-
Notifications
You must be signed in to change notification settings - Fork 1
/
process.py
169 lines (154 loc) · 5.4 KB
/
process.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
import os
import shutil
import csv
import pandas as pd
import triad_tuples
def csv_alter(triad, img_tuple, text_tuple, csv_path, csv_save_path):
ori_f = csv.reader(open(csv_path, 'r'))
ori_data = []
for i in ori_f:
ori_data.append(i)
triad_alter_num = 0
img_tuple_alter_num = 0
text_tuple_alter_num = 0
# alter 3-tuple
for ids in triad:
for data in ori_data:
if data[0] == str(ids[0]):
data[1] = '1'
data[2] = '1'
triad_alter_num += 1
elif data[0] == str(ids[1]):
data[1] = '0'
data[2] = '0'
triad_alter_num += 1
elif data[0] == str(ids[2]):
data[1] = '0'
data[2] = '0'
triad_alter_num += 1
# alter img_tuple
for ids in img_tuple:
id_0 = str(ids[0])
id_1 = str(ids[1])
prob_0 = None
label_0 = None
prob_1 = None
label_1 = None
# find ids
for data in ori_data:
if data[0] == id_0:
prob_0 = data[1]
label_0 = data[2]
break
for data in ori_data:
if data[0] == id_1:
prob_1 = data[1]
label_1 = data[2]
break
# compare and alter
if prob_0 is not None and prob_1 is not None:
if float(prob_0) >= float(prob_1):
for data in ori_data:
if data[0] == id_0:
data[1] = '1'
data[2] = '1'
img_tuple_alter_num += 1
break
for data in ori_data:
if data[0] == id_1:
data[1] = '0'
data[2] = '0'
img_tuple_alter_num += 1
break
else:
for data in ori_data:
if data[0] == id_0:
data[1] = '0'
data[2] = '0'
img_tuple_alter_num += 1
break
for data in ori_data:
if data[0] == id_1:
data[1] = '1'
data[2] = '1'
img_tuple_alter_num += 1
break
# alter text_tuple
for ids in text_tuple:
id_0 = str(ids[0])
id_1 = str(ids[1])
prob_0 = None
label_0 = None
prob_1 = None
label_1 = None
# find ids
for data in ori_data:
if data[0] == id_0:
prob_0 = data[1]
label_0 = data[2]
break
for data in ori_data:
if data[0] == id_1:
prob_1 = data[1]
label_1 = data[2]
break
# compare and alter
if prob_0 is not None and prob_1 is not None:
if float(prob_0) >= float(prob_1):
for data in ori_data:
if data[0] == id_0:
data[1] = '1'
data[2] = '1'
text_tuple_alter_num += 1
break
for data in ori_data:
if data[0] == id_1:
data[1] = '0'
data[2] = '0'
text_tuple_alter_num += 1
break
else:
for data in ori_data:
if data[0] == id_0:
data[1] = '0'
data[2] = '0'
text_tuple_alter_num += 1
break
for data in ori_data:
if data[0] == id_1:
data[1] = '1'
data[2] = '1'
text_tuple_alter_num += 1
break
with open(csv_save_path, 'w') as alter_f:
writer = csv.writer(alter_f)
for i in ori_data:
writer.writerow(i)
print('triad_alter_num:{}, img_tuple_alter_num:{}, text_tuple_alter_num:{}'.format(
triad_alter_num, img_tuple_alter_num, text_tuple_alter_num))
def treat_csv(path, save_path):
triad, img_tuple, text_tuple = triad_tuples.get_triad_and_tuples(
triad_tuples.test_unseen_img_path, triad_tuples.test_unseen_path)
if os.path.exists(save_path):
shutil.rmtree(save_path)
os.mkdir(save_path)
for file in os.listdir(path):
input_csv = os.path.join(path, file)
output_csv = os.path.join(save_path, file)
csv_alter(triad, img_tuple, text_tuple, input_csv, output_csv)
def stacking(path, save_path):
results = []
for file in os.listdir(path):
data = pd.read_csv(path+'/'+file, index_col=0)
results.append(data)
proba_test = pd.concat([rst.proba for rst in results], axis=1)
proba_test.columns = range(proba_test.shape[1])
rst = pd.DataFrame()
rst['proba'] = proba_test.mean(1)
rst['label'] = (rst['proba'] >= 0.5).map(int)
rst.to_csv(save_path)
def main():
treat_csv('./csv', './treated_csv')
stacking('./treated_csv', './final_result.csv')
if __name__ == '__main__':
main()