-
Notifications
You must be signed in to change notification settings - Fork 34
/
501_clickTran_feat_all.py
132 lines (113 loc) · 6.78 KB
/
501_clickTran_feat_all.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
import numpy as np
import pandas as pd
import scipy as sp
import gc
import datetime
import time
import random
import scipy.special as special
import pickle
##########################################
# 获取click_time之前所有天的点击数和转化数, 为转化率和进行贝叶斯平滑做准备#
##########################################
print("Preprocessing...")
path = './data/'
train = pd.read_csv(path+'train_all.csv')
test = pd.read_csv(path+'test_all.csv')
data = pd.concat([train, test])
print(data['instance_id'].head())
data['period'] = data['hour'] // 12 + 1
data['period'] = data['day'] % 31 * 2 + data['period']
max_p = data['period'].max()+1
cols = data.columns.tolist()
cols.remove('instance_id')
# print("-----------------转化率（贝叶斯平滑）--------------------------")
# f = open('./smooth_parameters.txt','rb')
# smooth_para = pickle.load(f)
print("单特征........")
temp=data[['user_age_level','user_gender_id','user_occupation_id','item_id','shop_id','item_brand_id','item_city_id',
'context_page_id','period','hour','is_trade']]
# for feat_1 in ['user_age_level','user_gender_id','user_occupation_id','item_id','shop_id','item_brand_id','item_city_id',
# 'context_page_id']:
for feat_1 in ['item_id','shop_id','user_age_level']:
gc.collect()
res=pd.DataFrame()
for period in range(2,max_p):
print(feat_1,'The',period,'starting')
count=temp.groupby([feat_1]).apply(lambda x: x['is_trade'][(x['period']<period).values].count()).reset_index(name=feat_1+'_all')
count1=temp.groupby([feat_1]).apply(lambda x: x['is_trade'][(x['period']<period).values].sum()).reset_index(name=feat_1+'_1')
count_shift = temp.groupby([feat_1]).apply(lambda x: x['is_trade'][(x['period']==(period-1)).values].count()).reset_index(name=feat_1+'_shift_all')
count_shift_1 = temp.groupby([feat_1]).apply(lambda x: x['is_trade'][(x['period']==(period-1)).values].sum()).reset_index(name=feat_1+'_shift_1')
count[feat_1+'_shift_all'] = count_shift[feat_1+'_shift_all']
count[feat_1+'_shift_1'] = count_shift_1[feat_1+'_shift_1']
count[feat_1+'_1']=count1[feat_1+'_1']
count.fillna(value=0, inplace=True)
######hyper########
# hyper = smooth_para['{}_{}'.format(feat_1, period)]
# count[feat_1+'_rate_hyper'] = round((count[feat_1+'_1']+hyper[0]) / (count[feat_1+'_all'] + hyper[0] + hyper[1]), 5)
# count[feat_1+'_rate_hyper'].fillna(round(hyper[0]/(hyper[0] + hyper[1]), 5), inplace = True)
###################
count[feat_1+'_rate'] = round(count[feat_1+'_1'] / count[feat_1+'_all'], 5)
count[feat_1+'_shift_rate'] = round(count[feat_1+'_shift_1'] / count[feat_1+'_shift_all'], 5)
count['period']=period
count.drop([feat_1+'_all', feat_1+'_1', feat_1+'_shift_all', feat_1+'_shift_1'],axis=1,inplace=True)
count.fillna(value=0, inplace=True)
res=res.append(count,ignore_index=True)
data = pd.merge(data,res, how='left', on=[feat_1,'period'])
print(feat_1,' over')
del temp
gc.collect()
print("双特征........")
temp=data[['user_age_level','user_gender_id','user_occupation_id','item_id','shop_id','item_brand_id','item_city_id',
'context_page_id','period','hour','is_trade']]
# for feat_1,feat_2 in [('user_age_level','item_id'), ('user_occupation_id','item_brand_id'),
# ('item_id','item_city_id'),('item_city_id','shop_id'),('item_brand_id','shop_id')]:
for feat_1,feat_2 in [('user_age_level','item_id'),('item_id','item_city_id'),('item_brand_id','shop_id'), ('user_occupation_id','item_brand_id')]:
gc.collect()
res=pd.DataFrame()
for period in range(2,17):
print(feat_1,feat_2,'The',period,'starting')
count=temp.groupby([feat_1, feat_2]).apply(lambda x: x['is_trade'][(x['period']<period).values].count()).reset_index(name=feat_1+'_'+feat_2+'_all')
count1=temp.groupby([feat_1, feat_2]).apply(lambda x: x['is_trade'][(x['period']<period).values].sum()).reset_index(name=feat_1+'_'+feat_2+'_1')
count.fillna(value=0, inplace=True)
count[feat_1+'_'+feat_2+'_1']=count1[feat_1+'_'+feat_2+'_1']
count[feat_1+'_'+feat_2+'_rate'] = round(count[feat_1+'_'+feat_2+'_1'] / count[feat_1+'_'+feat_2+'_all'], 5)
count['period'] = period
count.drop([feat_1+'_'+feat_2+'_all', feat_1+'_'+feat_2+'_1'],axis=1,inplace=True)
count.fillna(value=0, inplace=True)
res=res.append(count,ignore_index=True)
data = data.merge(res, how='left', on=[feat_1,feat_2,'period'])
print(feat_1,feat_2,' over')
del temp
gc.collect()
print("仨特征........")
temp=data[['user_age_level','user_gender_id','user_occupation_id','item_id','shop_id','item_brand_id','item_city_id',
'context_page_id','period','hour','is_trade']]
# for feat_1,feat_2,feat_3 in [('user_gender_id','user_age_level','item_brand_id'),('user_gender_id','user_age_level','shop_id'),
# ('user_gender_id','user_occupation_id','item_brand_id'),('item_id','item_brand_id','item_city_id'),
# ('item_id','item_brand_id','context_page_id'),('item_brand_id','shop_id','context_page_id'),
# ('user_age_level','user_occupation_id','item_brand_id')]:
for feat_1,feat_2,feat_3 in [('user_gender_id','user_age_level','shop_id'),('item_id','item_brand_id','item_city_id'),
('item_id','item_brand_id','context_page_id'),('user_age_level','user_occupation_id','item_brand_id')]:
gc.collect()
res=pd.DataFrame()
for period in range(2,17):
print(feat_1,feat_2,feat_3,'The',period,'starting')
count=temp.groupby([feat_1, feat_2, feat_3]).apply(lambda x: x['is_trade'][(x['period']<period).values].count()).reset_index(name=feat_1+'_'+feat_2+'_'+feat_3+'_all')
count1=temp.groupby([feat_1, feat_2, feat_3]).apply(lambda x: x['is_trade'][(x['period']<period).values].sum()).reset_index(name=feat_1+'_'+feat_2+'_'+feat_3+'_1')
count[feat_1+'_'+feat_2+'_'+feat_3+'_1']=count1[feat_1+'_'+feat_2+'_'+feat_3+'_1']
count.fillna(value=0, inplace=True)
count[feat_1+'_'+feat_2+'_'+feat_3+'_rate'] = round(count[feat_1+'_'+feat_2+'_'+feat_3+'_1'] / count[feat_1+'_'+feat_2+'_'+feat_3+'_all'], 5)
count['period'] = period
count.drop([feat_1+'_'+feat_2+'_'+feat_3+'_all', feat_1+'_'+feat_2+'_'+feat_3+'_1'], axis=1, inplace=True)
count.fillna(value=0, inplace=True)
res=res.append(count,ignore_index=True)
data = data.merge(res, how='left', on=[feat_1,feat_2,feat_3,'period'])
print(feat_1,feat_2,feat_3,' over')
del temp
gc.collect()
print(data.columns.tolist())
print(cols)
data = data.drop(cols, axis=1)
print('经过处理后,最终维度:', data.shape)
data.to_csv(path+'501_clickTran_feat_all.csv', index=False)