-
Notifications
You must be signed in to change notification settings - Fork 2
/
pre_poi_transformer.py
107 lines (69 loc) · 2.43 KB
/
pre_poi_transformer.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
import pandas as pd
import pickle
from shapely.geometry import Point, LineString
from shapely.geometry import Polygon,MultiPoint
import torch
from torch import nn
import numpy as np
def load_data(file):
data_load_file = []
file_1 = open(file, "rb")
data_load_file = pickle.load(file_1)
return data_load_file
region_back = load_data("../data/region_back_merge.pickle")
reg_poi = load_data("../data/reg_incld_poi_new.pickle")
# reg_spatial = load_data("../data/region_spatial.pickle")
poi_max=[]
for key,value in reg_poi.items():
poi_max.extend(value)
print(max(poi_max)) #there are 120 fin-grained pois
# println()
reg_poi_={}
s = 0
emb = nn.Embedding(120, 512)
embedding_spatial = torch.nn.Embedding(15, 512) # spatial
for key,value in reg_poi.items():
# print("value:",value)
if value!=[]:
reg_poi_[key]=[]
# print("value:",value)
if len(value)>s:
s = len(value)
for item in value:
reg_poi_[key].append(emb(torch.tensor(item)).tolist())
# spa_vec= embedding_spatial(torch.tensor(reg_spatial[idx]))
# reg_poi_t = {}
reg_poi_list = []
for iii in range(180):
# for key,value in reg_poi_.items():
if iii not in reg_poi_.keys():
reg_poi_list.append(np.array([0.0]*512))
# reg_poi_list.append(ci)
else:
# print("value:",value)
tp = np.mean(reg_poi_[key],axis=0)
reg_poi_list.append(tp)
reg_poi_list_ = torch.tensor(np.array(reg_poi_list)).float()
reg_poi_list_tensor = torch.unsqueeze(reg_poi_list_,0)
print(reg_poi_list_tensor.size())
reg_idx= [key for key in reg_poi_.keys()]
from torch import nn
encoder_layer = nn.TransformerEncoderLayer(d_model=512, nhead=8 )
transformer_encoder = nn.TransformerEncoder(encoder_layer, num_layers=6)
# src = torch.rand(1, 172, 512)
src = reg_poi_list_tensor
out = transformer_encoder(src)
# print(out.size())
out_ = torch.squeeze(out,0)
print(out_.size())
print(reg_idx)
print(len(reg_idx))
# reg_poi_vec = {}
# for idx,vec in zip(reg_idx,out_):
# reg_poi_vec[idx] = vec
file=open(r"../data/reg_poi_vec.pickle","wb")
pickle.dump(out_,file) #storing_list
file.close()
file=open(r"../data/reg_poi_idx.pickle","wb")
pickle.dump(reg_idx,file) #storing_list
file.close()