-
Notifications
You must be signed in to change notification settings - Fork 19
/
feature_dig_v2.py
161 lines (135 loc) · 4.38 KB
/
feature_dig_v2.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
# -*- coding: utf-8 -*-
#V2版特征，对列表特征的编码
import pandas as pd
import numpy as np
def encoding(flist):
num_all=len(flist)
i=1
j=1
dict_re={'':0}
re_list=[]
for item in flist:
if isinstance(item,float):
re_list.append(0)
else:
list_t=item.strip().split(' ')
list_t.sort()
key=' '.join(list_t)
if key in dict_re:
re_list.append(dict_re[key])
else:
re_list.append(j)
dict_re[key]=j
j+=1
i+=1
if i%1000000==0:
print('转化已完成:%.2f%%'%(i*100.0/num_all))
print('转化完成')
return np.array(re_list)
def count_encoding(flist):
num_all=len(flist)
i=1
dict_re={0:0}
for item in flist:
if item!=0:
if item in dict_re:
dict_re[item]+=1
else:
dict_re[item]=1
i+=1
if i%1000000==0:
print('统计字典:%.2f%%'%(i*100.0/num_all))
print('统计完成,开始记录')
re_list=[]
i=1
for item in flist:
re_list.append(dict_re[item])
return np.array(re_list)
def digEncoding():
df_all=pd.read_csv('data/origin/userFeature_detail.csv')
df_out=pd.DataFrame({'uid':df_all['uid'].values})
#对兴趣编码
for i in range(1,6):
print('统计interest%d'%i)
e=df_all['interest%d'%i].values
e=encoding(e)
df_out['i%d_e'%i]=e
del e
#对关键词编码
for i in range(1,4):
print('统计kw%d'%i)
e=df_all['kw%d'%i].values
e=encoding(e)
df_out['k%d_e'%i]=e
del e
#对主题编码
for i in range(1,4):
print('统计topic%d'%i)
e=df_all['topic%d'%i].values
e=encoding(e)
df_out['t%d_e'%i]=e
del e
os_e=df_all.os.values
ct_e=df_all.ct.values
mar_e=df_all.marriageStatus.values
del df_all
print('开始对os数据进行转化')
os_e=encoding(os_e)
df_out['os_e']=os_e
del os_e
print('开始对ct数据进行转化')
ct_e=encoding(ct_e)
df_out['ct_e']=ct_e
del ct_e
print('开始对婚姻数据进行转化')
mar_e=encoding(mar_e)
df_out['mar_e']=mar_e
del mar_e
print('开始保存')
df_out.to_csv('data/extra/EncodingFeature.csv',index=False)
#----分割线，对id进行次数编码，而非直接利用（效果待验证）----
def digEncoding_C():
print('开始加载特征')
df_out=pd.read_csv('data/extra/EncodingFeature.csv')
for i in range(1,6):
print('计数转化interest%d'%i)
df_out['i%d_e'%i]=count_encoding(df_out['i%d_e'%i].values)
#对关键词编码
for i in range(1,4):
print('计数转化kw%d'%i)
df_out['k%d_e'%i]=count_encoding(df_out['k%d_e'%i].values)
#对主题编码
for i in range(1,4):
print('计数转化topic%d'%i)
df_out['t%d_e'%i]=count_encoding(df_out['t%d_e'%i].values)
print('开始保存')
df_out.to_csv('data/extra/EncodingFeature_C.csv',index=False)
def mergeEncoding(dtype='train',EC=False):
if dtype=='train':
df_join=pd.read_csv('data/feature/train_v1.csv')
outpath='data/feature/train_v2.csv'
elif dtype=='test1':
df_join=pd.read_csv('data/feature/test1_v1.csv')
outpath='data/feature/test1_v2.csv'
elif dtype=='test2':
df_join=pd.read_csv('data/feature/test2_v1.csv')
outpath='data/feature/test2_v2.csv'
else:
print('error type')
return
print('开始拼接特征！')
if EC==True:
df_feature=pd.read_csv('data/extra/EncodingFeature_C.csv')
else:
df_feature=pd.read_csv('data/extra/EncodingFeature.csv')
df_join=pd.merge(df_join,df_feature,how='left',on='uid')#拼接position信息
del df_feature
print('拼接完成，开始保存')
df_join.to_csv(outpath,index=False)
print('保存完毕')
if __name__=='__main__':
digEncoding()
digEncoding_C()
mergeEncoding(dtype='train',EC=True)
mergeEncoding(dtype='test1',EC=True)
mergeEncoding(dtype='test2',EC=True)