-
Notifications
You must be signed in to change notification settings - Fork 7
/
feature_extraction_modify.py
463 lines (399 loc) · 26.4 KB
/
feature_extraction_modify.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
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
# -*- coding: utf-8 -*-
"""
-------------------------------------------------
File Name： main
Description :
Author : joleo
date： 2018/5/8 0008
-------------------------------------------------
Change Activity:
2018/5/8 0008:
-------------------------------------------------
"""
__author__ = 'joleo'
import pandas as pd
import numpy as np
import datetime
import time
import gc
from math import radians, cos, sin, asin, sqrt
class LYFeatureExtraction(object):
def haversine1(self, lon1, lat1, lon2, lat2): # 经度1，纬度1，经度2，纬度2 （十进制度数）
"""
Calculate the great circle distance between two points
on the earth (specified in decimal degrees)
"""
# 将十进制度数转化为弧度
lon1, lat1, lon2, lat2 = map(radians, [lon1, lat1, lon2, lat2])
# haversine公式
dlon = lon2 - lon1
dlat = lat2 - lat1
a = sin(dlat / 2) ** 2 + cos(lat1) * cos(lat2) * sin(dlon / 2) ** 2
c = 2 * asin(sqrt(a))
r = 6371 # 地球平均半径，单位为公里
return c * r * 1000
def timestamp_datetime(self, value):
fmt = '%Y-%m-%d %H:%M:%S'
value = time.localtime(value)
dt = time.strftime(fmt, value)
return dt
def get_user_Y(self, data):
"""
获取用户Y值
:param data:
:return:
"""
user_id = data.groupby(['TERMINALNO'], as_index = False)['Y'].first()
return user_id
def is_holiday_week(self, day):
if day in holiday_data:
return 1
else:
return 0
def to_datetime(self, value):
fmt = '%Y-%m-%d'
value = time.localtime(value)
dt = time.strftime(fmt, value)
return dt
def test(self, data):
# holiday_data week_data
# data['TIME'] = data['TIME'].apply(lambda x: self.timestamp_datetime(x), 1)
# data['TIME'] = data['TIME'].apply(lambda x: str(x)[:10], 1)
# data = data.sort_values(by=["TERMINALNO", 'TIME'])
# data.loc[:, 'isholiday'] = data['TIME'].apply(lambda x: self.isholiday(str(x)))
res = self.user_holiday_stats(data)
# relust = self.isholiday(data['TIME'])
return res
def user_driver_time(self, data):
"""
用户时间特征特征
"""
user_time_data = data.loc[:, ['TERMINALNO', 'TRIP_ID', 'TIME']]
user_time_data['hour'] = user_time_data['TIME'].map(lambda x: datetime.datetime.fromtimestamp(x).hour)
user_time_per = user_time_data.groupby('TERMINALNO', as_index=False)['TRIP_ID'].count()
user_time_data.loc[:, 'hour_5'] = user_time_data.hour.map(lambda x: 1 if x == 5 else 0)
user_time_data.loc[:, 'hour_6'] = user_time_data.hour.map(lambda x: 1 if x == 6 else 0)
user_time_data.loc[:, 'hour_10'] = user_time_data.hour.map(lambda x: 1 if x == 10 else 0)
user_time_data.loc[:, 'hour_15'] = user_time_data.hour.map(lambda x: 1 if x == 15 else 0)
user_time_data.loc[:, 'hour_7'] = user_time_data.hour.map(lambda x: 1 if x == 7 else 0)
user_time_data.loc[:, 'hour_8'] = user_time_data.hour.map(lambda x: 1 if x == 8 else 0)
user_time_data.loc[:, 'hour_9'] = user_time_data.hour.map(lambda x: 1 if x == 9 else 0)
user_time_data.loc[:, 'hour_17'] = user_time_data.hour.map(lambda x: 1 if x == 17 else 0)
user_time_data.loc[:, 'hour_18'] = user_time_data.hour.map(lambda x: 1 if x == 18 else 0)
user_time_data.loc[:, 'hour_11'] = user_time_data.hour.map(lambda x: 1 if x == 11 else 0)
user_time_data.loc[:, 'hour_12'] = user_time_data.hour.map(lambda x: 1 if x == 12 else 0)
user_time_data.loc[:, 'hour_13'] = user_time_data.hour.map(lambda x: 1 if x == 13 else 0)
user_time_data.loc[:, 'hour_14'] = user_time_data.hour.map(lambda x: 1 if x == 14 else 0)
user_time_data.loc[:, 'hour_0'] = user_time_data.hour.map(lambda x: 1 if x == 0 else 0)
user_time_data.loc[:, 'hour_1'] = user_time_data.hour.map(lambda x: 1 if x == 1 else 0)
user_time_data.loc[:, 'hour_2'] = user_time_data.hour.map(lambda x: 1 if x == 2 else 0)
user_time_data.loc[:, 'hour_3'] = user_time_data.hour.map(lambda x: 1 if x == 3 else 0)
user_time_data.loc[:, 'hour_4'] = user_time_data.hour.map(lambda x: 1 if x == 4 else 0)
driver_time_colmns = ['hour_5', 'hour_6', 'hour_10', 'hour_15',
'hour_7', 'hour_8', 'hour_9', 'hour_17', 'hour_18',
'hour_11', 'hour_12', 'hour_13',
'hour_14', 'hour_0', 'hour_1', 'hour_2', 'hour_3',
'hour_4']
# 计算每个时间段的比例
driver_time_num = user_time_data.groupby(['TERMINALNO'], as_index=False)[driver_time_colmns].sum()
user_time_per.loc[:, 'hour_5_per'] = (driver_time_num['hour_5'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_6_per'] = (driver_time_num['hour_6'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_10_per'] = (driver_time_num['hour_10'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_15_per'] = (driver_time_num['hour_15'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_7_per'] = (driver_time_num['hour_7'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_8_per'] = (driver_time_num['hour_8'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_9_per'] = (driver_time_num['hour_9'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_17_per'] = (driver_time_num['hour_17'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_18_per'] = (driver_time_num['hour_18'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_11_per'] = (driver_time_num['hour_11'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_12_per'] = (driver_time_num['hour_12'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_13_per'] = (driver_time_num['hour_13'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_14_per'] = (driver_time_num['hour_14'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_0_per'] = (driver_time_num['hour_0'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_1_per'] = (driver_time_num['hour_1'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_2_per'] = (driver_time_num['hour_2'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_3_per'] = (driver_time_num['hour_3'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per.loc[:, 'hour_4_per'] = (driver_time_num['hour_4'] / user_time_per['TRIP_ID']).astype(np.float32)
user_time_per['TERMINALNO'] = user_time_per['TERMINALNO'].astype(np.int32)
user_time_per.drop('TRIP_ID', axis=1, inplace=True)
del user_time_data
gc.collect()
return user_time_per
def user_night_stat(self, data):
"""
夜晚开车和每天用车时长的统计
"""
# train['TIME'] = train['TIME'].apply(lambda x: self.timestamp_datetime(x), 1)
data['TIME'] = data['TIME'].apply(lambda x: datetime.datetime.fromtimestamp(x), 1)
data['TIME'] = data['TIME'].apply(lambda x: str(x)[:13], 1)
train = data.sort_values(by=["TERMINALNO", 'TIME'])
train.index = range(len(train))
train['hour'] = train.TIME.apply(lambda x: str(x)[11:13], 1)
train['hour'] = train['hour'].astype(int)
train['hour_0'] = train.hour.apply(lambda x: 1 if x == 0 else 0, 1)
train['hour_1'] = train.hour.apply(lambda x: 1 if x == 1 else 0, 1)
train['hour_2'] = train.hour.apply(lambda x: 1 if x == 2 else 0, 1)
train['hour_3'] = train.hour.apply(lambda x: 1 if x == 3 else 0, 1)
train['hour_4'] = train.hour.apply(lambda x: 1 if x == 4 else 0, 1)
train['hour_21'] = train.hour.apply(lambda x: 1 if x == 21 else 0, 1)
train['hour_22'] = train.hour.apply(lambda x: 1 if x == 22 else 0, 1)
train['hour_23'] = train.hour.apply(lambda x: 1 if x == 23 else 0, 1)
train_hour = train.groupby(['TERMINALNO', 'TIME'], as_index=False).count()
train_hour.TIME = train_hour.TIME.apply(lambda x: str(x)[:10], 1)
train_day = train_hour.groupby(
['TERMINALNO', 'TIME'], as_index=False).count()
train_hour_count = train_day.groupby('TERMINALNO')['LONGITUDE'].agg(
{"hour_count_max": "max", "hour_count_mean": "mean", "hour_count_std": "std"}).reset_index()
train_hour_first = train.groupby(
['TERMINALNO', 'TIME'], as_index=False).first()
train_hour_first.TIME = train_hour_first.TIME.apply(
lambda x: str(x)[:10], 1)
train_day_sum = train_hour_first.groupby(
['TERMINALNO', 'TIME'], as_index=False).sum()
train_day_sum['hour_count'] = train_day['LONGITUDE']
train_day_sum['night_drive_count'] = train_day_sum.apply(lambda x: x['hour_0'] + x['hour_1'] +
x['hour_2'] + x['hour_3'] + x[
'hour_4'] +
x['hour_21'] + x['hour_22'] + x[
'hour_23'], 1)
train_day_sum['night_delta'] = train_day_sum['night_drive_count'] / \
train_day_sum['hour_count']
train_day_sum['night'] = train_day_sum['night_drive_count'].apply(
lambda x: 1 if x != 0 else 0, 1)
train_night_count = train_day_sum.groupby('TERMINALNO')['night_delta'].agg(
{"night_count_max": "max", "night_count_mean": "mean", "night_count_std": "std"}).reset_index()
feature = pd.merge(
train_hour_count, train_night_count, on="TERMINALNO", how="left")
# 假期特征
data['TIME_2'] = data['TIME'].apply(lambda x: str(x)[:10], 1)
data['time_flag'] = data['TIME_2'].map(lambda x: self.is_holiday_week(x))
holiday_count = data.groupby('TERMINALNO', as_index=False)['TRIP_ID'].count()
holiday_num_sum = data.groupby(['TERMINALNO'])['time_flag'].agg(
{'holiday_num_sum': 'sum'}).reset_index()
holiday_num_per = pd.merge(holiday_count, holiday_num_sum, how='left', on='TERMINALNO')
holiday_num_per['holiday_num_per'] = holiday_num_per['holiday_num_sum'] / holiday_num_per['TRIP_ID']
holiday_num_per.drop(['TRIP_ID', 'holiday_num_sum'], axis=1, inplace=True)
# holiday_data = data[data['time_flag'] == 1]
# holiday_speed = holiday_data.groupby(['TERMINALNO'])['SPEED'].agg(
# {'holiday_mean_speed': 'mean', 'holiday_speed_std': 'std'}).reset_index()
# holiday_height = holiday_data.groupby(['TERMINALNO'])['HEIGHT'].agg(
# {'holiday_mean_height': 'mean', 'holiday_height_std': 'std'}).reset_index()
# holiday_direc = holiday_data.groupby(['TERMINALNO'])['DIRECTION'].agg(
# {'holiday_mean_direc': 'mean', 'holiday_direc_std': 'std'}).reset_index()
#
# holiday_feature = pd.merge(holiday_speed, holiday_height, how='left', on='TERMINALNO')
# holiday_feature = pd.merge(holiday_feature, holiday_direc, how='left', on='TERMINALNO')
# holiday_feature = pd.merge(holiday_feature, train_data, how='left', on='TERMINALNO')
# feature = pd.merge(train_data, holiday_num_per, how='left', on='TERMINALNO')
feature['TERMINALNO'] = feature['TERMINALNO'].astype(np.int32)
feature['hour_count_max'] = feature['hour_count_max'].astype(np.int32)
feature['hour_count_mean'] = feature['hour_count_mean'].astype(np.int32)
feature['hour_count_std'] = feature['hour_count_std'].astype(np.float32)
feature['night_count_max'] = feature['night_count_max'].astype(np.float32)
feature['night_count_mean'] = feature['night_count_mean'].astype(np.float32)
feature['night_count_std'] = feature['night_count_std'].astype(np.float32)
# feature['holiday_num_per'] = feature['holiday_num_per'].astype(np.float32)
del train_hour_count
del train_night_count
del holiday_count
del holiday_num_sum
gc.collect()
return feature
def user_driver_stat(self, data):
"""
用户驾驶行为基本特征统计
"""
# trid 特征
df_user_trid_num = data.groupby('TERMINALNO', as_index=False)['TRIP_ID'].max()
user_time_per = data.groupby('TERMINALNO', as_index=False)['TRIP_ID'].count()
# 驾驶行为特征
user_speed_stat = data.groupby('TERMINALNO')['SPEED'].agg(
{'user_speed_mean': 'mean', 'user_speed_std': 'std', 'user_speed_max': 'max'}).reset_index()
user_height_stat = data.groupby('TERMINALNO')['HEIGHT'].agg(
{'user_height_mean': 'mean', 'user_height_std': 'std'}).reset_index()
user_direction_stat = data.groupby('TERMINALNO')['DIRECTION'].agg(
{'user_direction_mean': 'mean'}).reset_index()
user_long_stat = data.groupby('TERMINALNO')['LONGITUDE'].agg(
{'user_lon_mean': 'mean', 'user_lon_std': 'std'}).reset_index()
# 峰度系数是用来反映频数分布曲线顶端尖峭或扁平程度的指标
user_kurt = data.groupby('TERMINALNO')['HEIGHT','DIRECTION'].apply(lambda x: x.kurt()).reset_index()
# user_kurt = data.groupby(['TERMINALNO', 'TRIP_ID'], as_index=False)['HEIGHT','DIRECTION'].apply(lambda x: x.kurt())#.max().reset_index()
# user_kurt_1 = user_kurt.groupby(['TERMINALNO'], as_index=False)['HEIGHT','DIRECTION'].max().reset_index()
user_kurt.rename(columns={'HEIGHT': 'user_height_kurt', 'DIRECTION': 'user_direc_kurt'}, inplace=True)
# 偏态系数是根据众数、中位数与均值各自的性质，通过比较众数或中位数与均值来衡量偏斜度的，即偏态系数是对分布偏斜方向和程度的刻画
user_skew_1 = data.groupby(['TERMINALNO','TRIP_ID'], as_index=False)['HEIGHT','DIRECTION'].skew()
user_max_skew = user_skew_1.groupby('TERMINALNO')['HEIGHT','DIRECTION'].max().reset_index()
user_min_skew = user_skew_1.groupby(['TERMINALNO'])['HEIGHT', 'DIRECTION'].min().reset_index()
user_max_skew.rename(
columns={'HEIGHT': 'user_height_max_skew', 'DIRECTION': 'user_direc_max_skew'}, inplace=True)
user_min_skew.rename(
columns={'HEIGHT': 'user_height_min_skew', 'DIRECTION': 'user_direc_min_skew'}, inplace=True)
# 合并
user_driver_stat = pd.merge(df_user_trid_num, user_speed_stat, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_height_stat, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_direction_stat, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_long_stat, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_kurt, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_max_skew, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_min_skew, how='left', on='TERMINALNO')
del user_long_stat
del df_user_trid_num
del user_speed_stat
del user_height_stat
del user_direction_stat
del user_kurt
del user_max_skew
del user_min_skew
# 类型转换
user_driver_stat['TERMINALNO'] = user_driver_stat['TERMINALNO'].astype(np.int32)
user_driver_stat['TRIP_ID'] = user_driver_stat['TRIP_ID'].astype(np.int16)
user_driver_stat['user_speed_mean'] = user_driver_stat['user_speed_mean'].astype(np.float32)
user_driver_stat['user_speed_std'] = user_driver_stat['user_speed_std'].astype(np.float32)
user_driver_stat['user_speed_max'] = user_driver_stat['user_speed_max'].astype(np.float32)
user_driver_stat['user_height_mean'] = user_driver_stat['user_height_mean'].astype(np.float32)
user_driver_stat['user_height_std'] = user_driver_stat['user_height_std'].astype(np.float32)
user_driver_stat['user_direction_mean'] = user_driver_stat['user_direction_mean'].astype(np.float32)
user_driver_stat['user_lon_mean'] = user_driver_stat['user_lon_mean'].astype(np.float32)
user_driver_stat['user_lon_std'] = user_driver_stat['user_lon_std'].astype(np.float32)
user_driver_stat['user_height_kurt'] = user_driver_stat['user_height_kurt'].astype(np.float32)
user_driver_stat['user_direc_kurt'] = user_driver_stat['user_direc_kurt'].astype(np.float32)
user_driver_stat['user_height_max_skew'] = user_driver_stat['user_height_max_skew'].astype(np.float32)
user_driver_stat['user_direc_max_skew'] = user_driver_stat['user_direc_max_skew'].astype(np.float32)
user_driver_stat['user_height_min_skew'] = user_driver_stat['user_height_min_skew'].astype(np.float32)
user_driver_stat['user_direc_min_skew'] = user_driver_stat['user_direc_min_skew'].astype(np.float32)
gc.collect()
return user_driver_stat
def user_test_driver_stat(self, data):
"""
用户驾驶行为基本特征统计
"""
# trid 特征
df_user_trid_num = data.groupby('TERMINALNO', as_index=False)['TRIP_ID'].max()
# 驾驶行为特征
user_speed_stat = data.groupby('TERMINALNO')['SPEED'].agg(
{'user_speed_mean': 'mean', 'user_speed_std': 'std', 'user_speed_max': 'max'}).reset_index()
user_height_stat = data.groupby('TERMINALNO')['HEIGHT'].agg(
{'user_height_mean': 'mean', 'user_height_std': 'std'}).reset_index()
user_direction_stat = data.groupby('TERMINALNO')['DIRECTION'].agg(
{'user_direction_mean': 'mean'}).reset_index()
user_long_stat = data.groupby('TERMINALNO')['LONGITUDE'].agg(
{'user_lon_mean': 'mean', 'user_lon_std': 'std'}).reset_index()
user_kurt = data.groupby('TERMINALNO')['HEIGHT', 'DIRECTION'].apply(lambda x: x.kurt()).reset_index()
user_kurt.rename(columns={'HEIGHT': 'user_height_kurt', 'DIRECTION': 'user_direc_kurt'}, inplace=True)
# user_kurt = data.groupby(['TERMINALNO', 'TRIP_ID'], as_index=False)['HEIGHT','DIRECTION'].apply(lambda x: x.kurt())#.max().reset_index()
# user_kurt_1 = user_kurt.groupby(['TERMINALNO'], as_index=False)['HEIGHT', 'DIRECTION'].max().reset_index()
user_kurt.rename(
columns={'HEIGHT': 'user_height_kurt', 'DIRECTION': 'user_direc_kurt'}, inplace=True)
user_skew_1 = data.groupby(['TERMINALNO','TRIP_ID'], as_index=False)['HEIGHT','DIRECTION'].skew()
user_max_skew = user_skew_1.groupby('TERMINALNO')['HEIGHT', 'DIRECTION'].max().reset_index()
user_min_skew = user_skew_1.groupby(['TERMINALNO'])['HEIGHT', 'DIRECTION'].min().reset_index()
user_max_skew.rename(
columns={'HEIGHT': 'user_height_max_skew', 'DIRECTION': 'user_direc_max_skew'}, inplace=True)
user_min_skew.rename(
columns={'HEIGHT': 'user_height_min_skew', 'DIRECTION': 'user_direc_min_skew'}, inplace=True)
# 合并
user_driver_stat = pd.merge(df_user_trid_num, user_speed_stat, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_height_stat, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_direction_stat, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_long_stat, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_kurt, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_max_skew, how='left', on='TERMINALNO')
user_driver_stat = pd.merge(user_driver_stat, user_min_skew, how='left', on='TERMINALNO')
user_driver_stat.loc[:, 'Y'] = -1
del user_long_stat
del df_user_trid_num
del user_speed_stat
del user_height_stat
del user_direction_stat
del user_kurt
del user_max_skew
del user_min_skew
# 类型转换
user_driver_stat['TERMINALNO'] = user_driver_stat['TERMINALNO'].astype(np.int32)
user_driver_stat['TRIP_ID'] = user_driver_stat['TRIP_ID'].astype(np.int16)
user_driver_stat['user_speed_mean'] = user_driver_stat['user_speed_mean'].astype(np.float32)
user_driver_stat['user_speed_std'] = user_driver_stat['user_speed_std'].astype(np.float32)
user_driver_stat['user_speed_max'] = user_driver_stat['user_speed_max'].astype(np.float32)
user_driver_stat['user_height_mean'] = user_driver_stat['user_height_mean'].astype(np.float32)
user_driver_stat['user_height_std'] = user_driver_stat['user_height_std'].astype(np.float32)
user_driver_stat['user_direction_mean'] = user_driver_stat['user_direction_mean'].astype(np.float32)
user_driver_stat['user_lon_mean'] = user_driver_stat['user_lon_mean'].astype(np.float32)
user_driver_stat['user_lon_std'] = user_driver_stat['user_lon_std'].astype(np.float32)
user_driver_stat['user_height_kurt'] = user_driver_stat['user_height_kurt'].astype(np.float32)
user_driver_stat['user_direc_kurt'] = user_driver_stat['user_direc_kurt'].astype(np.float32)
user_driver_stat['user_height_max_skew'] = user_driver_stat['user_height_max_skew'].astype(np.float32)
user_driver_stat['user_direc_max_skew'] = user_driver_stat['user_direc_max_skew'].astype(np.float32)
user_driver_stat['user_height_min_skew'] = user_driver_stat['user_height_min_skew'].astype(np.float32)
user_driver_stat['user_direc_min_skew'] = user_driver_stat['user_direc_min_skew'].astype(np.float32)
gc.collect()
return user_driver_stat
# def user_speed_kurt(self, data):
# # 速度的峰度系数
# speed_kurt = data.groupby(['TERMINALNO', 'TRIP_ID'])['SPEED'].apply(lambda x: x.kurt()).reset_index()
# speed_kurt.rename(columns={'SPEED': 'speed_kurt'}, inplace=True)
# # user_speed_max_kurt = speed_kurt.groupby(['TERMINALNO'])['speed_kurt'].agg(
# # {'user_speed_max_kurt': 'max'}).reset_index()
# user_speed_min_kurt = speed_kurt.groupby(['TERMINALNO'])['speed_kurt'].agg(
# {'user_speed_min_kurt': 'min'}).reset_index()
# user_speed_mean_kurt = speed_kurt.groupby(['TERMINALNO'])['speed_kurt'].agg(
# {'user_speed_mean_kurt': 'mean'}).reset_index()
#
# user_speed_kurt = pd.merge(user_speed_mean_kurt, user_speed_min_kurt, how='left', on='TERMINALNO')
# # user_speed_kurt = pd.merge(user_speed_min_kurt, user_speed_mean_kurt, how='left', on='TERMINALNO')
#
# return user_speed_kurt
def get_distance(self, data):
"""
经纬度转距离
"""
lon_lat_data = data.loc[:, ['TERMINALNO', 'LONGITUDE', 'LATITUDE']]
user_distance = lon_lat_data.groupby('TERMINALNO', as_index=False)[['LONGITUDE', 'LATITUDE']].first()
user_distance.loc[:, 'user_distance'] = user_distance.apply(
lambda row: self.haversine1(row['LONGITUDE'], row['LATITUDE'], 113.9177317, 22.54334333), axis=1)
user_distance.drop(['LONGITUDE', 'LATITUDE'], axis=1, inplace=True)
user_distance['TERMINALNO'] = user_distance['TERMINALNO'].astype(np.int32)
user_distance['user_distance'] = user_distance['user_distance'].astype(np.float32)
del lon_lat_data
gc.collect()
return user_distance
# def get_similar_position(self, data):
# from sklearn.cluster import KMeans
#
# similar_position_labels = self.get_distance(data)
# Kmeans_mod = KMeans(n_clusters=100)
# similar_position_labels['similar_position_labels'] = Kmeans_mod.fit(
# similar_position_labels['user_distance'].reshape(-1, 1)).labels_
# similar_position_labels.drop(['user_distance'], axis=1, inplace=True)
#
# return similar_position_labels
def user_callstate_stat(self, data):
"""
用户通话状态统计
"""
callstate_data = data.loc[:, ['TERMINALNO', 'CALLSTATE']]
user_callstate_stat = callstate_data.groupby(['TERMINALNO'], as_index=False)['CALLSTATE'].count()
# 未知状态
callstate_data['call_unknow_state'] = callstate_data['CALLSTATE'].map(lambda x: 1 if x == 0 else 0)
user_callstate_stat.loc[:, 'call_unknow_state'] = callstate_data.groupby(['TERMINALNO'], as_index=False)[
'call_unknow_state'].sum()
user_callstate_stat.loc[:, 'call_unknow_state_per'] = user_callstate_stat['call_unknow_state'] / \
user_callstate_stat['CALLSTATE']
# 打电话状态
callstate_data['user_call_num'] = callstate_data['CALLSTATE'].map(lambda x: 1 if x != 0 else 0)
user_callstate_stat.loc[:, 'user_call_num'] = callstate_data.groupby(['TERMINALNO'], as_index=False)[
'user_call_num'].sum()
user_callstate_stat.loc[:, 'user_call_num_per'] = user_callstate_stat['user_call_num'] / \
user_callstate_stat['CALLSTATE']
user_callstate_stat.drop(['CALLSTATE', 'call_unknow_state', 'user_call_num'], axis=1, inplace=True)
return user_callstate_stat
holiday_data = ['2016-09-15','2016-09-16','2016-09-17','2016-10-01','2016-10-02','2016-10-03'
, '2016-10-04','2016-10-05','2016-10-06','2016-10-07','2016-12-24','2016-12-25'
, '2016-12-03','2016-12-31','2017-01-01','2016-07-02','2016-07-03','2016-07-09','2016-07-10','2016-07-16','2016-07-17',
'2016-07-23''2016-07-24','2016-07-30','2016-07-31',
'2016-08-06','2016-08-07''2016-08-13''2016-08-14','2016-08-20','2016-08-21',
'2016-08-27','2016-08-28',
'2016-09-03','2016-09-04','2016-09-10','2016-09-11','2016-09-04','2016-09-24'
, '2016-09-25','2016-10-15','2016-10-16','2016-10-22','2016-10-23','2016-10-29'
, '2016-10-30','2016-11-05','2016-11-06','2016-11-12','2016-11-13','2016-11-19'
, '2016-11-20','2016-11-26','2016-11-27','2016-12-03','2016-12-04','2016-12-10'
, '2016-12-11','2016-12-17','2016-12-18' ]