-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathtest.py
185 lines (165 loc) · 7.89 KB
/
test.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
import torch
import os
import yaml
import numpy as np
import torch.nn.functional as F
from tqdm import tqdm
from xarray import open_dataset
from net.CombinatorialNetwork import CombinatorialNet
from net.confidence import confidenceNetwork
from newdataset import gridNewDataset
class Test():
def __init__(self, combinatorial_args, device):
self.needed = [0, 8, 14, 17, 22, 28, 31, 35, 40]
self.tsthreas = [0.1, 3, 10, 20]
self.mean = torch.load(
'/mnt/pami23/stma/weather/processed_data/mean.pth').numpy()
self.std = torch.load(
'/mnt/pami23/stma/weather/processed_data/std.pth').numpy()
self.confidence = confidenceNetwork().to(device)
self.prediction = CombinatorialNet(
combinatorial_args['encoder']['in_channels'],
combinatorial_args['encoder']['mid_channels'],
combinatorial_args['encoder']['out_channels'],
combinatorial_args['ordinal']['mid_channels'],
combinatorial_args['ordinal']['out_channels'],
combinatorial_args['decoder']['mid_channels'],
combinatorial_args['decoder']['out_channels'],
combinatorial_args['nclass'],
noise_mean=0,
noise_std=1e-1).to(device)
self.device = device
self.nClass = combinatorial_args['nclass']
def initialize(self, confidence_path, encoder_path, decoder_path,
ordinal_path):
confidence_ckpt = torch.load(confidence_path)
encoder_ckpt = torch.load(encoder_path)
decoder_ckpt = torch.load(decoder_path)
ordinal_ckpt = torch.load(ordinal_path)
self.confidence.load_state_dict(confidence_ckpt)
self.confidence.eval()
self.prediction.encoder.load_state_dict(encoder_ckpt)
self.prediction.decoder.load_state_dict(decoder_ckpt)
self.prediction.OD.load_state_dict(ordinal_ckpt)
self.prediction.eval()
def test(self):
config = yaml.load(open('config.yaml', 'r'), Loader=yaml.FullLoader)
test_path = config['test_dir']
out_path = '/mnt/pami23/stma/weather/output/0116/'
for i in tqdm(range(400)):
file_dir_name = os.path.join(test_path,
'example' + '{:0>5d}'.format(i + 1))
write_dir_name = os.path.join(out_path,
'example' + '{:0>5d}'.format(i + 1))
if not os.path.exists(write_dir_name):
os.mkdir(write_dir_name)
for j in range(9):
loc_file_name = os.path.join(
file_dir_name,
'ji_loc_inputs_' + '{:0>2d}'.format(j + 1) + '.txt')
input_file_name = os.path.join(
file_dir_name,
'grid_inputs_' + '{:0>2d}'.format(j + 1) + '.nc')
if not os.path.exists(input_file_name) or not os.path.exists(
loc_file_name):
continue
input_list = self.get_input_list(input_file_name)
input_list = torch.from_numpy(input_list).to(self.device)
idx = 0
out_file_name = os.path.join(
write_dir_name, 'pred_' + '{:0>2d}'.format(j + 1) + '.txt')
with open(
out_file_name, 'w'
) as f: # 如果filename不存在会自动创建， 'w'表示写数据，写之前会清空文件中的原有数据！
for line in open(loc_file_name):
line = line.strip().split()
row, col = line
row, col = int(row), int(col)
#print(row, col)
temp_input = input_list[:, row - 8:row + 9,
col - 8:col + 9].unsqueeze(0)
#print(temp_input.shape)
confidence_result = self.confidence(temp_input)
ordinal_results = self.prediction(temp_input,
isOrdinal=True)
ordinal_results = ordinal_results[0]
if ordinal_results[0] < 0.5:
prediction = 0
elif ordinal_results[0] > 0.5 and ordinal_results[
1] < 0.5:
prediction = 0.1
elif ordinal_results[1] > 0.5 and ordinal_results[
2] < 0.5:
prediction = 3
elif ordinal_results[2] > 0.5 and ordinal_results[
3] < 0.5:
prediction = 10
else:
prediction = 20
f.write(str(prediction) + '\n')
f.close()
def get_input_list(self, input_file_name):
input = open_dataset(input_file_name)
input_values = self.read_file(input)
temp_list = []
i = 0
for values in input_values:
if values.ndim == 3:
values = np.transpose(values, (2, 0, 1))
for num in range(values.shape[0]):
#temp_list.append(values[num].flatten().tolist())
if i in self.needed:
temp_list.append((values[num] - self.mean[i]) /
self.std[i].tolist())
i += 1
else:
#temp_list.append(values.flatten().tolist())
temp_list.append(
(values - self.mean[i]) / self.std[i].tolist())
i += 1
input = np.asarray(temp_list)
return input
def read_file(self, file):
vars_name = [name for name in file]
return [file.data_vars[name].values for name in vars_name]
def generateOneHot(self, softmax):
maxIdxs = torch.argmax(softmax, dim=1, keepdim=True).cpu().long()
oneHotMask = torch.zeros(softmax.shape, dtype=torch.float32)
oneHotMask = oneHotMask.scatter_(1, maxIdxs, 1.0)
#oneHotMask = oneHotMask.unsqueeze(-2)
return oneHotMask
def updateFile(self):
out_path = '/mnt/pami23/stma/weather/output/0114/'
new_path = '/mnt/pami23/stma/weather/output/0114new/'
for i in tqdm(range(400)):
new_dir_name = os.path.join(new_path,
'example' + '{:0>5d}'.format(i + 1))
write_dir_name = os.path.join(out_path,
'example' + '{:0>5d}'.format(i + 1))
if not os.path.exists(new_dir_name):
os.mkdir(new_dir_name)
for j in range(9):
out_file_name = os.path.join(
write_dir_name, 'pred_' + '{:0>2d}'.format(j + 1) + '.txt')
if not os.path.exists(out_file_name):
continue
new_file_name = os.path.join(
new_dir_name, 'Pred_' + '{:0>2d}'.format(j + 1) + '.txt')
with open(
new_file_name, 'w'
) as f: # 如果filename不存在会自动创建， 'w'表示写数据，写之前会清空文件中的原有数据！
for line in open(out_file_name):
line = line.strip().split()
row, col, pred = line
#print(row, col, pred)
f.write(str(pred) + '\n')
f.close()
if __name__ == "__main__":
config = yaml.load(open('config.yaml', 'r'), Loader=yaml.FullLoader)
device = 'cuda'
test = Test(config['combinatotorial'], device)
#test.test()
test.initialize('checkpoint/confidence2.pth',
'checkpoint/encoderwithodr3.pth', 'checkpoint/decoder.pth',
'checkpoint/odr3.pth')
test.test()