python-selected-topics/python-pctr/data_processing.py at master · slxiao/python-selected-topics · GitHub

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
import pandas as pd
import lightgbm as lgb
import numpy as np
from sklearn.metrics import log_loss
import sys
import sklearn
from datetime import datetime
import pytz
import time

def timestamp_datetime(value):
    fmt = '%Y-%m-%d %H:%M:%S'
    utc_dt = pytz.utc.localize(datetime.utcfromtimestamp(float(value)))
    sh_tz = pytz.timezone("Asia/Shanghai")
    sh_dt = utc_dt.astimezone(sh_tz)
    return sh_dt.strftime(fmt)

def drop_duplicate(data):
    return data.drop_duplicates()

def instance_time(data):
    def map_hour(x):
        if x >= 7 and x<= 12:
            return 1
        elif x >= 13 and x <= 20:
            return 2
        return 3
    data.loc[:,'time'] = data.context_timestamp.apply(timestamp_datetime)
    data.loc[:,'day'] = data.time.apply(lambda x: int(x[8:10]))
    data.loc[:,'hour'] = data.time.apply(lambda x: int(x[11:13]))
    data.loc[:,'hour_seg'] = data.hour.apply(map_hour)
    return data

def user_query_counts(data):
    user_query_day = data.groupby(['user_id', 'day']).size().reset_index().rename(columns={0: 'user_query_counts_day'})
    data = pd.merge(data, user_query_day, 'left', on=['user_id', 'day'])
    user_query_day_hour = data.groupby(['user_id', 'day', 'hour']).size().reset_index().rename(columns={0: 'user_query_counts_dayhour'})
    data = pd.merge(data, user_query_day_hour, 'left', on=['user_id', 'day', 'hour'])
    return data

def user_already_queried_item(data):
    counts = data.groupby(['item_id', "user_id", "context_timestamp"]).size().reset_index().rename(columns={0: 'counts'})
    counts.sort_values(["user_id", "item_id", "context_timestamp"], inplace=True)
    counts['same_item']=counts['item_id'].diff().eq(0)
    counts['same_user']=counts['user_id'].diff().eq(0)
    counts['user_already_queried_item']=counts.same_item & counts.same_user
    counts.drop(columns=["counts", "same_item", "same_user"], inplace=True)
    data = pd.merge(data, counts, 'left', on=['item_id', 'user_id', 'context_timestamp'])
    return data

def item_total_query_counts(data):
    counts = data.groupby(['item_id']).size().reset_index().rename(columns={0: 'item_total_query_counts'})
    return data.merge(counts, 'left', on=['item_id'])

def item_total_trade_counts(data):
    counts = data[data.is_trade == 1].groupby(['item_id']).size().reset_index().rename(columns={0: 'item_total_trade_counts'})
    data = data.merge(counts, 'left', on=['item_id'])
    data.item_total_trade_counts.fillna(0, inplace=True)
    return data

def item_total_trade_rate(data):
    data['item_total_trade_rate'] = data.item_total_trade_counts / data.item_total_query_counts
    return data

def further_processing_on_train(data):
    data = item_total_query_counts(data)
    data = item_total_trade_counts(data)
    data = item_total_trade_rate(data)
    data = user_already_queried_item(data)
    return data

def user_already_queried_item2(data, train):
    counts1 = train.groupby(['item_id', "user_id", "context_timestamp"]).size().reset_index().rename(columns={0: 'counts'})
    counts2 = data.groupby(['item_id', "user_id", "context_timestamp"]).size().reset_index().rename(columns={0: 'counts'})
    counts = counts1.append(counts2)
    counts.sort_values(["user_id", "item_id", "context_timestamp"], inplace=True)
    counts['same_item']=counts['item_id'].diff().eq(0)
    counts['same_user']=counts['user_id'].diff().eq(0)
    counts['user_already_queried_item']=counts.same_item & counts.same_user
    counts.drop(columns=["counts", "same_item", "same_user"], inplace=True)
    data = pd.merge(data, counts, 'left', on=['item_id', 'user_id', 'context_timestamp'])
    return data

def item_total_query_counts2(data, train):
    counts = train.groupby(["item_id"]).size().reset_index().rename(columns={0:"item_total_query_counts"})
    data = data.merge(counts, 'left', on=['item_id'])
    data.item_total_query_counts.fillna(1, inplace=True)
    return data

def item_total_trade_counts2(data, train):
    counts = train[train.is_trade == 1].groupby(['item_id']).size().reset_index().rename(columns={0: 'item_total_trade_counts'})
    data = data.merge(counts, 'left', on=['item_id'])
    data.item_total_trade_counts.fillna(0, inplace=True)
    return data

def further_processing_on_test(data, train):
    data = item_total_query_counts2(data, train)
    data = item_total_trade_counts2(data, train)
    data = item_total_trade_rate(data)
    data = user_already_queried_item2(data, train)
    return data

def basic_processing(data, lbl):
    def map_user_star(star):
        if star in [-1, 3000]:
            return 1
        elif star in [3009, 3010]:
            return 3
        return 2

    print "basic processing on time features"
    data = instance_time(data)

    print "basic processing on user features"
    data.loc[:,'user_gender'] = data['user_gender_id'].apply(lambda x: 1 if x == -1 else 2)
    data.loc[:,'user_occupation'] = data['user_occupation_id'].apply(lambda x: 1 if x == -1 or x == 2003 else 2)
    data.loc[:,'user_star'] = data['user_star_level'].apply(map_user_star)
    data.loc[:,'user_age'] = data['user_age_level'].apply(lambda x: 1 if x in  [1004, 1005, 1006, 1007] else 2)
    data.loc[:,'user_id'] = lbl.fit_transform(data["user_id"])
    data = user_query_counts(data)

    print "basic processing on item features"
    data['item_property_len'] = data['item_property_list'].map(lambda x: len(str(x).split(';')))
    for i in range(10):
        data.loc[:,'item_property_' + str(i)] = lbl.fit_transform(data['item_property_list'].map(lambda x: str(str(x).split(';')[i]) if len(str(x).split(';')) > i else ''))
    data['item_category_len'] = data['item_category_list'].map(lambda x: len(str(x).split(';')))
    for i in range(1, 3):
        data.loc[:,'item_category_' + str(i)] = lbl.fit_transform(data['item_category_list'].map(lambda x: str(str(x).split(';')[i]) if len(str(x).split(';')) > i else ''))
    for col in ['item_id', 'item_brand_id', 'item_city_id']:
        data.loc[:,col] = lbl.fit_transform(data[col])

    print "basic processing on shop features"
    data.loc[:,"shop_id"] = lbl.fit_transform(data["shop_id"])
    data.loc[:,'shop_score_delivery'] = data['shop_score_delivery'].apply(lambda x: 0 if x <= 0.98 and x >= 0.96  else 1)

    print "basic processing on context features"
    data.loc[:,'context_page'] = data['context_page_id'].apply(lambda x: 1 if x in [4001, 4002, 4003, 4004, 4007] else 2)
    data.loc[:,'context_predict_category_property_len'] = data['predict_category_property'].map(lambda x: len(str(x).split(';')))
    for i in range(5):
        data.loc[:,'context_predict_category_property_' + str(i)] = lbl.fit_transform(data['predict_category_property'].map(
            lambda x: str(str(x).split(';')[i]) if len(str(x).split(';')) > i else ''))
    return data