-
Notifications
You must be signed in to change notification settings - Fork 7
/
topNscore.py
167 lines (145 loc) · 5.83 KB
/
topNscore.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
# encoding: utf-8
import os
import sys
import pandas as pd
import numpy as np
from numpy import array
import time
import calendar
import math
from progress.bar import Bar
N = 10
sample_num = 100
target_ratio = 0.1 # 第25天权重降至比例
delta_time_weight = 25/(-math.log(target_ratio)) * 24 * 3600
test_data = pd.read_csv('./test_data.txt',header=None,sep='\t').sample(n=sample_num, random_state=0, axis=0)
train_data = pd.read_csv('./train_data.txt',header=None,sep='\t')
poi_vec_data = pd.read_csv('./net_POI_vec.txt',header=None,sep=' ')[[x for x in range(101)]]
time_vec_data = pd.read_csv('./net_time_vec.txt',header=None,sep=' ')[[x for x in range(101)]]
region_vec_data = pd.read_csv('./net_reg_vec.txt',header=None,sep=' ')[[x for x in range(101)]]
user_vec = pd.DataFrame(columns=['userID','VenueId','Time_Slot','Time','Region']+[x for x in range(100)])
poi_vec_dict = {}
time_vec_dict = {}
region_vec_dict = {}
month_dict = dict((v,k) for k,v in enumerate(calendar.month_abbr))
def get_time_slot(time_str):
return time_str[11:13]
def get_timestamp(time_read):
week = time_read[0:3]
month = time_read[4:7]
day = time_read[8:10]
hour = time_read[11:13]
minute = time_read[14:16]
second = time_read[17:19]
year = time_read[-4:]
try:
format_time = year+'-'+str(month_dict[month])+'-'+day+' '+hour+':'+minute+':'+second
ts = time.strptime(format_time, "%Y-%m-%d %H:%M:%S")
time_ = time.mktime(ts)
print_str = str(int(time_)) + '\t' + str(hour) + '\n'
return time_
except:
print(time_read)
def get_region(loc_str):
location = loc_str[1:-1]
locations = location.split(',')
state, country = locations[3], locations[4]
if len(locations) is 5:
city = locations[2]
else: # len(locations) is 6:
city = locations[-4] + locations[-3]
region = ''.join(city.split(' ')) + '_' + ''.join(state.split(' ')) + '_' + ''.join(country.split(' '))
return region
def write_in_file(f_out_path,str_out):
f_out = open(f_out_path, 'w+')
f_out.write(str_out)
f_out.close()
def get_dicts():
print('get_dicts')
global poi_vec_dict
global time_vec_dict
global region_vec_dict
poi_vec_dict = eval(open('./poi_vec_dict.txt','r').read())
time_vec_dict = eval(open('./time_vec_dict.txt','r').read())
region_vec_dict = eval(open('./region_vec_dict.txt','r').read())
def get_user_vec():
print('get_user_vec')
global user_vec
user_vec = pd.read_csv('./user_vec.txt',header=0,sep='\t')
def dicts_gen():
print('dicts_gen')
# poi_vec_dict
for index, row in poi_vec_data.iterrows():
if row[0] not in poi_vec_dict:
poi_vec_dict[row[0]] = np.array(row[1:])
else:
print('poi_vec_dict重复:',row[0])
write_in_file('./poi_vec_dict.txt',str(poi_vec_dict))
# time_vec_dict
for index, row in time_vec_data.iterrows():
if int(row[0]) not in time_vec_dict:
time_vec_dict[int(row[0])] = np.array(row[1:])
else:
print('time_vec_dict重复:',row[0])
write_in_file('./time_vec_dict.txt',str(time_vec_dict))
# region_vec_dict
for index, row in region_vec_data.iterrows():
if row[0] not in region_vec_dict:
region_vec_dict[row[0]] = np.array(row[1:])
else:
print('region_vec_dict重复:',row[0])
write_in_file('./region_vec_dict.txt',str(region_vec_dict))
def user_vec_gen():
# print('user_vec_gen')
user_vec['userID'] = test_data[1]
user_vec['VenueId'] = test_data[3]
user_vec['Time_Slot'] = test_data[2].apply(get_time_slot)
user_vec['Time'] = test_data[2].apply(get_timestamp)
user_vec['Region'] = test_data[5].apply(get_region)
train_data['Time'] = train_data[2].apply(get_timestamp)
bar = Bar('user_vec_gen', max=sample_num)
for index, row in user_vec.iterrows():
vec = np.zeros(100)
pois_time = train_data[(train_data['Time']<row['Time'])&(train_data[1]==row['userID'])][[3,'Time']]
# 方法1：除以delta_time_weight
for index_, row_ in pois_time.iterrows():
p_v = np.array(poi_vec_data[poi_vec_data[0]==row_[3]][poi_vec_data.columns[1:]].iloc[0])
vec = vec + math.exp(-(row['Time']-row_['Time'])/delta_time_weight) * p_v
user_vec.loc[index,[x for x in range(100)]] = vec
bar.next()
bar.finish()
user_vec.to_csv('./user_vec.txt',sep='\t',header=True,index=0)
def topNscore():
print('topNscore')
hit_num = 0
sample_num = 0
for index, row in user_vec.iterrows():
sample_num += 1
userID = row['userID']
poiID = row['VenueId']
time_slot = row['Time_Slot']
region = row['Region']
userID_vec = np.array(row[5:])
# # poiID_vec = np.array(poi_vec_data[poi_vec_data[0]==poiID][poi_vec_data.columns[1:]].iloc[0])
# time_slot_vec = np.array(time_vec_data[time_vec_data[0]==int(time_slot)][time_vec_data.columns[1:]].iloc[0])
# region_vec = np.array(region_vec_data[region_vec_data[0]==region][region_vec_data.columns[1:]].iloc[0])
time_slot_vec = time_vec_dict[int(time_slot)]
region_vec = region_vec_dict[region]
topN = [('',-sys.maxsize-1)] * N
# for index_, row_ in poi_vec_data.iterrows():
for key, value in poi_vec_dict.items():
# poi_vec = np.array(row_[1:])
score = (userID_vec+region_vec+time_slot_vec).dot(value)
topN.sort(key=lambda x:x[1])
if topN[0][1] < score:
topN[0] = (key, score)
topN_pois = [x[0] for x in topN]
if poiID in topN_pois:
hit_num += 1
print('第',sample_num,'个样本',hit_num,sample_num, '{:.3}'.format(hit_num/sample_num))
if __name__ == '__main__':
# dicts_gen()
get_dicts()
# user_vec_gen()
get_user_vec()
topNscore()