-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathrun_micromobility.py
149 lines (125 loc) · 7.38 KB
/
run_micromobility.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
import os
import torch
from datetime import datetime
from experiments.exp_micromobility import Exp_Micromobility
import argparse
import pandas as pd
import numpy as np
from torch.utils.tensorboard import SummaryWriter
parser = argparse.ArgumentParser(description='ICN on shared micromobility datasets')
### ------- dataset settings --------------
parser.add_argument('--dataset', type=str, default='Austin', choices=['Austin', 'Chicago', 'Chicago_Halfyear'])
parser.add_argument('--norm_method', type=str, default='z_score')
parser.add_argument('--normtype', type=int, default=2)
### ------- device settings --------------
parser.add_argument('--use_gpu', type=bool, default=True, help='use gpu')
parser.add_argument('--use_multi_gpu', action='store_true', help='use multiple gpus', default=False)
parser.add_argument('--gpu', type=int, default=0, help='gpu')
parser.add_argument('--device', type=str, default='cuda:0')
### ------- input/output length settings --------------
parser.add_argument('--window_size', type=int, default=12)
parser.add_argument('--horizon', type=int, default=12)
parser.add_argument('--concat_len', type=int, default=0)
parser.add_argument('--single_step_output_One', type=int, default=0)
parser.add_argument('--train_length', type=float, default=6)
parser.add_argument('--valid_length', type=float, default=2)
parser.add_argument('--test_length', type=float, default=2)
### ------- training settings --------------
parser.add_argument('--train', type=bool, default=False)
parser.add_argument('--resume', type=bool, default=False)
parser.add_argument('--evaluate', type=bool, default=False)
parser.add_argument('--tune', type=bool, default=False)
parser.add_argument('--validate_freq', type=int, default=1)
parser.add_argument('--ablation', type=str, default=None, choices=[None, 'poi', 'demo', 'tran'])
parser.add_argument('--weather', type=bool, default=False)
parser.add_argument('--epoch', type=int, default=80)
parser.add_argument('--lr', type=float, default=0.001)
parser.add_argument('--batch_size', type=int, default=8)
parser.add_argument('--optimizer', type=str, default='RMSProp') #
parser.add_argument('--early_stop', type=bool, default=False)
parser.add_argument('--exponential_decay_step', type=int, default=5)
parser.add_argument('--decay_rate', type=float, default=0.5)
parser.add_argument('--lradj', type=int, default=1,help='adjust learning rate')
parser.add_argument('--weight_decay', type=float, default=1e-5)
parser.add_argument('--model_name', type=str, default='ICN')
### ------- model settings --------------
parser.add_argument('--hidden_size', default=0.0625, type=float, help='hidden channel scale of module')
parser.add_argument('--INN', default=1, type=int, help='use INN or basic strategy')
parser.add_argument('--kernel', default=5, type=int, help='kernel size for the first layer')
parser.add_argument('--dilation', default=1, type=int, help='dilation')
parser.add_argument('--positionalEcoding', type=bool , default = True)
parser.add_argument('--groups', type=int, default=1)
parser.add_argument('--levels', type=int, default=2)
parser.add_argument('--stacks', type=int, default=1)
parser.add_argument('--dropout', type=float, default=0.5)
parser.add_argument('--num_decoder_layer', type=int, default=1)
parser.add_argument('--RIN', type=bool, default=False)
args = parser.parse_args()
if __name__ == '__main__':
torch.manual_seed(4321) # reproducible. 4321 is a magical seed lol.
torch.cuda.manual_seed_all(4321)
torch.backends.cudnn.benchmark = False
torch.backends.cudnn.deterministic = True # Can change it to False --> default: False
torch.backends.cudnn.enabled = True
Exp=Exp_Micromobility
# exp=Exp(args)
if args.evaluate:
exp=Exp(args)
before_evaluation = datetime.now().timestamp()
exp.test()
after_evaluation = datetime.now().timestamp()
print(f'Evaluation took {(after_evaluation - before_evaluation) / 60} minutes')
elif args.train or args.resume:
exp=Exp(args)
before_train = datetime.now().timestamp()
print("===================Normal-Start=========================")
best_epoch_val, best_validate_metrics, best_epoch_test, best_test_metrics = exp.train()
after_train = datetime.now().timestamp()
# print best epoch
print('Best validation epoch:', best_epoch_val)
print('Best test epoch:', best_epoch_test)
print('Performance on test set: | MAE: {:5.4f} | MAPE: {:5.4f} | RMSE: {:5.4f}'.format(best_test_metrics['mae'],
best_test_metrics['mape'],
best_test_metrics['rmse']))
print(f'Training took {(after_train - before_train) / 60} minutes')
print("===================Normal-End=========================")
elif args.tune:
before_tune = datetime.now().timestamp()
horizon_list = [1, 6, 12]
window_list = [24, 48, 96, 192]
batch_list = [8, 16, 32, 64]
hidden_list = [0.5, 1]
# lr_list = [0.001, 0.0005]
# horizon_list = [12]
# window_list = [96, 192]
# batch_list = [8, 16, 32, 64]
# hidden_list = [0.5, 1]
lr_list = [0.001]
results = pd.DataFrame(columns=['Horizon', 'Window', 'Batch_Size', 'Hidden_Size',
'Initial_Lr', 'MAE', 'MAPE', 'RMSE'])
print("===================Tuning-Start=========================")
for horizon in horizon_list:
for window in window_list:
for batch in batch_list:
for hidden in hidden_list:
for lr in lr_list:
args.horizon = horizon
args.window_size = window
args.batch_size = batch
args.hidden_size = hidden
args.lr = lr
exp=Exp(args)
best_epoch_val, best_validate_metrics, best_epoch_test, best_test_metrics = exp.train()
best_case = pd.DataFrame([[horizon, window, batch, hidden, lr,
best_test_metrics['mae'], best_test_metrics['mape'],
best_test_metrics['rmse']]], columns=results.columns)
results = pd.concat([results, best_case])
with open(args.dataset + '_Tuning.txt', 'a') as f:
f.write(str(horizon) + ',' + str(window) + ',' + str(batch) + ',' +
str(hidden) + ',' + str(lr) + ',' + str(best_test_metrics['mae']) +
',' + str(best_test_metrics['mape']) + ',' + str(best_test_metrics['rmse'])
+ '\n')
after_tune = datetime.now().timestamp()
results.to_csv('Results/' + args.dataset + '_Tuning.csv')
print(f'Tuning took {(after_tune - before_tune) / 60} minutes')
print("===================Tuning-End=========================")