-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathrun.py
181 lines (155 loc) · 9.54 KB
/
run.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
import argparse, os, torch, random, json
from exp.exp_long_term_forecasting import *
import numpy as np
def set_random_seed(seed):
torch.manual_seed(seed)
torch.cuda.manual_seed_all(seed)
np.random.seed(seed)
random.seed(seed)
def initial_setup(args):
args.use_gpu = True if torch.cuda.is_available() else False
if args.use_gpu:
print(torch.cuda.get_device_name(0))
if args.use_gpu and args.use_multi_gpu:
args.devices = args.devices.replace(' ', '')
device_ids = args.devices.split(',')
args.device_ids = [int(id_) for id_ in device_ids]
args.gpu = args.device_ids[0]
args.enc_in = args.dec_in = args.c_out = args.n_features
args.task_name = 'long_term_forecast'
def main(args):
initial_setup(args)
set_random_seed(args.seed)
print(f'Args in experiment: {args}')
if args.itrs == 1:
exp = Exp_Long_Term_Forecast(args)
if not args.test:
if os.path.exists(exp.best_model_path):
print(f'Checkpoint exists already. Skipping...')
else:
print('>>>>>>> start training :>>>>>>>>>>')
exp.train()
print('\n>>>>>>> testing : <<<<<<<<<<<<<<<<<<<')
exp.test(flag='test', dump_output=args.dump_output)
else:
parent_seed = args.seed
np.random.seed(parent_seed)
experiment_seeds = np.random.randint(1e3, size=args.itrs)
experiment_seeds = [int(seed) for seed in experiment_seeds]
args.experiment_seeds = experiment_seeds
original_itr = args.itr_no
for itr_no in range(1, args.itrs+1):
if (original_itr is not None) and original_itr != itr_no: continue
args.seed = experiment_seeds[itr_no-1]
print(f'\n>>>> itr_no: {itr_no}, seed: {args.seed} <<<<<<')
set_random_seed(args.seed)
args.itr_no = itr_no
exp = Exp_Long_Term_Forecast(args)
if not args.test:
if os.path.exists(exp.best_model_path):
print(f'Checkpoint exists already. Skipping...')
else:
print('>>>>>>> start training :>>>>>>>>>>')
exp.train()
print('\n>>>>>>> testing : <<<<<<<<<<<<<<<<<<<')
exp.test(flag='test', dump_output=args.dump_output)
data_name = args.data_path.split('.')[0]
config_filepath = os.path.join(
args.result_path, data_name,
stringify_setting(args), 'config.json'
)
args.seed = parent_seed
with open(config_filepath, 'w') as output_file:
json.dump(vars(args), output_file, indent=4)
def get_basic_parser(
name='TimeSeries Model',
non_stationary=False, timemixer=False
):
parser = argparse.ArgumentParser(description=name)
# basic config
parser.add_argument('--test', action='store_true', help='test the model')
parser.add_argument('--seed', type=int, default=2024, help='random seed')
parser.add_argument('--result_path', type=str, default='results', help='result output folder')
parser.add_argument('--disable_progress', action='store_true', help='do not show progress bar')
# data loader
parser.add_argument('--root_path', type=str, default='./data', help='root path of the data file')
parser.add_argument('--data_path', type=str, default='Exchange_Rate_Report.csv', help='data file')
parser.add_argument('--features', type=str, default='M', choices=['M', 'S', 'MS'],
help='forecasting task, options:[M, S, MS]; M:multivariate predict multivariate, S:univariate predict univariate, MS:multivariate predict univariate')
parser.add_argument('--n_features', type=int, required=True, help='Number of input features')
parser.add_argument('--group_id', type=str, default=None, help='group id for multi-time series')
parser.add_argument('--target', type=str, default='OT', help='target feature in S or MS task')
parser.add_argument(
'--freq', type=str, default='d', choices=['s', 't', 'h', 'd', 'b', 'w', 'm', 'a'],
help='freq for time features encoding, options:[s:secondly, t:minutely, h:hourly, d:daily, b:business days, w:weekly, m:monthly, a:yearly], you can also use more detailed freq like 15min or 3h'
)
parser.add_argument('--no_scale', action='store_true', help='do not scale the dataset')
# forecasting task
parser.add_argument('--seq_len', type=int, default=96, help='input sequence length')
parser.add_argument('--label_len', type=int, default=48, help='start token length')
parser.add_argument('--pred_len', type=int, default=24, help='prediction sequence length')
# parser.add_argument('--seasonal_patterns', type=str, default='Monthly', help='subset for M4')
# parser.add_argument('--inverse', action='store_true', help='inverse output data', default=False)
# model define
parser.add_argument('--top_k', type=int, default=5, help='for TimesBlock')
parser.add_argument('--num_kernels', type=int, default=6, help='for Inception')
# parser.add_argument('--enc_in', type=int, default=7, help='encoder input size')
# parser.add_argument('--dec_in', type=int, default=7, help='decoder input size')
# parser.add_argument('--c_out', type=int, default=7, help='output size')
parser.add_argument('--d_model', type=int, default=64, help='dimension of model')
parser.add_argument('--n_heads', type=int, default=4, help='num of heads')
parser.add_argument('--e_layers', type=int, default=2, help='num of encoder layers')
parser.add_argument('--d_layers', type=int, default=1, help='num of decoder layers')
parser.add_argument('--d_ff', type=int, default=128, help='dimension of fcn')
parser.add_argument('--moving_avg', type=int, default=3, help='window size of moving average')
parser.add_argument('--factor', type=int, default=1, help='attn factor')
parser.add_argument('--dropout', type=float, default=0.1, help='dropout')
parser.add_argument('--embed', type=str, default='timeF',
help='time features encoding, options:[timeF, fixed, learned]')
parser.add_argument('--activation', type=str, default='gelu', help='activation')
parser.add_argument('--output_attention', action='store_true', help='whether to output attention in ecoder')
# TimeMixer parameters
if timemixer:
parser.add_argument('--channel_independence', type=int, default=1,
help='1: channel dependence 0: channel independence for FreTS model')
parser.add_argument('--decomp_method', type=str, default='moving_avg',
help='method of series decompsition, only support moving_avg or dft_decomp')
parser.add_argument('--use_norm', type=int, default=1, help='whether to use normalize; True 1 False 0')
parser.add_argument('--down_sampling_layers', type=int, default=0, help='num of down sampling layers')
parser.add_argument('--down_sampling_window', type=int, default=1, help='down sampling window size')
parser.add_argument('--down_sampling_method', type=str, default=None,
help='down sampling method, only support avg, max, conv')
parser.add_argument('--seg_len', type=int, default=48,
help='the length of segmen-wise iteration of SegRNN')
# optimization
parser.add_argument('--num_workers', type=int, default=10, help='data loader num workers')
parser.add_argument('--itrs', type=int, default=1, help='experiments times')
parser.add_argument('--itr_no', type=int, default=None, help='experiments number among itrs. 1<= itr_no <= itrs .')
parser.add_argument('--train_epochs', type=int, default=10, help='train epochs')
parser.add_argument('--batch_size', type=int, default=32, help='batch size of train input data')
parser.add_argument('--patience', type=int, default=3, help='early stopping patience')
parser.add_argument('--learning_rate', type=float, default=1e-3, help='optimizer learning rate')
parser.add_argument('--des', type=str, default=None, help='exp description')
parser.add_argument('--loss', type=str, default='MSE', help='loss function')
parser.add_argument('--lradj', type=str, default='type1', help='adjust learning rate')
# GPU
parser.add_argument('--gpu', type=int, default=0, help='gpu')
parser.add_argument('--use_multi_gpu', action='store_true', help='use multiple gpus', default=False)
parser.add_argument('--devices', type=str, default='0,1,2,3', help='device ids of multile gpus')
# de-stationary projector params
if non_stationary:
parser.add_argument('--p_hidden_dims', type=int, nargs='+', default=[128, 128],
help='hidden layer dimensions of projector (List)')
parser.add_argument('--p_hidden_layers', type=int, default=2, help='number of hidden layers in projector')
parser.add_argument('--dry_run', action='store_true', help='run only one batch for test')
parser.add_argument('--percent', type=int, default=100, help='Percent of recent examples (0 to 100) to keep for training. percent = 0 means no training or zeroshot.')
parser.add_argument('--dump_output', action='store_true', help='dump output as a csv file')
return parser
if __name__ == '__main__':
parser = get_basic_parser(
non_stationary=True, timemixer=True
)
parser.add_argument('--model', type=str, default='DLinear',
choices=list(Exp_Basic.model_dict.keys()), help='model name')
args = parser.parse_args()
main(args)