: 32 90 | epoch: 2000 91 | lr: 0.0002 92 | weight_decay: 0.0001 93 | tran_dropout: 0.5 94 | feat_dropout: 0.8 95 | prop_dropout: 0.5 96 | norm: 'none' 97 | penn: 98 | nclass: 2 99 | nlayer: 1 100 | num_heads: 1 101 | hidden_dim: 64 102 | epoch: 2000 103 | lr: 0.001 104 | weight_decay: 0.001 105 | tran_dropout: 0.0 106 | feat_dropout: 0.4 107 | prop_dropout: 0.4 108 | norm: 'batch' 109 | -------------------------------------------------------------------------------- /Node/data/chameleon.pt.zip: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/data/chameleon.pt.zip -------------------------------------------------------------------------------- /Node/main_node.py: -------------------------------------------------------------------------------- 1 | import time 2 | import yaml 3 | import copy 4 | import math 5 | import random 6 | import argparse 7 | import numpy as np 8 | import torch 9 | import torch.nn as nn 10 | import torch.nn.functional as F 11 | import torchmetrics 12 | from sklearn.metrics import roc_auc_score, mean_absolute_error, accuracy_score, r2_score 13 | from model_node import Specformer 14 | from utils import count_parameters, init_params, seed_everything, get_split 15 | 16 | 17 | def main_worker(args, config): 18 | print(args, config) 19 | seed_everything(args.seed) 20 | device = 'cuda:{}'.format(args.cuda) 21 | torch.cuda.set_device(args.seed) 22 | 23 | epoch = config['epoch'] 24 | lr = config['lr'] 25 | weight_decay = config['weight_decay'] 26 | nclass = config['nclass'] 27 | nlayer = config['nlayer'] 28 | hidden_dim = config['hidden_dim'] 29 | num_heads = config['num_heads'] 30 | tran_dropout = config['tran_dropout'] 31 | feat_dropout = config['feat_dropout'] 32 | prop_dropout = config['prop_dropout'] 33 | norm = config['norm'] 34 | 35 | if 'signal' in args.dataset: 36 | e, u, x, y, m = torch.load('data/{}.pt'.format(args.dataset)) 37 | e, u, x, y, m = e.cuda(), u.cuda(), x.cuda(), y.cuda(), m.cuda() 38 | mask = torch.where(m == 1) 39 | x = x[:, args.image].unsqueeze(1) 40 | y = y[:, args.image] 41 | else: 42 | e, u, x, y = torch.load('data/{}.pt'.format(args.dataset)) 43 | e, u, x, y = e.cuda(), u.cuda(), x.cuda(), y.cuda() 44 | 45 | if len(y.size()) > 1: 46 | if y.size(1) > 1: 47 | y = torch.argmax(y, dim=1) 48 | else: 49 | y = y.view(-1) 50 | 51 | train, valid, test = get_split(args.dataset, y, nclass, args.seed) 52 | train, valid, test = map(torch.LongTensor, (train, valid, test)) 53 | train, valid, test = train.cuda(), valid.cuda(), test.cuda() 54 | 55 | nfeat = x.size(1) 56 | net = Specformer(nclass, nfeat, nlayer, hidden_dim, num_heads, tran_dropout, feat_dropout, prop_dropout, norm).cuda() 57 | net.apply(init_params) 58 | optimizer = torch.optim.Adam(net.parameters(), lr=lr, weight_decay=weight_decay) 59 | print(count_parameters(net)) 60 | 61 | res = [] 62 | min_loss = 100.0 63 | max_acc = 0 64 | counter = 0 65 | evaluation = torchmetrics.Accuracy(task='multiclass', num_classes=nclass) 66 | 67 | for idx in range(epoch): 68 | 69 | net.train() 70 | optimizer.zero_grad() 71 | logits = net(e, u, x) 72 | 73 | if 'signal' in args.dataset: 74 | logits = logits.view(y.size()) 75 | loss = torch.square((logits[mask] - y[mask])).sum() 76 | else: 77 | loss = F.cross_entropy(logits[train], y[train]) 78 | 79 | loss.backward() 80 | optimizer.step() 81 | 82 | net.eval() 83 | logits = net(e, u, x) 84 | 85 | if 'signal' in args.dataset: 86 | logits = logits.view(y.size()) 87 | r2 = r2_score(y[mask].data.cpu().numpy(), logits[mask].data.cpu().numpy()) 88 | sse = torch.square(logits[mask] - y[mask]).sum().item() 89 | print(r2, sse) 90 | else: 91 | val_loss = F.cross_entropy(logits[valid], y[valid]).item() 92 | 93 | val_acc = evaluation(logits[valid].cpu(), y[valid].cpu()).item() 94 | test_acc = evaluation(logits[test].cpu(), y[test].cpu()).item() 95 | res.append([val_loss, val_acc, test_acc]) 96 | 97 | print(idx, val_loss, val_acc, test_acc) 98 | 99 | if val_loss < min_loss: 100 | min_loss = val_loss 101 | counter = 0 102 | else: 103 | counter += 1 104 | 105 | if counter == 200: 106 | max_acc1 = sorted(res, key=lambda x: x[0], reverse=False)[0][-1] 107 | max_acc2 = sorted(res, key=lambda x: x[1], reverse=True)[0][-1] 108 | print(max_acc1, max_acc2) 109 | break 110 | 111 | 112 | if __name__ == '__main__': 113 | parser = argparse.ArgumentParser() 114 | parser.add_argument('--seed', type=int, default=1) 115 | parser.add_argument('--cuda', type=int, default=0) 116 | parser.add_argument('--dataset', default='cora') 117 | parser.add_argument('--image', type=int, default=0) 118 | 119 | args = parser.parse_args() 120 | 121 | if 'signal' in args.dataset: 122 | config = yaml.load(open('config.yaml'), Loader=yaml.SafeLoader)['signal'] 123 | else: 124 | config = yaml.load(open('config.yaml'), Loader=yaml.SafeLoader)[args.dataset] 125 | 126 | main_worker(args, config) 127 | 128 | -------------------------------------------------------------------------------- /Node/master.csv: -------------------------------------------------------------------------------- 1 | ,ogbn-proteins,ogbn-products,ogbn-arxiv,ogbn-mag,ogbn-papers100M 2 | num tasks,112,1,1,1,1 3 | num classes,2,47,40,349,172 4 | eval metric,rocauc,acc,acc,acc,acc 5 | task type,binary classification,multiclass classification,multiclass classification,multiclass classification,multiclass classification 6 | download_name,proteins,products,arxiv,mag,papers100M-bin 7 | version,1,1,1,2,1 8 | url,http://snap.stanford.edu/ogb/data/nodeproppred/proteins.zip,http://snap.stanford.edu/ogb/data/nodeproppred/products.zip,http://snap.stanford.edu/ogb/data/nodeproppred/arxiv.zip,http://snap.stanford.edu/ogb/data/nodeproppred/mag.zip,http://snap.stanford.edu/ogb/data/nodeproppred/papers100M-bin.zip 9 | add_inverse_edge,True,True,False,False,False 10 | has_node_attr,False,True,True,True,True 11 | has_edge_attr,True,False,False,False,False 12 | split,species,sales_ranking,time,time,time 13 | additional node files,node_species,None,node_year,node_year,node_year 14 | additional edge files,None,None,None,edge_reltype,None 15 | is hetero,False,False,False,True,False 16 | binary,False,False,False,False,True 17 | -------------------------------------------------------------------------------- /Node/model_node.py: -------------------------------------------------------------------------------- 1 | import time 2 | import math 3 | import random 4 | import torch 5 | import torch.nn as nn 6 | import torch.nn.functional as F 7 | import numpy as np 8 | from torch.nn.init import xavier_uniform_, xavier_normal_, constant_ 9 | 10 | 11 | class SineEncoding(nn.Module): 12 | def __init__(self, hidden_dim=128): 13 | super(SineEncoding, self).__init__() 14 | self.constant = 100 15 | self.hidden_dim = hidden_dim 16 | self.eig_w = nn.Linear(hidden_dim + 1, hidden_dim) 17 | 18 | def forward(self, e): 19 | # input: [N] 20 | # output: [N, d] 21 | 22 | ee = e * self.constant 23 | div = torch.exp(torch.arange(0, self.hidden_dim, 2) * (-math.log(10000)/self.hidden_dim)).to(e.device) 24 | pe = ee.unsqueeze(1) * div 25 | eeig = torch.cat((e.unsqueeze(1), torch.sin(pe), torch.cos(pe)), dim=1) 26 | 27 | return self.eig_w(eeig) 28 | 29 | 30 | class FeedForwardNetwork(nn.Module): 31 | 32 | def __init__(self, input_dim, hidden_dim, output_dim): 33 | super(FeedForwardNetwork, self).__init__() 34 | self.layer1 = nn.Linear(input_dim, hidden_dim) 35 | self.gelu = nn.GELU() 36 | self.layer2 = nn.Linear(hidden_dim, output_dim) 37 | 38 | def forward(self, x): 39 | x = self.layer1(x) 40 | x = self.gelu(x) 41 | x = self.layer2(x) 42 | return x 43 | 44 | 45 | class SpecLayer(nn.Module): 46 | 47 | def __init__(self, nbases, ncombines, prop_dropout=0.0, norm='none'): 48 | super(SpecLayer, self).__init__() 49 | self.prop_dropout = nn.Dropout(prop_dropout) 50 | 51 | if norm == 'none': 52 | self.weight = nn.Parameter(torch.ones((1, nbases, ncombines))) 53 | else: 54 | self.weight = nn.Parameter(torch.empty((1, nbases, ncombines))) 55 | nn.init.normal_(self.weight, mean=0.0, std=0.01) 56 | 57 | if norm == 'layer': # Arxiv 58 | self.norm = nn.LayerNorm(ncombines) 59 | elif norm == 'batch': # Penn 60 | self.norm = nn.BatchNorm1d(ncombines) 61 | else: # Others 62 | self.norm = None 63 | 64 | def forward(self, x): 65 | x = self.prop_dropout(x) * self.weight # [N, m, d] * [1, m, d] 66 | x = torch.sum(x, dim=1) 67 | 68 | if self.norm is not None: 69 | x = self.norm(x) 70 | x = F.relu(x) 71 | 72 | return x 73 | 74 | 75 | class Specformer(nn.Module): 76 | 77 | def __init__(self, nclass, nfeat, nlayer=1, hidden_dim=128, nheads=1, 78 | tran_dropout=0.0, feat_dropout=0.0, prop_dropout=0.0, norm='none'): 79 | super(Specformer, self).__init__() 80 | 81 | self.norm = norm 82 | self.nfeat = nfeat 83 | self.nlayer = nlayer 84 | self.nheads = nheads 85 | self.hidden_dim = hidden_dim 86 | 87 | self.feat_encoder = nn.Sequential( 88 | nn.Linear(nfeat, hidden_dim), 89 | nn.ReLU(), 90 | nn.Linear(hidden_dim, nclass), 91 | ) 92 | 93 | # for arxiv & penn 94 | self.linear_encoder = nn.Linear(nfeat, hidden_dim) 95 | self.classify = nn.Linear(hidden_dim, nclass) 96 | 97 | self.eig_encoder = SineEncoding(hidden_dim) 98 | self.decoder = nn.Linear(hidden_dim, nheads) 99 | 100 | self.mha_norm = nn.LayerNorm(hidden_dim) 101 | self.ffn_norm = nn.LayerNorm(hidden_dim) 102 | self.mha_dropout = nn.Dropout(tran_dropout) 103 | self.ffn_dropout = nn.Dropout(tran_dropout) 104 | self.mha = nn.MultiheadAttention(hidden_dim, nheads, tran_dropout) 105 | self.ffn = FeedForwardNetwork(hidden_dim, hidden_dim, hidden_dim) 106 | 107 | self.feat_dp1 = nn.Dropout(feat_dropout) 108 | self.feat_dp2 = nn.Dropout(feat_dropout) 109 | if norm == 'none': 110 | self.layers = nn.ModuleList([SpecLayer(nheads+1, nclass, prop_dropout, norm=norm) for i in range(nlayer)]) 111 | else: 112 | self.layers = nn.ModuleList([SpecLayer(nheads+1, hidden_dim, prop_dropout, norm=norm) for i in range(nlayer)]) 113 | 114 | 115 | def forward(self, e, u, x): 116 | N = e.size(0) 117 | ut = u.permute(1, 0) 118 | 119 | if self.norm == 'none': 120 | h = self.feat_dp1(x) 121 | h = self.feat_encoder(h) 122 | h = self.feat_dp2(h) 123 | else: 124 | h = self.feat_dp1(x) 125 | h = self.linear_encoder(h) 126 | 127 | eig = self.eig_encoder(e) # [N, d] 128 | 129 | mha_eig = self.mha_norm(eig) 130 | mha_eig, attn = self.mha(mha_eig, mha_eig, mha_eig) 131 | eig = eig + self.mha_dropout(mha_eig) 132 | 133 | ffn_eig = self.ffn_norm(eig) 134 | ffn_eig = self.ffn(ffn_eig) 135 | eig = eig + self.ffn_dropout(ffn_eig) 136 | 137 | new_e = self.decoder(eig) # [N, m] 138 | 139 | for conv in self.layers: 140 | basic_feats = [h] 141 | utx = ut @ h 142 | for i in range(self.nheads): 143 | basic_feats.append(u @ (new_e[:, i].unsqueeze(1) * utx)) # [N, d] 144 | basic_feats = torch.stack(basic_feats, axis=1) # [N, m, d] 145 | h = conv(basic_feats) 146 | 147 | if self.norm == 'none': 148 | return h 149 | else: 150 | h = self.feat_dp2(h) 151 | h = self.classify(h) 152 | return h 153 | 154 | -------------------------------------------------------------------------------- /Node/node_raw_data/2Dgrid.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/2Dgrid.mat -------------------------------------------------------------------------------- /Node/node_raw_data/Penn94.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/Penn94.mat -------------------------------------------------------------------------------- /Node/node_raw_data/amazon_electronics_photo.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/amazon_electronics_photo.npz -------------------------------------------------------------------------------- /Node/node_raw_data/citeseer/ind.citeseer.allx: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/citeseer/ind.citeseer.allx -------------------------------------------------------------------------------- /Node/node_raw_data/citeseer/ind.citeseer.ally: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/citeseer/ind.citeseer.ally -------------------------------------------------------------------------------- /Node/node_raw_data/citeseer/ind.citeseer.graph: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/citeseer/ind.citeseer.graph -------------------------------------------------------------------------------- /Node/node_raw_data/citeseer/ind.citeseer.test.index: -------------------------------------------------------------------------------- 1 | 2488 2 | 2644 3 | 3261 4 | 2804 5 | 3176 6 | 2432 7 | 3310 8 | 2410 9 | 2812 10 | 2520 11 | 2994 12 | 3282 13 | 2680 14 | 2848 15 | 2670 16 | 3005 17 | 2977 18 | 2592 19 | 2967 20 | 2461 21 | 3184 22 | 2852 23 | 2768 24 | 2905 25 | 2851 26 | 3129 27 | 3164 28 | 2438 29 | 2793 30 | 2763 31 | 2528 32 | 2954 33 | 2347 34 | 2640 35 | 3265 36 | 2874 37 | 2446 38 | 2856 39 | 3149 40 | 2374 41 | 3097 42 | 3301 43 | 2664 44 | 2418 45 | 2655 46 | 2464 47 | 2596 48 | 3262 49 | 3278 50 | 2320 51 | 2612 52 | 2614 53 | 2550 54 | 2626 55 | 2772 56 | 3007 57 | 2733 58 | 2516 59 | 2476 60 | 2798 61 | 2561 62 | 2839 63 | 2685 64 | 2391 65 | 2705 66 | 3098 67 | 2754 68 | 3251 69 | 2767 70 | 2630 71 | 2727 72 | 2513 73 | 2701 74 | 3264 75 | 2792 76 | 2821 77 | 3260 78 | 2462 79 | 3307 80 | 2639 81 | 2900 82 | 3060 83 | 2672 84 | 3116 85 | 2731 86 | 3316 87 | 2386 88 | 2425 89 | 2518 90 | 3151 91 | 2586 92 | 2797 93 | 2479 94 | 3117 95 | 2580 96 | 3182 97 | 2459 98 | 2508 99 | 3052 100 | 3230 101 | 3215 102 | 2803 103 | 2969 104 | 2562 105 | 2398 106 | 3325 107 | 2343 108 | 3030 109 | 2414 110 | 2776 111 | 2383 112 | 3173 113 | 2850 114 | 2499 115 | 3312 116 | 2648 117 | 2784 118 | 2898 119 | 3056 120 | 2484 121 | 3179 122 | 3132 123 | 2577 124 | 2563 125 | 2867 126 | 3317 127 | 2355 128 | 3207 129 | 3178 130 | 2968 131 | 3319 132 | 2358 133 | 2764 134 | 3001 135 | 2683 136 | 3271 137 | 2321 138 | 2567 139 | 2502 140 | 3246 141 | 2715 142 | 3066 143 | 2390 144 | 2381 145 | 3162 146 | 2741 147 | 2498 148 | 2790 149 | 3038 150 | 3321 151 | 2481 152 | 3050 153 | 3161 154 | 3122 155 | 2801 156 | 2957 157 | 3177 158 | 2965 159 | 2621 160 | 3208 161 | 2921 162 | 2802 163 | 2357 164 | 2677 165 | 2519 166 | 2860 167 | 2696 168 | 2368 169 | 3241 170 | 2858 171 | 2419 172 | 2762 173 | 2875 174 | 3222 175 | 3064 176 | 2827 177 | 3044 178 | 2471 179 | 3062 180 | 2982 181 | 2736 182 | 2322 183 | 2709 184 | 2766 185 | 2424 186 | 2602 187 | 2970 188 | 2675 189 | 3299 190 | 2554 191 | 2964 192 | 2597 193 | 2753 194 | 2979 195 | 2523 196 | 2912 197 | 2896 198 | 2317 199 | 3167 200 | 2813 201 | 2482 202 | 2557 203 | 3043 204 | 3244 205 | 2985 206 | 2460 207 | 2363 208 | 3272 209 | 3045 210 | 3192 211 | 2453 212 | 2656 213 | 2834 214 | 2443 215 | 3202 216 | 2926 217 | 2711 218 | 2633 219 | 2384 220 | 2752 221 | 3285 222 | 2817 223 | 2483 224 | 2919 225 | 2924 226 | 2661 227 | 2698 228 | 2361 229 | 2662 230 | 2819 231 | 3143 232 | 2316 233 | 3196 234 | 2739 235 | 2345 236 | 2578 237 | 2822 238 | 3229 239 | 2908 240 | 2917 241 | 2692 242 | 3200 243 | 2324 244 | 2522 245 | 3322 246 | 2697 247 | 3163 248 | 3093 249 | 3233 250 | 2774 251 | 2371 252 | 2835 253 | 2652 254 | 2539 255 | 2843 256 | 3231 257 | 2976 258 | 2429 259 | 2367 260 | 3144 261 | 2564 262 | 3283 263 | 3217 264 | 3035 265 | 2962 266 | 2433 267 | 2415 268 | 2387 269 | 3021 270 | 2595 271 | 2517 272 | 2468 273 | 3061 274 | 2673 275 | 2348 276 | 3027 277 | 2467 278 | 3318 279 | 2959 280 | 3273 281 | 2392 282 | 2779 283 | 2678 284 | 3004 285 | 2634 286 | 2974 287 | 3198 288 | 2342 289 | 2376 290 | 3249 291 | 2868 292 | 2952 293 | 2710 294 | 2838 295 | 2335 296 | 2524 297 | 2650 298 | 3186 299 | 2743 300 | 2545 301 | 2841 302 | 2515 303 | 2505 304 | 3181 305 | 2945 306 | 2738 307 | 2933 308 | 3303 309 | 2611 310 | 3090 311 | 2328 312 | 3010 313 | 3016 314 | 2504 315 | 2936 316 | 3266 317 | 3253 318 | 2840 319 | 3034 320 | 2581 321 | 2344 322 | 2452 323 | 2654 324 | 3199 325 | 3137 326 | 2514 327 | 2394 328 | 2544 329 | 2641 330 | 2613 331 | 2618 332 | 2558 333 | 2593 334 | 2532 335 | 2512 336 | 2975 337 | 3267 338 | 2566 339 | 2951 340 | 3300 341 | 2869 342 | 2629 343 | 2747 344 | 3055 345 | 2831 346 | 3105 347 | 3168 348 | 3100 349 | 2431 350 | 2828 351 | 2684 352 | 3269 353 | 2910 354 | 2865 355 | 2693 356 | 2884 357 | 3228 358 | 2783 359 | 3247 360 | 2770 361 | 3157 362 | 2421 363 | 2382 364 | 2331 365 | 3203 366 | 3240 367 | 2351 368 | 3114 369 | 2986 370 | 2688 371 | 2439 372 | 2996 373 | 3079 374 | 3103 375 | 3296 376 | 2349 377 | 2372 378 | 3096 379 | 2422 380 | 2551 381 | 3069 382 | 2737 383 | 3084 384 | 3304 385 | 3022 386 | 2542 387 | 3204 388 | 2949 389 | 2318 390 | 2450 391 | 3140 392 | 2734 393 | 2881 394 | 2576 395 | 3054 396 | 3089 397 | 3125 398 | 2761 399 | 3136 400 | 3111 401 | 2427 402 | 2466 403 | 3101 404 | 3104 405 | 3259 406 | 2534 407 | 2961 408 | 3191 409 | 3000 410 | 3036 411 | 2356 412 | 2800 413 | 3155 414 | 3224 415 | 2646 416 | 2735 417 | 3020 418 | 2866 419 | 2426 420 | 2448 421 | 3226 422 | 3219 423 | 2749 424 | 3183 425 | 2906 426 | 2360 427 | 2440 428 | 2946 429 | 2313 430 | 2859 431 | 2340 432 | 3008 433 | 2719 434 | 3058 435 | 2653 436 | 3023 437 | 2888 438 | 3243 439 | 2913 440 | 3242 441 | 3067 442 | 2409 443 | 3227 444 | 2380 445 | 2353 446 | 2686 447 | 2971 448 | 2847 449 | 2947 450 | 2857 451 | 3263 452 | 3218 453 | 2861 454 | 3323 455 | 2635 456 | 2966 457 | 2604 458 | 2456 459 | 2832 460 | 2694 461 | 3245 462 | 3119 463 | 2942 464 | 3153 465 | 2894 466 | 2555 467 | 3128 468 | 2703 469 | 2323 470 | 2631 471 | 2732 472 | 2699 473 | 2314 474 | 2590 475 | 3127 476 | 2891 477 | 2873 478 | 2814 479 | 2326 480 | 3026 481 | 3288 482 | 3095 483 | 2706 484 | 2457 485 | 2377 486 | 2620 487 | 2526 488 | 2674 489 | 3190 490 | 2923 491 | 3032 492 | 2334 493 | 3254 494 | 2991 495 | 3277 496 | 2973 497 | 2599 498 | 2658 499 | 2636 500 | 2826 501 | 3148 502 | 2958 503 | 3258 504 | 2990 505 | 3180 506 | 2538 507 | 2748 508 | 2625 509 | 2565 510 | 3011 511 | 3057 512 | 2354 513 | 3158 514 | 2622 515 | 3308 516 | 2983 517 | 2560 518 | 3169 519 | 3059 520 | 2480 521 | 3194 522 | 3291 523 | 3216 524 | 2643 525 | 3172 526 | 2352 527 | 2724 528 | 2485 529 | 2411 530 | 2948 531 | 2445 532 | 2362 533 | 2668 534 | 3275 535 | 3107 536 | 2496 537 | 2529 538 | 2700 539 | 2541 540 | 3028 541 | 2879 542 | 2660 543 | 3324 544 | 2755 545 | 2436 546 | 3048 547 | 2623 548 | 2920 549 | 3040 550 | 2568 551 | 3221 552 | 3003 553 | 3295 554 | 2473 555 | 3232 556 | 3213 557 | 2823 558 | 2897 559 | 2573 560 | 2645 561 | 3018 562 | 3326 563 | 2795 564 | 2915 565 | 3109 566 | 3086 567 | 2463 568 | 3118 569 | 2671 570 | 2909 571 | 2393 572 | 2325 573 | 3029 574 | 2972 575 | 3110 576 | 2870 577 | 3284 578 | 2816 579 | 2647 580 | 2667 581 | 2955 582 | 2333 583 | 2960 584 | 2864 585 | 2893 586 | 2458 587 | 2441 588 | 2359 589 | 2327 590 | 3256 591 | 3099 592 | 3073 593 | 3138 594 | 2511 595 | 2666 596 | 2548 597 | 2364 598 | 2451 599 | 2911 600 | 3237 601 | 3206 602 | 3080 603 | 3279 604 | 2934 605 | 2981 606 | 2878 607 | 3130 608 | 2830 609 | 3091 610 | 2659 611 | 2449 612 | 3152 613 | 2413 614 | 2722 615 | 2796 616 | 3220 617 | 2751 618 | 2935 619 | 3238 620 | 2491 621 | 2730 622 | 2842 623 | 3223 624 | 2492 625 | 3074 626 | 3094 627 | 2833 628 | 2521 629 | 2883 630 | 3315 631 | 2845 632 | 2907 633 | 3083 634 | 2572 635 | 3092 636 | 2903 637 | 2918 638 | 3039 639 | 3286 640 | 2587 641 | 3068 642 | 2338 643 | 3166 644 | 3134 645 | 2455 646 | 2497 647 | 2992 648 | 2775 649 | 2681 650 | 2430 651 | 2932 652 | 2931 653 | 2434 654 | 3154 655 | 3046 656 | 2598 657 | 2366 658 | 3015 659 | 3147 660 | 2944 661 | 2582 662 | 3274 663 | 2987 664 | 2642 665 | 2547 666 | 2420 667 | 2930 668 | 2750 669 | 2417 670 | 2808 671 | 3141 672 | 2997 673 | 2995 674 | 2584 675 | 2312 676 | 3033 677 | 3070 678 | 3065 679 | 2509 680 | 3314 681 | 2396 682 | 2543 683 | 2423 684 | 3170 685 | 2389 686 | 3289 687 | 2728 688 | 2540 689 | 2437 690 | 2486 691 | 2895 692 | 3017 693 | 2853 694 | 2406 695 | 2346 696 | 2877 697 | 2472 698 | 3210 699 | 2637 700 | 2927 701 | 2789 702 | 2330 703 | 3088 704 | 3102 705 | 2616 706 | 3081 707 | 2902 708 | 3205 709 | 3320 710 | 3165 711 | 2984 712 | 3185 713 | 2707 714 | 3255 715 | 2583 716 | 2773 717 | 2742 718 | 3024 719 | 2402 720 | 2718 721 | 2882 722 | 2575 723 | 3281 724 | 2786 725 | 2855 726 | 3014 727 | 2401 728 | 2535 729 | 2687 730 | 2495 731 | 3113 732 | 2609 733 | 2559 734 | 2665 735 | 2530 736 | 3293 737 | 2399 738 | 2605 739 | 2690 740 | 3133 741 | 2799 742 | 2533 743 | 2695 744 | 2713 745 | 2886 746 | 2691 747 | 2549 748 | 3077 749 | 3002 750 | 3049 751 | 3051 752 | 3087 753 | 2444 754 | 3085 755 | 3135 756 | 2702 757 | 3211 758 | 3108 759 | 2501 760 | 2769 761 | 3290 762 | 2465 763 | 3025 764 | 3019 765 | 2385 766 | 2940 767 | 2657 768 | 2610 769 | 2525 770 | 2941 771 | 3078 772 | 2341 773 | 2916 774 | 2956 775 | 2375 776 | 2880 777 | 3009 778 | 2780 779 | 2370 780 | 2925 781 | 2332 782 | 3146 783 | 2315 784 | 2809 785 | 3145 786 | 3106 787 | 2782 788 | 2760 789 | 2493 790 | 2765 791 | 2556 792 | 2890 793 | 2400 794 | 2339 795 | 3201 796 | 2818 797 | 3248 798 | 3280 799 | 2570 800 | 2569 801 | 2937 802 | 3174 803 | 2836 804 | 2708 805 | 2820 806 | 3195 807 | 2617 808 | 3197 809 | 2319 810 | 2744 811 | 2615 812 | 2825 813 | 2603 814 | 2914 815 | 2531 816 | 3193 817 | 2624 818 | 2365 819 | 2810 820 | 3239 821 | 3159 822 | 2537 823 | 2844 824 | 2758 825 | 2938 826 | 3037 827 | 2503 828 | 3297 829 | 2885 830 | 2608 831 | 2494 832 | 2712 833 | 2408 834 | 2901 835 | 2704 836 | 2536 837 | 2373 838 | 2478 839 | 2723 840 | 3076 841 | 2627 842 | 2369 843 | 2669 844 | 3006 845 | 2628 846 | 2788 847 | 3276 848 | 2435 849 | 3139 850 | 3235 851 | 2527 852 | 2571 853 | 2815 854 | 2442 855 | 2892 856 | 2978 857 | 2746 858 | 3150 859 | 2574 860 | 2725 861 | 3188 862 | 2601 863 | 2378 864 | 3075 865 | 2632 866 | 2794 867 | 3270 868 | 3071 869 | 2506 870 | 3126 871 | 3236 872 | 3257 873 | 2824 874 | 2989 875 | 2950 876 | 2428 877 | 2405 878 | 3156 879 | 2447 880 | 2787 881 | 2805 882 | 2720 883 | 2403 884 | 2811 885 | 2329 886 | 2474 887 | 2785 888 | 2350 889 | 2507 890 | 2416 891 | 3112 892 | 2475 893 | 2876 894 | 2585 895 | 2487 896 | 3072 897 | 3082 898 | 2943 899 | 2757 900 | 2388 901 | 2600 902 | 3294 903 | 2756 904 | 3142 905 | 3041 906 | 2594 907 | 2998 908 | 3047 909 | 2379 910 | 2980 911 | 2454 912 | 2862 913 | 3175 914 | 2588 915 | 3031 916 | 3012 917 | 2889 918 | 2500 919 | 2791 920 | 2854 921 | 2619 922 | 2395 923 | 2807 924 | 2740 925 | 2412 926 | 3131 927 | 3013 928 | 2939 929 | 2651 930 | 2490 931 | 2988 932 | 2863 933 | 3225 934 | 2745 935 | 2714 936 | 3160 937 | 3124 938 | 2849 939 | 2676 940 | 2872 941 | 3287 942 | 3189 943 | 2716 944 | 3115 945 | 2928 946 | 2871 947 | 2591 948 | 2717 949 | 2546 950 | 2777 951 | 3298 952 | 2397 953 | 3187 954 | 2726 955 | 2336 956 | 3268 957 | 2477 958 | 2904 959 | 2846 960 | 3121 961 | 2899 962 | 2510 963 | 2806 964 | 2963 965 | 3313 966 | 2679 967 | 3302 968 | 2663 969 | 3053 970 | 2469 971 | 2999 972 | 3311 973 | 2470 974 | 2638 975 | 3120 976 | 3171 977 | 2689 978 | 2922 979 | 2607 980 | 2721 981 | 2993 982 | 2887 983 | 2837 984 | 2929 985 | 2829 986 | 3234 987 | 2649 988 | 2337 989 | 2759 990 | 2778 991 | 2771 992 | 2404 993 | 2589 994 | 3123 995 | 3209 996 | 2729 997 | 3252 998 | 2606 999 | 2579 1000 | 2552 1001 | -------------------------------------------------------------------------------- /Node/node_raw_data/citeseer/ind.citeseer.tx: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/citeseer/ind.citeseer.tx -------------------------------------------------------------------------------- /Node/node_raw_data/citeseer/ind.citeseer.ty: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/citeseer/ind.citeseer.ty -------------------------------------------------------------------------------- /Node/node_raw_data/citeseer/ind.citeseer.x: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/citeseer/ind.citeseer.x -------------------------------------------------------------------------------- /Node/node_raw_data/citeseer/ind.citeseer.y: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/citeseer/ind.citeseer.y -------------------------------------------------------------------------------- /Node/node_raw_data/cora/ind.cora.allx: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/cora/ind.cora.allx -------------------------------------------------------------------------------- /Node/node_raw_data/cora/ind.cora.ally: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/cora/ind.cora.ally -------------------------------------------------------------------------------- /Node/node_raw_data/cora/ind.cora.graph: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/cora/ind.cora.graph -------------------------------------------------------------------------------- /Node/node_raw_data/cora/ind.cora.test.index: -------------------------------------------------------------------------------- 1 | 2692 2 | 2532 3 | 2050 4 | 1715 5 | 2362 6 | 2609 7 | 2622 8 | 1975 9 | 2081 10 | 1767 11 | 2263 12 | 1725 13 | 2588 14 | 2259 15 | 2357 16 | 1998 17 | 2574 18 | 2179 19 | 2291 20 | 2382 21 | 1812 22 | 1751 23 | 2422 24 | 1937 25 | 2631 26 | 2510 27 | 2378 28 | 2589 29 | 2345 30 | 1943 31 | 1850 32 | 2298 33 | 1825 34 | 2035 35 | 2507 36 | 2313 37 | 1906 38 | 1797 39 | 2023 40 | 2159 41 | 2495 42 | 1886 43 | 2122 44 | 2369 45 | 2461 46 | 1925 47 | 2565 48 | 1858 49 | 2234 50 | 2000 51 | 1846 52 | 2318 53 | 1723 54 | 2559 55 | 2258 56 | 1763 57 | 1991 58 | 1922 59 | 2003 60 | 2662 61 | 2250 62 | 2064 63 | 2529 64 | 1888 65 | 2499 66 | 2454 67 | 2320 68 | 2287 69 | 2203 70 | 2018 71 | 2002 72 | 2632 73 | 2554 74 | 2314 75 | 2537 76 | 1760 77 | 2088 78 | 2086 79 | 2218 80 | 2605 81 | 1953 82 | 2403 83 | 1920 84 | 2015 85 | 2335 86 | 2535 87 | 1837 88 | 2009 89 | 1905 90 | 2636 91 | 1942 92 | 2193 93 | 2576 94 | 2373 95 | 1873 96 | 2463 97 | 2509 98 | 1954 99 | 2656 100 | 2455 101 | 2494 102 | 2295 103 | 2114 104 | 2561 105 | 2176 106 | 2275 107 | 2635 108 | 2442 109 | 2704 110 | 2127 111 | 2085 112 | 2214 113 | 2487 114 | 1739 115 | 2543 116 | 1783 117 | 2485 118 | 2262 119 | 2472 120 | 2326 121 | 1738 122 | 2170 123 | 2100 124 | 2384 125 | 2152 126 | 2647 127 | 2693 128 | 2376 129 | 1775 130 | 1726 131 | 2476 132 | 2195 133 | 1773 134 | 1793 135 | 2194 136 | 2581 137 | 1854 138 | 2524 139 | 1945 140 | 1781 141 | 1987 142 | 2599 143 | 1744 144 | 2225 145 | 2300 146 | 1928 147 | 2042 148 | 2202 149 | 1958 150 | 1816 151 | 1916 152 | 2679 153 | 2190 154 | 1733 155 | 2034 156 | 2643 157 | 2177 158 | 1883 159 | 1917 160 | 1996 161 | 2491 162 | 2268 163 | 2231 164 | 2471 165 | 1919 166 | 1909 167 | 2012 168 | 2522 169 | 1865 170 | 2466 171 | 2469 172 | 2087 173 | 2584 174 | 2563 175 | 1924 176 | 2143 177 | 1736 178 | 1966 179 | 2533 180 | 2490 181 | 2630 182 | 1973 183 | 2568 184 | 1978 185 | 2664 186 | 2633 187 | 2312 188 | 2178 189 | 1754 190 | 2307 191 | 2480 192 | 1960 193 | 1742 194 | 1962 195 | 2160 196 | 2070 197 | 2553 198 | 2433 199 | 1768 200 | 2659 201 | 2379 202 | 2271 203 | 1776 204 | 2153 205 | 1877 206 | 2027 207 | 2028 208 | 2155 209 | 2196 210 | 2483 211 | 2026 212 | 2158 213 | 2407 214 | 1821 215 | 2131 216 | 2676 217 | 2277 218 | 2489 219 | 2424 220 | 1963 221 | 1808 222 | 1859 223 | 2597 224 | 2548 225 | 2368 226 | 1817 227 | 2405 228 | 2413 229 | 2603 230 | 2350 231 | 2118 232 | 2329 233 | 1969 234 | 2577 235 | 2475 236 | 2467 237 | 2425 238 | 1769 239 | 2092 240 | 2044 241 | 2586 242 | 2608 243 | 1983 244 | 2109 245 | 2649 246 | 1964 247 | 2144 248 | 1902 249 | 2411 250 | 2508 251 | 2360 252 | 1721 253 | 2005 254 | 2014 255 | 2308 256 | 2646 257 | 1949 258 | 1830 259 | 2212 260 | 2596 261 | 1832 262 | 1735 263 | 1866 264 | 2695 265 | 1941 266 | 2546 267 | 2498 268 | 2686 269 | 2665 270 | 1784 271 | 2613 272 | 1970 273 | 2021 274 | 2211 275 | 2516 276 | 2185 277 | 2479 278 | 2699 279 | 2150 280 | 1990 281 | 2063 282 | 2075 283 | 1979 284 | 2094 285 | 1787 286 | 2571 287 | 2690 288 | 1926 289 | 2341 290 | 2566 291 | 1957 292 | 1709 293 | 1955 294 | 2570 295 | 2387 296 | 1811 297 | 2025 298 | 2447 299 | 2696 300 | 2052 301 | 2366 302 | 1857 303 | 2273 304 | 2245 305 | 2672 306 | 2133 307 | 2421 308 | 1929 309 | 2125 310 | 2319 311 | 2641 312 | 2167 313 | 2418 314 | 1765 315 | 1761 316 | 1828 317 | 2188 318 | 1972 319 | 1997 320 | 2419 321 | 2289 322 | 2296 323 | 2587 324 | 2051 325 | 2440 326 | 2053 327 | 2191 328 | 1923 329 | 2164 330 | 1861 331 | 2339 332 | 2333 333 | 2523 334 | 2670 335 | 2121 336 | 1921 337 | 1724 338 | 2253 339 | 2374 340 | 1940 341 | 2545 342 | 2301 343 | 2244 344 | 2156 345 | 1849 346 | 2551 347 | 2011 348 | 2279 349 | 2572 350 | 1757 351 | 2400 352 | 2569 353 | 2072 354 | 2526 355 | 2173 356 | 2069 357 | 2036 358 | 1819 359 | 1734 360 | 1880 361 | 2137 362 | 2408 363 | 2226 364 | 2604 365 | 1771 366 | 2698 367 | 2187 368 | 2060 369 | 1756 370 | 2201 371 | 2066 372 | 2439 373 | 1844 374 | 1772 375 | 2383 376 | 2398 377 | 1708 378 | 1992 379 | 1959 380 | 1794 381 | 2426 382 | 2702 383 | 2444 384 | 1944 385 | 1829 386 | 2660 387 | 2497 388 | 2607 389 | 2343 390 | 1730 391 | 2624 392 | 1790 393 | 1935 394 | 1967 395 | 2401 396 | 2255 397 | 2355 398 | 2348 399 | 1931 400 | 2183 401 | 2161 402 | 2701 403 | 1948 404 | 2501 405 | 2192 406 | 2404 407 | 2209 408 | 2331 409 | 1810 410 | 2363 411 | 2334 412 | 1887 413 | 2393 414 | 2557 415 | 1719 416 | 1732 417 | 1986 418 | 2037 419 | 2056 420 | 1867 421 | 2126 422 | 1932 423 | 2117 424 | 1807 425 | 1801 426 | 1743 427 | 2041 428 | 1843 429 | 2388 430 | 2221 431 | 1833 432 | 2677 433 | 1778 434 | 2661 435 | 2306 436 | 2394 437 | 2106 438 | 2430 439 | 2371 440 | 2606 441 | 2353 442 | 2269 443 | 2317 444 | 2645 445 | 2372 446 | 2550 447 | 2043 448 | 1968 449 | 2165 450 | 2310 451 | 1985 452 | 2446 453 | 1982 454 | 2377 455 | 2207 456 | 1818 457 | 1913 458 | 1766 459 | 1722 460 | 1894 461 | 2020 462 | 1881 463 | 2621 464 | 2409 465 | 2261 466 | 2458 467 | 2096 468 | 1712 469 | 2594 470 | 2293 471 | 2048 472 | 2359 473 | 1839 474 | 2392 475 | 2254 476 | 1911 477 | 2101 478 | 2367 479 | 1889 480 | 1753 481 | 2555 482 | 2246 483 | 2264 484 | 2010 485 | 2336 486 | 2651 487 | 2017 488 | 2140 489 | 1842 490 | 2019 491 | 1890 492 | 2525 493 | 2134 494 | 2492 495 | 2652 496 | 2040 497 | 2145 498 | 2575 499 | 2166 500 | 1999 501 | 2434 502 | 1711 503 | 2276 504 | 2450 505 | 2389 506 | 2669 507 | 2595 508 | 1814 509 | 2039 510 | 2502 511 | 1896 512 | 2168 513 | 2344 514 | 2637 515 | 2031 516 | 1977 517 | 2380 518 | 1936 519 | 2047 520 | 2460 521 | 2102 522 | 1745 523 | 2650 524 | 2046 525 | 2514 526 | 1980 527 | 2352 528 | 2113 529 | 1713 530 | 2058 531 | 2558 532 | 1718 533 | 1864 534 | 1876 535 | 2338 536 | 1879 537 | 1891 538 | 2186 539 | 2451 540 | 2181 541 | 2638 542 | 2644 543 | 2103 544 | 2591 545 | 2266 546 | 2468 547 | 1869 548 | 2582 549 | 2674 550 | 2361 551 | 2462 552 | 1748 553 | 2215 554 | 2615 555 | 2236 556 | 2248 557 | 2493 558 | 2342 559 | 2449 560 | 2274 561 | 1824 562 | 1852 563 | 1870 564 | 2441 565 | 2356 566 | 1835 567 | 2694 568 | 2602 569 | 2685 570 | 1893 571 | 2544 572 | 2536 573 | 1994 574 | 1853 575 | 1838 576 | 1786 577 | 1930 578 | 2539 579 | 1892 580 | 2265 581 | 2618 582 | 2486 583 | 2583 584 | 2061 585 | 1796 586 | 1806 587 | 2084 588 | 1933 589 | 2095 590 | 2136 591 | 2078 592 | 1884 593 | 2438 594 | 2286 595 | 2138 596 | 1750 597 | 2184 598 | 1799 599 | 2278 600 | 2410 601 | 2642 602 | 2435 603 | 1956 604 | 2399 605 | 1774 606 | 2129 607 | 1898 608 | 1823 609 | 1938 610 | 2299 611 | 1862 612 | 2420 613 | 2673 614 | 1984 615 | 2204 616 | 1717 617 | 2074 618 | 2213 619 | 2436 620 | 2297 621 | 2592 622 | 2667 623 | 2703 624 | 2511 625 | 1779 626 | 1782 627 | 2625 628 | 2365 629 | 2315 630 | 2381 631 | 1788 632 | 1714 633 | 2302 634 | 1927 635 | 2325 636 | 2506 637 | 2169 638 | 2328 639 | 2629 640 | 2128 641 | 2655 642 | 2282 643 | 2073 644 | 2395 645 | 2247 646 | 2521 647 | 2260 648 | 1868 649 | 1988 650 | 2324 651 | 2705 652 | 2541 653 | 1731 654 | 2681 655 | 2707 656 | 2465 657 | 1785 658 | 2149 659 | 2045 660 | 2505 661 | 2611 662 | 2217 663 | 2180 664 | 1904 665 | 2453 666 | 2484 667 | 1871 668 | 2309 669 | 2349 670 | 2482 671 | 2004 672 | 1965 673 | 2406 674 | 2162 675 | 1805 676 | 2654 677 | 2007 678 | 1947 679 | 1981 680 | 2112 681 | 2141 682 | 1720 683 | 1758 684 | 2080 685 | 2330 686 | 2030 687 | 2432 688 | 2089 689 | 2547 690 | 1820 691 | 1815 692 | 2675 693 | 1840 694 | 2658 695 | 2370 696 | 2251 697 | 1908 698 | 2029 699 | 2068 700 | 2513 701 | 2549 702 | 2267 703 | 2580 704 | 2327 705 | 2351 706 | 2111 707 | 2022 708 | 2321 709 | 2614 710 | 2252 711 | 2104 712 | 1822 713 | 2552 714 | 2243 715 | 1798 716 | 2396 717 | 2663 718 | 2564 719 | 2148 720 | 2562 721 | 2684 722 | 2001 723 | 2151 724 | 2706 725 | 2240 726 | 2474 727 | 2303 728 | 2634 729 | 2680 730 | 2055 731 | 2090 732 | 2503 733 | 2347 734 | 2402 735 | 2238 736 | 1950 737 | 2054 738 | 2016 739 | 1872 740 | 2233 741 | 1710 742 | 2032 743 | 2540 744 | 2628 745 | 1795 746 | 2616 747 | 1903 748 | 2531 749 | 2567 750 | 1946 751 | 1897 752 | 2222 753 | 2227 754 | 2627 755 | 1856 756 | 2464 757 | 2241 758 | 2481 759 | 2130 760 | 2311 761 | 2083 762 | 2223 763 | 2284 764 | 2235 765 | 2097 766 | 1752 767 | 2515 768 | 2527 769 | 2385 770 | 2189 771 | 2283 772 | 2182 773 | 2079 774 | 2375 775 | 2174 776 | 2437 777 | 1993 778 | 2517 779 | 2443 780 | 2224 781 | 2648 782 | 2171 783 | 2290 784 | 2542 785 | 2038 786 | 1855 787 | 1831 788 | 1759 789 | 1848 790 | 2445 791 | 1827 792 | 2429 793 | 2205 794 | 2598 795 | 2657 796 | 1728 797 | 2065 798 | 1918 799 | 2427 800 | 2573 801 | 2620 802 | 2292 803 | 1777 804 | 2008 805 | 1875 806 | 2288 807 | 2256 808 | 2033 809 | 2470 810 | 2585 811 | 2610 812 | 2082 813 | 2230 814 | 1915 815 | 1847 816 | 2337 817 | 2512 818 | 2386 819 | 2006 820 | 2653 821 | 2346 822 | 1951 823 | 2110 824 | 2639 825 | 2520 826 | 1939 827 | 2683 828 | 2139 829 | 2220 830 | 1910 831 | 2237 832 | 1900 833 | 1836 834 | 2197 835 | 1716 836 | 1860 837 | 2077 838 | 2519 839 | 2538 840 | 2323 841 | 1914 842 | 1971 843 | 1845 844 | 2132 845 | 1802 846 | 1907 847 | 2640 848 | 2496 849 | 2281 850 | 2198 851 | 2416 852 | 2285 853 | 1755 854 | 2431 855 | 2071 856 | 2249 857 | 2123 858 | 1727 859 | 2459 860 | 2304 861 | 2199 862 | 1791 863 | 1809 864 | 1780 865 | 2210 866 | 2417 867 | 1874 868 | 1878 869 | 2116 870 | 1961 871 | 1863 872 | 2579 873 | 2477 874 | 2228 875 | 2332 876 | 2578 877 | 2457 878 | 2024 879 | 1934 880 | 2316 881 | 1841 882 | 1764 883 | 1737 884 | 2322 885 | 2239 886 | 2294 887 | 1729 888 | 2488 889 | 1974 890 | 2473 891 | 2098 892 | 2612 893 | 1834 894 | 2340 895 | 2423 896 | 2175 897 | 2280 898 | 2617 899 | 2208 900 | 2560 901 | 1741 902 | 2600 903 | 2059 904 | 1747 905 | 2242 906 | 2700 907 | 2232 908 | 2057 909 | 2147 910 | 2682 911 | 1792 912 | 1826 913 | 2120 914 | 1895 915 | 2364 916 | 2163 917 | 1851 918 | 2391 919 | 2414 920 | 2452 921 | 1803 922 | 1989 923 | 2623 924 | 2200 925 | 2528 926 | 2415 927 | 1804 928 | 2146 929 | 2619 930 | 2687 931 | 1762 932 | 2172 933 | 2270 934 | 2678 935 | 2593 936 | 2448 937 | 1882 938 | 2257 939 | 2500 940 | 1899 941 | 2478 942 | 2412 943 | 2107 944 | 1746 945 | 2428 946 | 2115 947 | 1800 948 | 1901 949 | 2397 950 | 2530 951 | 1912 952 | 2108 953 | 2206 954 | 2091 955 | 1740 956 | 2219 957 | 1976 958 | 2099 959 | 2142 960 | 2671 961 | 2668 962 | 2216 963 | 2272 964 | 2229 965 | 2666 966 | 2456 967 | 2534 968 | 2697 969 | 2688 970 | 2062 971 | 2691 972 | 2689 973 | 2154 974 | 2590 975 | 2626 976 | 2390 977 | 1813 978 | 2067 979 | 1952 980 | 2518 981 | 2358 982 | 1789 983 | 2076 984 | 2049 985 | 2119 986 | 2013 987 | 2124 988 | 2556 989 | 2105 990 | 2093 991 | 1885 992 | 2305 993 | 2354 994 | 2135 995 | 2601 996 | 1770 997 | 1995 998 | 2504 999 | 1749 1000 | 2157 1001 | -------------------------------------------------------------------------------- /Node/node_raw_data/cora/ind.cora.tx: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/cora/ind.cora.tx -------------------------------------------------------------------------------- /Node/node_raw_data/cora/ind.cora.ty: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/cora/ind.cora.ty -------------------------------------------------------------------------------- /Node/node_raw_data/cora/ind.cora.x: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/cora/ind.cora.x -------------------------------------------------------------------------------- /Node/node_raw_data/cora/ind.cora.y: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/cora/ind.cora.y -------------------------------------------------------------------------------- /Node/node_raw_data/fb100-Penn94-splits.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/DSL-Lab/Specformer/40a35d0c4db0839c9d5e17f45ea7b4618e8fce71/Node/node_raw_data/fb100-Penn94-splits.npy -------------------------------------------------------------------------------- /Node/preprocess_node_data.py: -------------------------------------------------------------------------------- 1 | import sys 2 | import os 3 | import math 4 | import time 5 | import pickle as pkl 6 | import scipy as sp 7 | from scipy import io 8 | import numpy as np 9 | import pandas as pd 10 | import networkx as nx 11 | import dgl 12 | import torch 13 | from sklearn.preprocessing import label_binarize 14 | from ogb.nodeproppred.dataset_dgl import DglNodePropPredDataset 15 | from numpy.linalg import eig, eigh 16 | 17 | 18 | def generate_signal_data(): 19 | data = io.loadmat('node_raw_data/2Dgrid.mat') 20 | A = data['A'] 21 | x = data['F'].astype(np.float32) 22 | m = data['mask'] 23 | 24 | A = sp.sparse.coo_matrix(A).todense() 25 | 26 | D_vec = np.sum(A, axis=1).A1 27 | D_vec_invsqrt_corr = 1 / np.sqrt(D_vec) 28 | D_invsqrt_corr = np.diag(D_vec_invsqrt_corr) 29 | L = np.eye(10000) - D_invsqrt_corr @ A @ D_invsqrt_corr 30 | 31 | e, u = eigh(L) 32 | 33 | y_low = u @ np.diag(np.array([math.exp(-10*(ee-0)**2) for ee in e])) @ u.T @ x 34 | y_high = u @ np.diag(np.array([1 - math.exp(-10*(ee-0)**2) for ee in e])) @ u.T @ x 35 | y_band = u @ np.diag(np.array([math.exp(-10*(ee-1)**2) for ee in e])) @ u.T @ x 36 | y_rej = u @ np.diag(np.array([1 - math.exp(-10*(ee-1)**2) for ee in e])) @ u.T @ x 37 | y_comb = u @ np.diag(np.array([abs(np.sin(ee*math.pi)) for ee in e])) @ u.T @ x 38 | 39 | e = torch.FloatTensor(e) 40 | u = torch.FloatTensor(u) 41 | x = torch.FloatTensor(x) 42 | m = torch.LongTensor(m).squeeze() 43 | y_low = torch.FloatTensor(y_low) 44 | y_high = torch.FloatTensor(y_high) 45 | y_band = torch.FloatTensor(y_band) 46 | y_rej = torch.FloatTensor(y_rej) 47 | y_comb = torch.FloatTensor(y_comb) 48 | 49 | torch.save([e, u, x, y_low, m], 'data/signal_low.pt') 50 | torch.save([e, u, x, y_high, m], 'data/signal_high.pt') 51 | torch.save([e, u, x, y_band, m], 'data/signal_band.pt') 52 | torch.save([e, u, x, y_rej, m], 'data/signal_rej.pt') 53 | torch.save([e, u, x, y_comb, m], 'data/signal_comb.pt') 54 | 55 | 56 | def normalize_graph(g): 57 | g = np.array(g) 58 | g = g + g.T 59 | g[g > 0.] = 1.0 60 | deg = g.sum(axis=1).reshape(-1) 61 | deg[deg == 0.] = 1.0 62 | deg = np.diag(deg ** -0.5) 63 | adj = np.dot(np.dot(deg, g), deg) 64 | L = np.eye(g.shape[0]) - adj 65 | return L 66 | 67 | 68 | def eigen_decompositon(g): 69 | "The normalized (unit “length”) eigenvectors, " 70 | "such that the column v[:,i] is the eigenvector corresponding to the eigenvalue w[i]." 71 | g = normalize_graph(g) 72 | e, u = eigh(g) 73 | return e, u 74 | 75 | 76 | def parse_index_file(filename): 77 | """Parse index file.""" 78 | index = [] 79 | for line in open(filename): 80 | index.append(int(line.strip())) 81 | return index 82 | 83 | 84 | def feature_normalize(x): 85 | x = np.array(x) 86 | rowsum = x.sum(axis=1, keepdims=True) 87 | rowsum = np.clip(rowsum, 1, 1e10) 88 | return x / rowsum 89 | 90 | 91 | def load_data(dataset_str): 92 | names = ['x', 'y', 'tx', 'ty', 'allx', 'ally', 'graph'] 93 | objects = [] 94 | for i in range(len(names)): 95 | with open("node_raw_data/{}/ind.{}.{}".format(dataset_str, dataset_str, names[i]), 'rb') as f: 96 | if sys.version_info > (3, 0): 97 | objects.append(pkl.load(f, encoding='latin1')) 98 | else: 99 | objects.append(pkl.load(f)) 100 | 101 | x, y, tx, ty, allx, ally, graph = tuple(objects) 102 | test_idx_reorder = parse_index_file("node_raw_data/{}/ind.{}.test.index".format(dataset_str, dataset_str)) 103 | test_idx_range = np.sort(test_idx_reorder) 104 | 105 | if dataset_str == 'citeseer': 106 | # Fix citeseer dataset (there are some isolated nodes in the graph) 107 | # Find isolated nodes, add them as zero-vecs into the right position 108 | test_idx_range_full = range(min(test_idx_reorder), max(test_idx_reorder)+1) 109 | tx_extended = sp.sparse.lil_matrix((len(test_idx_range_full), x.shape[1])) 110 | tx_extended[test_idx_range-min(test_idx_range), :] = tx 111 | tx = tx_extended 112 | ty_extended = np.zeros((len(test_idx_range_full), y.shape[1])) 113 | ty_extended[test_idx_range-min(test_idx_range), :] = ty 114 | ty = ty_extended 115 | 116 | features = sp.sparse.vstack((allx, tx)).tolil() 117 | features[test_idx_reorder, :] = features[test_idx_range, :] 118 | adj = nx.adjacency_matrix(nx.from_dict_of_lists(graph)) 119 | 120 | labels = np.vstack((ally, ty)) 121 | labels[test_idx_reorder, :] = labels[test_idx_range, :] 122 | 123 | return adj, features, labels 124 | 125 | 126 | def eig_dgl_adj_sparse(g, sm=0, lm=0): 127 | A = g.adj(scipy_fmt='csr') 128 | deg = np.array(A.sum(axis=0)).flatten() 129 | D_ = sp.sparse.diags(deg ** -0.5) 130 | 131 | A_ = D_.dot(A.dot(D_)) 132 | L_ = sp.sparse.eye(g.num_nodes()) - A_ 133 | 134 | if sm > 0: 135 | e1, u1 = sp.sparse.linalg.eigsh(L_, k=sm, which='SM', tol=1e-5) 136 | e1, u1 = map(torch.FloatTensor, (e1, u1)) 137 | 138 | if lm > 0: 139 | e2, u2 = sp.sparse.linalg.eigsh(L_, k=lm, which='LM', tol=1e-5) 140 | e2, u2 = map(torch.FloatTensor, (e2, u2)) 141 | 142 | if sm > 0 and lm > 0: 143 | return torch.cat((e1, e2), dim=0), torch.cat((u1, u2), dim=1) 144 | elif sm > 0: 145 | return e1, u1 146 | elif lm > 0: 147 | return e2, u2 148 | else: 149 | pass 150 | 151 | 152 | def load_fb100_dataset(): 153 | mat = io.loadmat('node_raw_data/Penn94.mat') 154 | A = mat['A'] 155 | metadata = mat['local_info'] 156 | 157 | edge_index = A.nonzero() 158 | metadata = metadata.astype(int) 159 | label = metadata[:, 1] - 1 # gender label, -1 means unlabeled 160 | 161 | # make features into one-hot encodings 162 | feature_vals = np.hstack((np.expand_dims(metadata[:, 0], 1), metadata[:, 2:])) 163 | features = np.empty((A.shape[0], 0)) 164 | for col in range(feature_vals.shape[1]): 165 | feat_col = feature_vals[:, col] 166 | feat_onehot = label_binarize(feat_col, classes=np.unique(feat_col)) 167 | features = np.hstack((features, feat_onehot)) 168 | 169 | node_feat = torch.tensor(features, dtype=torch.float) 170 | num_nodes = metadata.shape[0] 171 | label = torch.LongTensor(label) 172 | 173 | g = dgl.graph((edge_index[0], edge_index[1]), num_nodes=num_nodes) 174 | 175 | return g, node_feat, label 176 | 177 | 178 | def generate_node_data(dataset): 179 | 180 | if dataset in ['cora', 'citeseer']: 181 | 182 | adj, x, y = load_data(dataset) 183 | adj = adj.todense() 184 | x = x.todense() 185 | x = feature_normalize(x) 186 | e, u = eigen_decompositon(adj) 187 | 188 | e = torch.FloatTensor(e) 189 | u = torch.FloatTensor(u) 190 | x = torch.FloatTensor(x) 191 | y = torch.LongTensor(y) 192 | 193 | torch.save([e, u, x, y], 'data/{}.pt'.format(dataset)) 194 | 195 | elif dataset in ['photo']: 196 | data = np.load('node_raw_data/amazon_electronics_photo.npz', allow_pickle=True) 197 | adj = sp.sparse.csr_matrix((data['adj_data'], data['adj_indices'], data['adj_indptr']), 198 | shape=data['adj_shape']).toarray() 199 | feat = sp.sparse.csr_matrix((data['attr_data'], data['attr_indices'], data['attr_indptr']), 200 | shape=data['attr_shape']).toarray() 201 | x = feature_normalize(feat) 202 | y = data['labels'] 203 | e, u = eigen_decompositon(adj) 204 | 205 | e = torch.FloatTensor(e) 206 | u = torch.FloatTensor(u) 207 | x = torch.FloatTensor(x) 208 | y = torch.LongTensor(y) 209 | 210 | torch.save([e, u, x, y], 'data/{}.pt'.format(dataset)) 211 | 212 | elif dataset in ['arxiv']: 213 | data = DglNodePropPredDataset('ogbn-arxiv') 214 | g = data[0][0] 215 | g = dgl.add_reverse_edges(g) 216 | g = dgl.to_simple(g) 217 | 218 | e, u = eig_dgl_adj_sparse(g, sm=5000) 219 | x = g.ndata['feat'] 220 | y = data[0][1] 221 | 222 | torch.save([e, u, x, y], 'data/arxiv.pt') 223 | 224 | elif dataset in ['penn']: 225 | g, x, y = load_fb100_dataset() 226 | g = dgl.add_reverse_edges(g) 227 | g = dgl.to_simple(g) 228 | 229 | e, u = eig_dgl_adj_sparse(g, sm=3000, lm=3000) 230 | 231 | torch.save([e, u, x, y], 'data/penn.pt') 232 | 233 | elif dataset in ['chameleon', 'squirrel', 'actor']: 234 | edge_df = pd.read_csv('node_raw_data/{}/'.format(dataset) + 'out1_graph_edges.txt', sep='\t') 235 | node_df = pd.read_csv('node_raw_data/{}/'.format(dataset) + 'out1_node_feature_label.txt', sep='\t') 236 | feature = node_df[node_df.columns[1]] 237 | y = node_df[node_df.columns[2]] 238 | 239 | num_nodes = len(y) 240 | adj = np.zeros((num_nodes, num_nodes)) 241 | 242 | source = list(edge_df[edge_df.columns[0]]) 243 | target = list(edge_df[edge_df.columns[1]]) 244 | 245 | for i in range(len(source)): 246 | adj[source[i], target[i]] = 1. 247 | adj[target[i], source[i]] = 1. 248 | 249 | if dataset == 'actor': 250 | # for sparse features 251 | nfeat = 932 252 | x = np.zeros((len(y), nfeat)) 253 | 254 | feature = list(feature) 255 | feature = [feat.split(',') for feat in feature] 256 | for ind, feat in enumerate(feature): 257 | for ff in feat: 258 | x[ind, int(ff)] = 1. 259 | 260 | x = feature_normalize(x) 261 | else: 262 | feature = list(feature) 263 | feature = [feat.split(',') for feat in feature] 264 | new_feat = [] 265 | 266 | for feat in feature: 267 | new_feat.append([int(f) for f in feat]) 268 | x = np.array(new_feat) 269 | x = feature_normalize(x) 270 | 271 | e, u = eigen_decompositon(adj) 272 | 273 | e = torch.FloatTensor(e) 274 | u = torch.FloatTensor(u) 275 | x = torch.FloatTensor(x) 276 | y = torch.LongTensor(y) 277 | 278 | torch.save([e, u, x, y], 'data/{}.pt'.format(dataset)) 279 | 280 | 281 | if __name__ == '__main__': 282 | #generate_node_data('cora') 283 | #generate_node_data('citeseer') 284 | #generate_node_data('photo') 285 | #generate_node_data('chameleon') 286 | #generate_node_data('squirrel') 287 | #generate_node_data('actor') 288 | generate_node_data('penn') 289 | 290 | -------------------------------------------------------------------------------- /Node/utils.py: -------------------------------------------------------------------------------- 1 | import time 2 | import math 3 | import random 4 | import numpy as np 5 | import scipy as sp 6 | import os 7 | import torch 8 | import torch.nn as nn 9 | import torch.nn.functional as F 10 | from ogb.nodeproppred.dataset_dgl import DglNodePropPredDataset 11 | 12 | 13 | def count_parameters(model): 14 | return sum(p.numel() for p in model.parameters() if p.requires_grad) 15 | 16 | 17 | def init_params(module): 18 | if isinstance(module, nn.Linear): 19 | module.weight.data.normal_(mean=0.0, std=0.01) 20 | if module.bias is not None: 21 | module.bias.data.zero_() 22 | 23 | 24 | def seed_everything(seed): 25 | random.seed(seed) 26 | os.environ['PYTHONHASHSEED'] = str(seed) 27 | np.random.seed(seed) 28 | torch.manual_seed(seed) 29 | torch.cuda.manual_seed(seed) 30 | torch.backends.cudnn.deterministic = True 31 | torch.backends.cudnn.benchmark = True 32 | torch.backends.cudnn.allow_tf32 = False 33 | 34 | 35 | def get_split(dataset, y, nclass, seed=0): 36 | 37 | if dataset == 'arxiv': 38 | dataset = DglNodePropPredDataset('ogbn-arxiv') 39 | split = dataset.get_idx_split() 40 | train, valid, test = split['train'], split['valid'], split['test'] 41 | return train, valid, test 42 | 43 | elif dataset == 'penn': 44 | split = np.load('node_raw_data/fb100-Penn94-splits.npy', allow_pickle=True)[0] 45 | train, valid, test = split['train'], split['valid'], split['test'] 46 | return train, valid, test 47 | 48 | else: 49 | y = y.cpu() 50 | 51 | percls_trn = int(round(0.6 * len(y) / nclass)) 52 | val_lb = int(round(0.2 * len(y))) 53 | 54 | indices = [] 55 | for i in range(nclass): 56 | index = (y == i).nonzero().view(-1) 57 | index = index[torch.randperm(index.size(0), device=index.device)] 58 | indices.append(index) 59 | 60 | train_index = torch.cat([i[:percls_trn] for i in indices], dim=0) 61 | rest_index = torch.cat([i[percls_trn:] for i in indices], dim=0) 62 | rest_index = rest_index[torch.randperm(rest_index.size(0))] 63 | valid_index = rest_index[:val_lb] 64 | test_index = rest_index[val_lb:] 65 | 66 | return train_index, valid_index, test_index 67 | 68 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Specformer 2 | Code of [Specformer: Spectral Graph Neural Networks Meet Transformers](https://openreview.net/forum?id=0pdSt3oyJa1) 3 | 4 | # How to run 5 | - For node-level task, e.g., signal regression and node classification, you should first run preprocess_node_data.py to generate .pt files for each dataset. 6 | - For graph-level taks, you can direcly run dgl_main.py. 7 | 8 | # Q&A 9 | Any suggestion/question is welcome. 10 | 11 | # Reference 12 | If you make advantage of Specformer in your research, please cite the following in your manuscript: 13 | 14 | ``` 15 | @inproceedings{specformer2023, 16 | author={Deyu Bo and 17 | Chuan Shi and 18 | Lele Wang and 19 | Renjie Liao}, 20 | title={Specformer: Spectral Graph Neural Networks Meet Transformers}, 21 | booktitle = {{ICLR}}, 22 | publisher = {OpenReview.net}, 23 | year = {2023} 24 | } 25 | ``` 26 | -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- 1 | dgl==1.0.0 2 | easydict==1.10 3 | ema_pytorch==0.2.1 4 | json5==0.9.11 5 | networkx==3.0 6 | numpy==1.24.2 7 | ogb==1.3.5 8 | pandas==1.5.3 9 | PyYAML==6.0 10 | scikit_learn==1.2.1 11 | scipy==1.10.1 12 | torch==1.13.1 13 | torch_geometric==2.2.0 14 | torchmetrics==0.11.1 15 | tqdm==4.64.1 16 | wandb==0.13.10 17 | --------------------------------------------------------------------------------