-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathutils.py
186 lines (157 loc) · 5.12 KB
/
utils.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 numpy as np
import json
import random
import scipy.sparse as ssp
import logging
def reltri2tri(trip_tasks, rel2id, ent2id):
trip_data = []
for key in trip_tasks.keys():
rel_trip = trip_tasks[key]
for j in range(len(rel_trip)):
trip_data.append([ent2id[rel_trip[j][0]], rel2id[rel_trip[j][1]], ent2id[rel_trip[j][2]]])
return trip_data
def h2hrt(train_id):
h_hrt = {}
for i in range(len(train_id)):
if train_id[i][0] not in h_hrt.keys():
h_hrt[train_id[i][0]] = []
h_hrt[train_id[i][0]].append(train_id[i])
return h_hrt
def trip2rel2tripid(trip_id_all,rel_all):
rel2trip = {}
for i in range(len(rel_all)):
rel2trip[rel_all[i]] = []
for j in range(len(trip_id_all)):
trip_one = trip_id_all[j]
rel2trip[trip_one[1]].append(trip_one)
return rel2trip
def item2id(item, rel2id):
item_id = []
for i in range(len(item)):
one = []
for j in range(len(item[i])):
one.append(rel2id[item[i][j]])
item_id.append(one)
return item_id
def path_read(path_dict_str, rel2id, ent2id):
path_str = {}
path_id = {}
for key,item in path_dict_str.items():
h,t = key.split('&')[0], key.split('&')[1]
key_update = (h,t)
key_update_id = (ent2id[h],ent2id[t])
if len(item)!=0:
item_id = item2id(item, rel2id)
else:
item_id = []
path_str[key_update] = item
path_id[key_update_id] = item_id
return path_str,path_id
def test_relkind(rel_test_trip, train_test_path_id):
test2relkind_dict = {}
test2relkind = {}
for key, item in rel_test_trip.items():
test2relkind_dict[key] = []
test2relkind[key] = []
for i in range(len(item)):
trip = item[i]
pair = (trip[0],trip[2])
path = train_test_path_id[pair]
path_set = set()
for m in range(len(path)):
for n in range(len(path[m])):
path_set.add(path[m][n])
test2relkind_dict[key].append(pair)
test2relkind[key].append(list(path_set))
return test2relkind_dict, test2relkind
def set_rel_sim_count(num_rel_id):
sim_set = {}
for key, item1 in num_rel_id.items():
list_rel_all = item1
set_rel_all = []
for i in range(len(list_rel_all)):
set_rel_all.append(set(list_rel_all[i]))
set_onerel_sim = []
set_avg = []
for setm in set_rel_all:
set_one_sim = []
for setn in set_rel_all:
bing = list(set(setm) | set(setn))
jiao = list(set(setm)&set(setn))
if len(bing)!=0:
set_one_sim.append(len(jiao)/len(bing))
else:
set_one_sim.append(0)
set_onerel_sim.append(set_one_sim)
sum_one = (sum(set_one_sim) - 1)/(len(set_one_sim)-1)
set_avg.append(sum_one)
sim_set[key] = set_avg
return sim_set
def train_generate(sp_num, dataset_path, batch_size, train_tasks, ent2id, rel2id, id2ent, id2rel, e1rel_e2, rel2candidates):
task_pool = list(train_tasks.keys())
num_tasks = len(task_pool)
rel_idx = 0
while True:
if rel_idx % num_tasks == 0:
random.shuffle(task_pool)
query = task_pool[rel_idx % num_tasks]
rel_idx += 1
candidates = rel2candidates[query]
candidates_id = []
for i in range(len(candidates)):
candidates_id.append(ent2id[candidates[i]])
if len(candidates) <= 20:
continue
rel_tt = train_tasks[query]
random.shuffle(rel_tt)
train_tri_id = [[ent2id[triple[0]], rel2id[triple[1]], ent2id[triple[2]]] for triple in rel_tt]
train_tri_id_fil = []
for trip in train_tri_id:
if trip[0] != trip[2]:
train_tri_id_fil.append(trip)
train_tri_id = train_tri_id_fil
support_pair = train_tri_id[:sp_num]
query_pair = train_tri_id[sp_num:]
if len(support_pair) == 0 or len(query_pair)==0:
continue
if len(query_pair) < batch_size:
query_pair_pos = [random.choice(query_pair) for _ in range(batch_size)]
else:
query_pair_pos = random.sample(query_pair, batch_size)
support_pair = [[pair[0],pair[2]] for pair in support_pair]
query_pair_pos = [[pair[0],pair[2]] for pair in query_pair_pos]
one_tomany_train = []
for i in range(len(query_pair_pos)):
one2many = e1rel_e2[id2ent[int(query_pair_pos[i][0])]+query]
one2many2id = [ent2id[_] for _ in one2many]
one_tomany_train.append(one2many2id)
yield support_pair, query_pair_pos, one_tomany_train,candidates_id
def rel_submit(pair,train_test_path_id):
rel_all = []
for i in range(len(pair)):
pair_one = (pair[i][0], pair[i][1])
rel_list = train_test_path_id[pair_one]
rel_set = set()
for m in range(len(rel_list)):
for n in range(len(rel_list[m])):
rel_set.add(rel_list[m][n])
rel_all.append(list(rel_set))
return rel_all
def path_submit(pair, train_test_path_id):
path_all = []
for i in range(len(pair)):
pair_one = (pair[i][0], pair[i][1])
path_list = train_test_path_id[pair_one]
path_all.append(path_list)
return path_all
def pad_tensor(tensor: torch.Tensor, length, value=0, dim=0) -> torch.Tensor:
return torch.cat(
(tensor, tensor.new_full((*tensor.size()[:dim], length - tensor.size(dim), *tensor.size()[dim + 1:]), value)),
dim=dim)
def list2tensor(data_list: list, padding_idx, dtype=torch.long, device=torch.device("cpu")):
max_len = max(map(len, data_list))
max_len = max(max_len, 1)
data_tensor = torch.stack(
tuple(pad_tensor(torch.tensor(data, dtype=dtype), max_len, padding_idx, 0) for data in data_list)).to(device)
return data_tensor