-
Notifications
You must be signed in to change notification settings - Fork 1
/
main.py
executable file
·210 lines (187 loc) · 10.7 KB
/
main.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
import os
import sys
import torch
import numpy as np
import torch.nn as nn
import argparse
import configparser
from datetime import datetime
from model.generator import DAAGCN as Generator
from model.discriminator import Discriminator, Discriminator_RF
from trainer import Trainer
from dataloader import get_dataloader, get_dataloader_meta_la
from utils.metrics import MAE_torch
from utils.util import *
from utils.adj_dis_matrix import get_adj_dis_matrix, norm_Adj as norm_adj
#*************************************************************************#
Mode = 'Train'
DEBUG = 'True'
DATASET = 'PEMS04' # PEMS03 or PEMS04 or PEMS07 or PEMS08 or METR-LA or PEMS-Bay
MODEL = 'TrendGCN'
ADJ_MATRIX = './dataset/{}/{}.csv'.format(DATASET, DATASET)
#*************************************************************************#
# get configuration
config_file = './config/{}.conf'.format(DATASET)
print('Reading configuration file: %s' % (config_file))
config = configparser.ConfigParser()
config.read(config_file)
def get_arguments():
# parser
parser = argparse.ArgumentParser(description='arguments')
parser.add_argument('--dataset', default=DATASET, type=str)
parser.add_argument('--mode', default=Mode, type=str)
parser.add_argument('--debug', default=DEBUG, type=eval)
parser.add_argument('--model', default=MODEL, type=str)
parser.add_argument('--adj_file', default=ADJ_MATRIX, type=str)
parser.add_argument('--gpu_id', default=0, type=int)
# data
parser.add_argument('--val_ratio', default=config['data']['val_ratio'], type=float)
parser.add_argument('--test_ratio', default=config['data']['test_ratio'], type=float)
parser.add_argument('--lag', default=config['data']['lag'], type=int)
parser.add_argument('--horizon', default=config['data']['horizon'], type=int)
parser.add_argument('--num_nodes', default=config['data']['num_nodes'], type=int)
parser.add_argument('--tod', default=config['data']['tod'], type=eval)
parser.add_argument('--normalizer', default=config['data']['normalizer'], type=str)
parser.add_argument('--column_wise', default=config['data']['column_wise'], type=eval)
parser.add_argument('--default_graph', default=config['data']['default_graph'], type=eval)
# model
parser.add_argument('--input_dim', default=config['model']['input_dim'], type=int)
parser.add_argument('--output_dim', default=config['model']['output_dim'], type=int)
parser.add_argument('--embed_dim', default=config['model']['embed_dim'], type=int)
parser.add_argument('--rnn_units', default=config['model']['rnn_units'], type=int)
parser.add_argument('--num_layers', default=config['model']['num_layers'], type=int)
parser.add_argument('--cheb_k', default=config['model']['cheb_order'], type=int)
# train
parser.add_argument('--loss_func', default=config['train']['loss_func'], type=str)
parser.add_argument('--seed', default=config['train']['seed'], type=int)
parser.add_argument('--batch_size', default=config['train']['batch_size'], type=int)
parser.add_argument('--epochs', default=config['train']['epochs'], type=int)
parser.add_argument('--lr_init', default=config['train']['lr_init'], type=float)
parser.add_argument('--lr_decay', default=config['train']['lr_decay'], type=eval)
parser.add_argument('--lr_decay_rate', default=config['train']['lr_decay_rate'], type=float)
parser.add_argument('--lr_decay_step', default=config['train']['lr_decay_step'], type=str)
parser.add_argument('--early_stop', default=config['train']['early_stop'], type=eval)
parser.add_argument('--early_stop_patience', default=config['train']['early_stop_patience'], type=int)
parser.add_argument('--grad_norm', default=config['train']['grad_norm'], type=eval)
parser.add_argument('--max_grad_norm', default=config['train']['max_grad_norm'], type=int)
parser.add_argument('--real_value', default=config['train']['real_value'], type=eval, help = 'use real value for loss calculation')
# test
parser.add_argument('--mae_thresh', default=config['test']['mae_thresh'], type=eval)
parser.add_argument('--mape_thresh', default=config['test']['mape_thresh'], type=float)
# log
parser.add_argument('--log_dir', default='./', type=str)
parser.add_argument('--log_step', default=config['log']['log_step'], type=int)
parser.add_argument('--plot', default=config['log']['plot'], type=eval)
args = parser.parse_args()
for arg, value in sorted(vars(args).items()):
print(f"{arg}: {value}")
return args
def init_model(model):
for p in model.parameters():
if p.dim() > 1:
nn.init.xavier_uniform_(p)
else:
nn.init.uniform_(p)
print_model_parameters(model, only_num=False)
return model
if __name__ == "__main__":
args = get_arguments()
args.device = get_device(args)
# init generator and discriminator model
generator = Generator(args)
generator = generator.to(args.device)
generator = init_model(generator)
discriminator = Discriminator(args)
discriminator = discriminator.to(args.device)
discriminator = init_model(discriminator)
discriminator_rf = Discriminator_RF(args)
discriminator_rf = discriminator_rf.to(args.device)
discriminator_rf = init_model(discriminator_rf)
if args.dataset in ['METR-LA', 'PEMS-Bay']:
train_loader, val_loader, test_loader, scaler = get_dataloader_meta_la(args,
normalizer=args.normalizer,
tod=args.tod,
dow=False,
weather=False,
single=False)
# load dataset X = [B', W, N, D], Y = [B', H, N, D]
else:
train_loader, val_loader, test_loader, scaler = get_dataloader(args,
normalizer=args.normalizer,
tod=args.tod,
dow=False,
weather=False,
single=False)
# get norm adj_matrix, norm dis_matrix
cuda = True if torch.cuda.is_available() else False
TensorFloat = torch.cuda.FloatTensor if cuda else torch.FloatTensor
if args.dataset.lower() == 'pems03':
adj_matrix, dis_matrix = get_adj_dis_matrix(args.adj_file, args.num_nodes, False, "./dataset/PEMS03/PEMS03.txt")
norm_adj_matrix, norm_dis_matrix = TensorFloat(norm_adj(adj_matrix)), TensorFloat(norm_adj(dis_matrix))
elif args.dataset.lower() in ['metr-la', 'pems-bay']:
norm_adj_matrix, norm_dis_matrix = None, None
else:
adj_matrix, dis_matrix = get_adj_dis_matrix(args.adj_file, args.num_nodes, False)
norm_adj_matrix, norm_dis_matrix = TensorFloat(norm_adj(adj_matrix)), TensorFloat(norm_adj(dis_matrix))
# loss function
if args.loss_func == 'mask_mae':
loss_G = masked_mae_loss(scaler, mask_value=0.0)
elif args.loss_func == 'mae':
loss_G = torch.nn.L1Loss().to(args.device)
elif args.loss_func == 'mse':
loss_G = torch.nn.MSELoss().to(args.device)
else:
raise ValueError
loss_D = torch.nn.BCELoss()
# optimizer
optimizer_G = torch.optim.Adam(params=generator.parameters(),
lr=args.lr_init,
eps=1.0e-8,
weight_decay=0,
amsgrad=False)
optimizer_D = torch.optim.Adam(params=discriminator.parameters(),
lr=args.lr_init*0.1,
eps=1.0e-8,
weight_decay=0,
amsgrad=False)
optimizer_D_RF = torch.optim.Adam(params=discriminator_rf.parameters(),
lr=args.lr_init*0.1,
eps=1.0e-8,
weight_decay=0,
amsgrad=False)
# learning rate decay scheduler
lr_scheduler_G, lr_scheduler_D, lr_scheduler_D_RF = None, None, None
if args.lr_decay:
print('Applying learning rate decay.')
lr_decay_steps = [int(i) for i in list(args.lr_decay_step.split(','))]
lr_scheduler_G = torch.optim.lr_scheduler.MultiStepLR(optimizer=optimizer_G,
milestones=lr_decay_steps,
gamma=args.lr_decay_rate)
lr_scheduler_D = torch.optim.lr_scheduler.MultiStepLR(optimizer=optimizer_D,
milestones=lr_decay_steps,
gamma=args.lr_decay_rate)
lr_scheduler_D_RF = torch.optim.lr_scheduler.MultiStepLR(optimizer=optimizer_D_RF,
milestones=lr_decay_steps,
gamma=args.lr_decay_rate)
# lr_scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(optimizer=optimizer, T_max=64)
# config log path
current_time = datetime.now().strftime('%Y%m%d%H%M%S')
current_dir = os.path.dirname(os.path.realpath(__file__))
log_dir = os.path.join(current_dir, 'log', args.dataset, current_time)
args.log_dir = log_dir
# model training or testing
trainer = Trainer(args,
generator, discriminator, discriminator_rf,
train_loader, val_loader, test_loader, scaler,
norm_dis_matrix,
loss_G, loss_D,
optimizer_G, optimizer_D, optimizer_D_RF,
lr_scheduler_G, lr_scheduler_D, lr_scheduler_D_RF)
if args.mode.lower() == 'train':
trainer.train()
elif args.mode.lower() == 'test':
# generator.load_state_dict(torch.load('./log/{}/20221128054144/best_model.pth'.format(args.dataset)))
print("Load saved model")
trainer.test(generator, norm_dis_matrix, trainer.args, test_loader, scaler, trainer.logger, path=f'./log/{args.dataset}/20221130052054/')
else:
raise ValueError