-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathpreprocess_multi.py
More file actions
189 lines (166 loc) · 9.17 KB
/
preprocess_multi.py
File metadata and controls
189 lines (166 loc) · 9.17 KB
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
import os
import shutil
import sys
import pickle
def filter(data):
new_data = []
for d in data:
if 'wikidata' in d:
new_data.append(d[1:5])
else:
new_data.append(d)
return new_data
def process_wiki_txt(original_path, processed_path):
new_data = []
with open(original_path) as f:
for line in f.readlines():
data = line.strip().split('\t')
new_data.append(filter(data))
with open(processed_path, 'w') as fw:
for new_d in new_data:
fw.write('\t'.join(new_d) + '\n')
def process_wiki2_txt(original_path, processed_path):
new_data = []
with open(original_path) as f:
for line in f.readlines():
data = line.strip().split(' ')
new_data.append(filter(data))
with open(processed_path, 'w') as fw:
for new_d in new_data:
fw.write('\t'.join(new_d) + '\n')
def get_graph_file(dataset_name, original_file_path, processed_folder_path):
new_data = []
if dataset_name in ['M-FB15K', 'FB-AUTO', 'FB-AUTO-2', 'FB-AUTO-4', 'FB-AUTO-5', 'JF17K-4', 'JF17K-2', 'JF17K-5', 'JF17K-6', 'JF17K', 'WikiPeople-4', 'JF17K_new', 'WD50K_100-3', 'WD50K_100-4', 'WD50K_100-5', 'WD50K_100-6']:
with open(original_file_path) as f:
for index, line in enumerate(f.readlines()):
data = line.strip().split('\t')
rel = data[0]
entities = data[1:]
for i,e1 in enumerate(entities):
for j,e2 in enumerate(entities):
if i != j:
new_data.append([e2, rel+'/'+str(index)+'/'+str(j)+'/'+str(i), e1])
elif dataset_name in ['JF17K-3', 'WikiPeople-3']:
with open(original_file_path) as f:
for index, line in enumerate(f.readlines()):
data = line.strip().split(' ')
rel = data[0]
entities = data[1:]
for i,e1 in enumerate(entities):
for j,e2 in enumerate(entities):
if i != j:
new_data.append([e2, rel+'/'+str(index)+'/'+str(j)+'/'+str(i), e1])
new_data = [list(j) for j in list(set([tuple(i) for i in new_data]))] # remove duplicates
with open(os.path.join(processed_folder_path, 'graph.txt'), 'w') as fw:
for new_d in new_data:
fw.write('\t'.join(new_d) + '\n')
def get_pickle_file(dataset_name, original_file_path, processed_folder_path):
all_data = []
all_edge = []
entity2instance = dict()
instance2entity = dict()
instance2rel = dict()
if dataset_name in ['M-FB15K', 'FB-AUTO', 'FB-AUTO-2', 'FB-AUTO-4', 'FB-AUTO-5', 'JF17K-4', 'JF17K-2', 'JF17K-5', 'JF17K-6', 'JF17K', 'WikiPeople-4', 'WD50K_100-3', 'WD50K_100-4', 'WD50K_100-5', 'WD50K_100-6']:
with open(original_file_path) as f:
for line in f.readlines():
data = line.strip().split('\t')
all_edge.append(data[0])
all_data.append(data)
elif dataset_name in ['JF17K-3', 'WikiPeople-3']:
with open(original_file_path) as f:
for line in f.readlines():
data = line.strip().split(' ')
all_edge.append(data[0])
all_data.append(data)
all_name = ['instance'+str(i) for i in range(len(all_data))]
for i in range(len(all_data)):
rel = all_data[i][0]
entities = all_data[i][1:]
name = all_name[i]
if name not in instance2entity: instance2entity[name] = []
for e in entities:
if e not in entity2instance: entity2instance[e] = set()
entity2instance[e].add(name)
instance2entity[name].append(e)
instance2rel[name] = rel
with open(os.path.join(processed_folder_path, 'entity2instance.pkl'), 'wb') as fw:
pickle.dump(entity2instance, fw)
with open(os.path.join(processed_folder_path, 'instance2entity.pkl'), 'wb') as fw:
pickle.dump(instance2entity, fw)
with open(os.path.join(processed_folder_path, 'instance2rel.pkl'), 'wb') as fw:
pickle.dump(instance2rel, fw)
def process_test_file(original_file, processed_file):
new_data = []
with open(original_file) as f:
for line in f.readlines():
data = line.strip().split('\t')
new_data.append(data[1:])
with open(processed_file, 'w') as fw:
for new_d in new_data:
fw.write('\t'.join(new_d) + '\n')
def format_file(original_file, processed_file):
new_data = []
with open(original_file) as f:
for line in f.readlines():
data = line.strip().split(' ')
new_data.append(data)
with open(processed_file, 'w') as fw:
for new_d in new_data:
fw.write('\t'.join(new_d) + '\n')
def process_folder(dataset_name):
if dataset_name in ['M-FB15K', 'FB-AUTO', 'JF17K-4']:
original_folder = './data_original/{}'.format(dataset_name)
processed_folder = './data_preprocessed/{}'.format(dataset_name)
if not os.path.isdir(processed_folder):
os.makedirs(processed_folder)
get_graph_file(dataset_name, os.path.join(original_folder, 'train.txt'), processed_folder)
get_pickle_file(dataset_name, os.path.join(original_folder, 'train.txt'), processed_folder)
shutil.copyfile(os.path.join(original_folder, 'train.txt'), os.path.join(processed_folder, 'train.txt'))
#shutil.copyfile(os.path.join(original_folder, 'valid.txt'), os.path.join(processed_folder, 'valid.txt'))
shutil.copyfile(os.path.join(original_folder, 'test.txt'), os.path.join(processed_folder, 'test.txt'))
elif dataset_name in ['WikiPeople-4']:
original_folder = './data_original/{}'.format(dataset_name)
processed_folder = './data_preprocessed/{}'.format(dataset_name)
if not os.path.isdir(processed_folder):
os.makedirs(processed_folder)
process_wiki_txt(os.path.join(original_folder, 'train.txt'), os.path.join(processed_folder, 'train.txt'))
process_wiki_txt(os.path.join(original_folder, 'valid.txt'), os.path.join(processed_folder, 'valid.txt'))
process_wiki_txt(os.path.join(original_folder, 'test.txt'), os.path.join(processed_folder, 'test.txt'))
get_graph_file(dataset_name, os.path.join(processed_folder, 'train.txt'), processed_folder)
get_pickle_file(dataset_name, os.path.join(processed_folder, 'train.txt'), processed_folder)
elif dataset_name in ['JF17K-3']:
original_folder = './data_original/{}'.format(dataset_name)
processed_folder = './data_preprocessed/{}'.format(dataset_name)
if not os.path.isdir(processed_folder):
os.makedirs(processed_folder)
get_graph_file(dataset_name, os.path.join(original_folder, 'train.txt'), processed_folder)
get_pickle_file(dataset_name, os.path.join(original_folder, 'train.txt'), processed_folder)
format_file(os.path.join(original_folder, 'train.txt'), os.path.join(processed_folder, 'train.txt'))
format_file(os.path.join(original_folder, 'valid.txt'), os.path.join(processed_folder, 'valid.txt'))
format_file(os.path.join(original_folder, 'test.txt'), os.path.join(processed_folder, 'test.txt'))
elif dataset_name in ['JF17K']:
original_folder = './data_original/{}/instances'.format(dataset_name)
processed_folder = './data_preprocessed/{}'.format(dataset_name)
if not os.path.isdir(processed_folder):
os.makedirs(processed_folder)
get_graph_file(dataset_name, os.path.join(original_folder, 'train.txt'), processed_folder)
shutil.copyfile(os.path.join(original_folder, 'train.txt'), os.path.join(processed_folder, 'train.txt'))
process_test_file(os.path.join(original_folder, 'test.txt'), os.path.join(processed_folder, 'test.txt'))
elif dataset_name in ['WikiPeople-3']:
original_folder = './data_original/{}'.format(dataset_name)
processed_folder = './data_preprocessed/{}'.format(dataset_name)
if not os.path.isdir(processed_folder):
os.makedirs(processed_folder)
process_wiki2_txt(os.path.join(original_folder, 'train.txt'), os.path.join(processed_folder, 'train.txt'))
process_wiki2_txt(os.path.join(original_folder, 'valid.txt'), os.path.join(processed_folder, 'valid.txt'))
process_wiki2_txt(os.path.join(original_folder, 'test.txt'), os.path.join(processed_folder, 'test.txt'))
get_graph_file(dataset_name, os.path.join(processed_folder, 'train.txt'), processed_folder)
get_pickle_file(dataset_name, os.path.join(processed_folder, 'train.txt'), processed_folder)
elif dataset_name in ['JF17K-2', 'JF17K-5', 'JF17K-6', 'FB-AUTO-2', 'FB-AUTO-4', 'FB-AUTO-5', 'WD50K_100-3', 'WD50K_100-4', 'WD50K_100-5', 'WD50K_100-6']:
# these datasets need to be constructed based on the original dataset
original_folder = './data_preprocessed/{}'.format(dataset_name)
processed_folder = './data_preprocessed/{}'.format(dataset_name)
get_graph_file(dataset_name, os.path.join(original_folder, 'train.txt'), processed_folder)
get_pickle_file(dataset_name, os.path.join(original_folder, 'train.txt'), processed_folder)
dataset = sys.argv[1]
process_folder(dataset)