A城市巡游车与网约车运营特征对比分析—数据读取及统计

初始化

import pandas as pd
import numpy as np

# 文件目录，绝对路径
INPUT_PATH = 'G:\DCIC\Data\\'

# 文件读取行数
MAX_ROWS = 100000

数据读取

巡游车GPS数据读取

taxigps2019 = pd.read_csv(INPUT_PATH + 'taxiGps20190531.csv', nrows=MAX_ROWS)
# 查看数据的统计描述
taxigps2019.describe()

	OPERATING_STATUS	GPS_SPEED	DRIVING_DIRECTION	LONGITUDE	LATITUDE
count	100000.000000	100000.000000	100000.000000	100000.000000	100000.000000
mean	2.415830	15.922179	161.451510	117.388134	24.354396
std	2.320472	22.837529	113.725946	9.273066	1.932351
min	1.000000	0.000000	0.000000	0.000000	0.000000
25%	1.000000	0.000000	59.000000	118.100982	24.480172
50%	1.000000	0.000000	167.000000	118.123175	24.493398
75%	6.000000	28.900000	261.000000	118.149498	24.516875
max	8.000000	381.300000	360.000000	129.110960	34.656481

taxigps2019.head()

	OPERATING_STATUS	GPS_SPEED	DRIVING_DIRECTION	GPS_TIME	LONGITUDE	LATITUDE	CARNO
0	1	0.0	0	2019/5/31 0:00:05	118.098451	24.493498	c3e622b913ca7085db129fa379121b90
1	1	0.0	4	2019/5/31 0:00:08	118.055053	24.564395	9b4802bcb6344a5772814557428aee85
2	1	0.0	0	2019/5/31 0:00:09	0.000000	0.000000	f7e16e0e76abf3460007f65cb1509565
3	1	0.0	0	2019/5/31 0:00:10	0.000000	0.000000	c607b7a160ab91bf05f8085ed221196f
4	1	0.8	32	2019/5/31 0:00:12	117.990452	24.569882	7be62de4645772356a3bb7190bc54c6d

taxigps2019.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 100000 entries, 0 to 99999
Data columns (total 7 columns):
 #   Column             Non-Null Count   Dtype  
---  ------             --------------   -----  
 0   OPERATING_STATUS   100000 non-null  int64  
 1   GPS_SPEED          100000 non-null  float64
 2   DRIVING_DIRECTION  100000 non-null  int64  
 3   GPS_TIME           100000 non-null  object 
 4   LONGITUDE          100000 non-null  float64
 5   LATITUDE           100000 non-null  float64
 6   CARNO              100000 non-null  object 
dtypes: float64(3), int64(2), object(2)
memory usage: 5.3+ MB

将字段类型根据取值空间进行修改，可以压缩内存使用需求

taxigps2019 = pd.read_csv(INPUT_PATH + 'taxiGps20190531.csv', nrows=MAX_ROWS,
                         dtype = {
                             'DRIVING_DIRECTION': np.uint16,
                             'OPERATING_STATUS': np.uint8,
                             'LONGITUDE': np.float32,
                             'LATITUDE': np.float32,
                             'GPS_SPEED': np.float16 
                         })

taxigps2019.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 100000 entries, 0 to 99999
Data columns (total 7 columns):
 #   Column             Non-Null Count   Dtype  
---  ------             --------------   -----  
 0   OPERATING_STATUS   100000 non-null  uint8  
 1   GPS_SPEED          100000 non-null  float16
 2   DRIVING_DIRECTION  100000 non-null  uint16 
 3   GPS_TIME           100000 non-null  object 
 4   LONGITUDE          100000 non-null  float32
 5   LATITUDE           100000 non-null  float32
 6   CARNO              100000 non-null  object 
dtypes: float16(1), float32(2), object(2), uint16(1), uint8(1)
memory usage: 2.8+ MB

对GPS数据进行排序

taxigps2019 = taxigps2019[taxigps2019.columns[::-1]] # 把列的顺序反过来
taxigps2019.sort_values(by=['CARNO','GPS_TIME'], inplace=True) # 根据车号以及GPS_TIME进行排序
taxigps2019.reset_index(inplace=True, drop=True)
taxigps2019.head()

	CARNO	LATITUDE	LONGITUDE	GPS_TIME	DRIVING_DIRECTION	OPERATING_STATUS
0	0006d282be70d06881a7513b69fcaa60	24.479755	118.146935	2019/5/31 1:31:20	292	1
1	0006d282be70d06881a7513b69fcaa60	24.479755	118.146935	2019/5/31 1:31:35	292	1
2	0006d282be70d06881a7513b69fcaa60	24.479755	118.146935	2019/5/31 1:31:50	292	1
3	0006d282be70d06881a7513b69fcaa60	24.479755	118.146935	2019/5/31 1:32:05	292	1
4	0006d282be70d06881a7513b69fcaa60	24.479755	118.146935	2019/5/31 1:32:20	292	1

将多天的数据拼接在一起

# 出租车2019年5月31日-6月1日的GPS数据
taxigps2019 = pd.concat([
    pd.read_csv(INPUT_PATH + 'taxiGps20190531.csv', nrows=MAX_ROWS,
                         dtype = {
                             'DRIVING_DIRECTION': np.uint16,
                             'OPERATING_STATUS': np.uint8,
                             'LONGITUDE': np.float32,
                             'LATITUDE': np.float32,
                             'GPS_SPEED': np.float16 
                         }),
    pd.read_csv(INPUT_PATH + 'taxiGps20190601.csv', nrows=MAX_ROWS,
                         dtype = {
                             'DRIVING_DIRECTION': np.uint16,
                             'OPERATING_STATUS': np.uint8,
                             'LONGITUDE': np.float32,
                             'LATITUDE': np.float32,
                             'GPS_SPEED': np.float16 
                         })
])
taxigps2019 = taxigps2019[taxigps2019.columns[::-1]]
taxigps2019.sort_values(by=['CARNO','GPS_TIME'], inplace=True)
taxigps2019.reset_index(inplace=True)
print(taxigps2019.head())
print(taxigps2019.tail())

   index                             CARNO   LATITUDE   LONGITUDE  \
0   7622  0006d282be70d06881a7513b69fcaa60  24.479755  118.146935   
1  11509  0006d282be70d06881a7513b69fcaa60  24.479755  118.146935   
2  15763  0006d282be70d06881a7513b69fcaa60  24.479755  118.146935   
3  19754  0006d282be70d06881a7513b69fcaa60  24.479755  118.146935   
4  23693  0006d282be70d06881a7513b69fcaa60  24.479755  118.146935   

            GPS_TIME  DRIVING_DIRECTION  GPS_SPEED  OPERATING_STATUS  
0  2019/5/31 1:31:20                292        0.0                 1  
1  2019/5/31 1:31:35                292        0.0                 1  
2  2019/5/31 1:31:50                292        0.0                 1  
3  2019/5/31 1:32:05                292        0.0                 1  
4  2019/5/31 1:32:20                292        0.0                 1  
        index                             CARNO   LATITUDE   LONGITUDE  \
199995  72693  fff20f025f560278d601b2fd47e1f6b7  24.465818  118.073380   
199996  81887  fff20f025f560278d601b2fd47e1f6b7  24.465914  118.071518   
199997  84249  fff20f025f560278d601b2fd47e1f6b7  24.465792  118.070946   
199998  92843  fff20f025f560278d601b2fd47e1f6b7  24.463242  118.070229   
199999  94174  fff20f025f560278d601b2fd47e1f6b7  24.463011  118.070023   

                GPS_TIME  DRIVING_DIRECTION  GPS_SPEED  OPERATING_STATUS  
199995  2019/6/1 0:04:10                266  26.296875                 1  
199996  2019/6/1 0:04:41                274  26.296875                 1  
199997  2019/6/1 0:04:49                214  27.203125                 1  
199998  2019/6/1 0:05:19                198  30.000000                 1  
199999  2019/6/1 0:05:24                258  27.703125                 1

巡游车订单读取

taxiorder2019 = pd.concat([
    pd.read_csv(INPUT_PATH + 'taxiOrder20190531.csv', nrows=MAX_ROWS,
                           dtype = {
                               'GETON_LONGITUDE': np.float32,
                               'GETON_LATITUDE': np.float32,
                               'GETOFF_LONGITUDE': np.float32,
                               'GETOFF_LATITUDE': np.float32,
                               'PASS_MILE': np.float16,
                               'NOPASS_MILE': np.float16,
                               'WAITING_TIME': np.float16
                           }),
    pd.read_csv(INPUT_PATH + 'taxiOrder20190601.csv', nrows=MAX_ROWS,
                           dtype = {
                               'GETON_LONGITUDE': np.float32,
                               'GETON_LATITUDE': np.float32,
                               'GETOFF_LONGITUDE': np.float32,
                               'GETOFF_LATITUDE': np.float32,
                               'PASS_MILE': np.float16,
                               'NOPASS_MILE': np.float16,
                               'WAITING_TIME': np.float16
                           })
])
taxiorder2019 = taxiorder2019.rename(columns={'CAR_NO':'CARNO'})
taxiorder2019.sort_values(by=['CARNO','GETON_DATE'], inplace=True) # 根据车号和上车时间进行排序
taxiorder2019.reset_index(inplace=True, drop=True)
print(taxiorder2019.head())
print(taxiorder2019.tail())

                              CARNO           GETON_DATE  GETON_LONGITUDE  \
0  0006d282be70d06881a7513b69fcaa60  2019-05-31 00:08:00       118.155060   
1  0006d282be70d06881a7513b69fcaa60  2019-05-31 02:30:00       118.154999   
2  0006d282be70d06881a7513b69fcaa60  2019-05-31 07:47:00       118.155083   
3  0006d282be70d06881a7513b69fcaa60  2019-05-31 08:19:00       118.146347   
4  0006d282be70d06881a7513b69fcaa60  2019-05-31 08:43:00       118.129448   

   GETON_LATITUDE          GETOFF_DATE  GETOFF_LONGITUDE  GETOFF_LATITUDE  \
0       24.506035  2019-05-31 00:13:00        118.176666        24.509895   
1       24.488607  2019-05-31 02:46:00        118.188156        24.498117   
2       24.506119  2019-05-31 08:00:00        118.181938        24.531134   
3       24.515144  2019-05-31 08:32:00        118.112679        24.520956   
4       24.496716  2019-05-31 08:56:00        118.157806        24.486753   

   PASS_MILE  NOPASS_MILE  WAITING_TIME  
0   3.099609     1.299805          62.0  
1   7.300781     0.000000         326.0  
2   4.398438    16.906250         328.0  
3   4.601562     0.000000         254.0  
4   3.300781     0.199951         438.0  
                                   CARNO           GETON_DATE  \
199995  fff20f025f560278d601b2fd47e1f6b7  2019-06-01 01:03:00   
199996  fff20f025f560278d601b2fd47e1f6b7  2019-06-01 01:19:00   
199997  fff20f025f560278d601b2fd47e1f6b7  2019-06-01 03:00:00   
199998  fff20f025f560278d601b2fd47e1f6b7  2019-06-01 10:05:00   
199999  fff20f025f560278d601b2fd47e1f6b7  2019-06-01 10:44:00   

        GETON_LONGITUDE  GETON_LATITUDE          GETOFF_DATE  \
199995       118.118797       24.428305  2019-06-01 01:07:00   
199996       118.119179       24.430992  2019-06-01 01:35:00   
199997       118.173531       24.481377  2019-06-01 03:00:00   
199998       118.158691       24.448475  2019-06-01 10:20:00   
199999       118.173271       24.463055  2019-06-01 11:00:00   

        GETOFF_LONGITUDE  GETOFF_LATITUDE  PASS_MILE  NOPASS_MILE  \
199995        118.111732        24.440104   2.000000    10.601562   
199996        118.170212        24.487490  12.203125     4.601562   
199997        118.171600        24.482180   0.099976    30.000000   
199998        118.118454        24.479401   8.898438    22.906250   
199999        118.088173        24.439388  11.796875    10.500000   

        WAITING_TIME  
199995          59.0  
199996         142.0  
199997           0.0  
199998         179.0  
199999         210.0

网约车GPS数据读取

wycgps2019 = pd.read_csv(INPUT_PATH + 'wycGps20190531.csv', nrows=MAX_ROWS,
                        dtype={
                            'LONGITUDE': np.float32,
                            'LATITUDE': np.float32,
                            'SPEED': np.float16
                        })

wycgps2019 = wycgps2019.rename(columns={'CAR_NO':'CARNO'})
wycgps2019 = wycgps2019[wycgps2019.columns[::-1]]
wycgps2019.sort_values(by=['CARNO','POSITION_TIME'], inplace=True)

wycgps2019['BIZ_STATUS'] = wycgps2019['BIZ_STATUS'].fillna(-1).astype(np.int8)
wycgps2019['ENCRYPT'] = wycgps2019['ENCRYPT'].fillna(-1).astype(np.int8)
wycgps2019.head()

	CARNO	BIZ_STATUS	SPEED	DIRECTION	ENCRYPT	LATITUDE	LONGITUDE	POSITION_TIME
88732	0006d282be70d06881a7513b69fcaa60	-1	8.0	92.0	1	24.503584	118.157280	2019/5/31 20:33:05
31769	0006d282be70d06881a7513b69fcaa60	-1	8.0	85.0	1	24.477930	118.083000	2019/5/31 23:17:35
58098	0006d282be70d06881a7513b69fcaa60	-1	5.0	267.0	1	24.474464	118.082787	2019/5/31 23:20:18
38891	0006d282be70d06881a7513b69fcaa60	-1	8.0	267.0	1	24.474428	118.080521	2019/5/31 23:20:49
23289	0006d282be70d06881a7513b69fcaa60	-1	18.0	298.0	1	24.493311	118.084457	2019/5/31 23:25:58

网约车订单数据读取

wycorder2019 = pd.read_csv(INPUT_PATH + 'wycOrder20190531.csv', nrows=MAX_ROWS,
                        dtype={
                            'DEP_LONGITUDE': np.float32,
                            'DEP_LATITUDE': np.float32,
                            'DEST_LONGITUDE': np.float32,
                            'DEST_LATITUDE': np.float32,
                        })
wycorder2019 = wycorder2019.rename(columns={'CAR_NO':'CARNO'})
wycorder2019.sort_values(by=['CARNO','DEP_TIME'], inplace=True)
wycorder2019.head()

	ORDER_ID	ON_AREA	CARNO	BOOK_DEP_TIME	DEP_LONGITUDE	DEP_LATITUDE	DEP_TIME	DEST_LONGITUDE	DEST_LATITUDE	DEST_TIME	DRIVE_MILE	DRIVE_TIME	WAIT_MILE	ORDER_MATCH_TIME
69056	P190531163151202000	350200	000ec98fb3522d1d3f3ffd9f43c6617b	20190531163100	118.183853	24.527201	20190531163743	118.164581	24.539051	20190531164441	3.8	418	0.0	20190531185034
33401	P190531190532186001	350200	000ec98fb3522d1d3f3ffd9f43c6617b	20190531190500	118.087311	24.440067	20190531191327	118.086441	24.468336	20190531193002	5.2	995	0.0	20190531214125
36568	B190531193036068000	350200	000ec98fb3522d1d3f3ffd9f43c6617b	20190531193036	118.082085	24.461121	20190531194114	118.119873	24.477873	20190531195730	5.4	976	0.0	20190531220126
58463	B190531195846071000	350200	000ec98fb3522d1d3f3ffd9f43c6617b	20190531195846	118.117836	24.476557	20190531200107	118.155640	24.508106	20190531201939	8.0	1112	0.1	20190531222130
5912	P190531202421202000	350200	000ec98fb3522d1d3f3ffd9f43c6617b	20190531202400	118.150558	24.509783	20190531202734	118.153320	24.479868	20190531203854	4.8	680	0.0	20190531230145

数据统计

对5月31日的数据进行统计

# 巡游车 GPS 数据
taxigps2019 = pd.read_csv(INPUT_PATH + 'taxiGps20190531.csv',
                         dtype = {
                             'DRIVING_DIRECTION': np.uint16,
                             'OPERATING_STATUS': np.uint8,
                             'LONGITUDE': np.float32,
                             'LATITUDE': np.float32,
                             'GPS_SPEED': np.float16 
                         })
taxigps2019 = taxigps2019[taxigps2019.columns[::-1]] # 把列的顺序反过来
taxigps2019.sort_values(by=['CARNO','GPS_TIME'], inplace=True) # 根据车号以及GPS_TIME进行排序
taxigps2019.reset_index(inplace=True, drop=True)

# 巡游车订单数据
taxiorder2019 = pd.read_csv(INPUT_PATH + 'taxiOrder20190531.csv',
                           dtype = {
                               'GETON_LONGITUDE': np.float32,
                               'GETON_LATITUDE': np.float32,
                               'GETOFF_LONGITUDE': np.float32,
                               'GETOFF_LATITUDE': np.float32,
                               'PASS_MILE': np.float16,
                               'NOPASS_MILE': np.float16,
                               'WAITING_TIME': np.float16
                           })
taxiorder2019 = taxiorder2019.rename(columns={'CAR_NO':'CARNO'})
taxiorder2019.sort_values(by=['CARNO','GETON_DATE'], inplace=True) # 根据车号和上车时间进行排序
taxiorder2019.reset_index(inplace=True, drop=True)

# 网约车 GPS 数据
wycgps2019 = pd.read_csv(INPUT_PATH + 'wycGps20190531.csv',
                        dtype={
                            'LONGITUDE': np.float32,
                            'LATITUDE': np.float32,
                            'SPEED': np.float16
                        })
wycgps2019 = wycgps2019.rename(columns={'CAR_NO':'CARNO'})
wycgps2019 = wycgps2019[wycgps2019.columns[::-1]]
wycgps2019.sort_values(by=['CARNO','POSITION_TIME'], inplace=True)
wycgps2019['BIZ_STATUS'] = wycgps2019['BIZ_STATUS'].fillna(-1).astype(np.int8)
wycgps2019['ENCRYPT'] = wycgps2019['ENCRYPT'].fillna(-1).astype(np.int8)

# 网约车订单数据
wycorder2019 = pd.read_csv(INPUT_PATH + 'wycOrder20190531.csv',
                        dtype={
                            'DEP_LONGITUDE': np.float32,
                            'DEP_LATITUDE': np.float32,
                            'DEST_LONGITUDE': np.float32,
                            'DEST_LATITUDE': np.float32,
                        })
wycorder2019 = wycorder2019.rename(columns={'CAR_NO':'CARNO'})
wycorder2019.sort_values(by=['CARNO','DEP_TIME'], inplace=True)

有多少辆巡游车：

taxigps2019['CARNO'].nunique()  # 获取唯一值的统计次数

有多少辆网约车：

wycgps2019['CARNO'].nunique()  # 获取唯一值的统计次数

巡游车平均GPS速度：

np.clip(taxigps2019['GPS_SPEED'].values, 0, 150).mean() 
# clip 将数组中的元素限制在 [a_min, a_max]，大于 a_max 的就使得它等于 a_max，小于 a_min 的就使得它等于 a_min

17.42

巡游车运营状态统计:

taxigps2019['OPERATING_STATUS'].value_counts()
# 查看表格的某列有多少个不同的值，并计算每个值在该列出现的次数

1    9540708
6    6916642
8     371497
Name: OPERATING_STATUS, dtype: int64

查看某辆巡游车的数据：

taxigps2019[taxigps2019['CARNO'] == '0006d282be70d06881a7513b69fcaa60'].head()

	CARNO	LATITUDE	LONGITUDE	GPS_TIME	DRIVING_DIRECTION	GPS_SPEED	OPERATING_STATUS
0	0006d282be70d06881a7513b69fcaa60	24.506195	118.128929	2019/5/31 10:00:11	12	9.203125	1
1	0006d282be70d06881a7513b69fcaa60	24.506210	118.128937	2019/5/31 10:00:26	10	0.000000	1
2	0006d282be70d06881a7513b69fcaa60	24.506210	118.128937	2019/5/31 10:00:41	10	0.000000	1
3	0006d282be70d06881a7513b69fcaa60	24.506210	118.128937	2019/5/31 10:00:56	10	0.000000	1
4	0006d282be70d06881a7513b69fcaa60	24.506210	118.128937	2019/5/31 10:01:11	10	0.000000	1

统计某个运行方向的车辆：

len(taxigps2019[taxigps2019['DRIVING_DIRECTION'] == 10]['CARNO'].unique())

统计最多的GPS小时（大部分数据都是几点）：

taxigps2019['GPS_TIME'] = pd.to_datetime(taxigps2019['GPS_TIME'])
taxigps2019['GPS_TIME'].dt.hour.value_counts()

23    1059242
22    1054752
10    1054140
9     1053138
21    1047899
20    1042931
8     1042160
19    1039729
17    1038851
18    1033548
7     1002225
6      903491
2      893999
4      859305
3      852674
5      816163
1      457019
11     439676
16     131282
0        2273
15       1394
14       1285
13        851
12        820
Name: GPS_TIME, dtype: int64

统计某个巡游车的在一天的载客时间：

$\sum$ 下车时间-上车时间

temp = taxiorder2019[taxiorder2019['CARNO'] == '0006d282be70d06881a7513b69fcaa60']  
dtime = pd.to_datetime(temp['GETOFF_DATE'])-pd.to_datetime(temp['GETON_DATE'])
dtime.sum()

Timedelta('0 days 11:04:00')

统计某个巡游车的在一天的有效运行时间：（以分钟为单位）

temp = taxigps2019[taxigps2019['CARNO'] == '0006d282be70d06881a7513b69fcaa60'] 
run_time = temp['GPS_TIME'].dt.hour.astype(str) +temp['GPS_TIME'].dt.minute.astype(str) 
run_time.nunique()

统计某个巡游车的行驶距离：

平均 GPS 速度 * 行驶时间

temp = taxigps2019[taxigps2019['CARNO'] == '0006d282be70d06881a7513b69fcaa60']
mean_speed  = np.clip(temp['GPS_SPEED'].values, 0, 150).mean() 
run_time = max(temp['GPS_TIME'])-min(temp['GPS_TIME'])
(run_time/ np.timedelta64(1, 'h')) * mean_speed

389.4881163194445

统计网约车订单数据上车经纬度的最大最小值：

print('上车经度最大值',end=': ')
print(taxiorder2019['GETON_LONGITUDE'].max())  
print('上车经度最小值',end=': ')
print(taxiorder2019[taxiorder2019['GETON_LONGITUDE'] > 0]['GETON_LONGITUDE'].min())  
print('上车纬度最大值',end=': ')
print(taxiorder2019['GETON_LATITUDE'].max())  
print('上车纬度最小值',end=': ')
print(taxiorder2019[taxiorder2019['GETON_LATITUDE'] > 0]['GETON_LATITUDE'].min())  
print('下车经度最大值',end=': ')
print(taxiorder2019['GETOFF_LONGITUDE'].max())  
print('下车经度最小值',end=': ')
print(taxiorder2019[taxiorder2019['GETOFF_LONGITUDE'] > 0]['GETOFF_LONGITUDE'].min())  
print('下车纬度最大值',end=': ')
print(taxiorder2019['GETOFF_LATITUDE'].max())  
print('下车纬度最小值',end=': ')
print(taxiorder2019[taxiorder2019['GETOFF_LATITUDE'] > 0]['GETOFF_LATITUDE'].min())

上车经度最大值: 126.43206
上车经度最小值: 108.93546
上车纬度最大值: 34.656197
上车纬度最小值: 12.239963
下车经度最大值: 126.995125
下车经度最小值: 100.401085
下车纬度最大值: 34.69632
下车纬度最小值: 12.239963

统计网约车订单数据集在中下车经纬度最常见的位置（经度+维度，各保留三位有效数字组合得到具体位置）：

f = lambda x:str(round(x,3))
taxiorder2019['GETOFF_POSITION'] = taxiorder2019['GETOFF_LONGITUDE'].apply(f)+','+taxiorder2019['GETOFF_LATITUDE'].apply(f)
taxiorder2019['GETOFF_POSITION'].value_counts()

0.0,0.0           3340
118.144,24.547    1860
118.07,24.638     1786
118.127,24.537    1731
118.116,24.481    1437
                  ... 
118.114,24.678       1
118.023,24.47        1
118.123,24.524       1
118.161,24.721       1
117.98,24.477        1
Name: GETOFF_POSITION, Length: 13267, dtype: int64

GPS 定位经纬度都为0的点是异常值（对应真实的物理坐标在非洲），可能是由于没有开启GPS定位功能、系统故障或者是没有网络造成的。

找到数据中的异常值：

切比雪夫定理：可用于判断异常值

约 75% 的数据，位于平均数两个标准差范围内

约 89% 的数据，位于平均数三个标准差范围内

约 96% 的数据，位于平均数五个标准差范围内
对于异常值可以利用插值的方法重新赋值

lb = taxiorder2019['GETOFF_LONGITUDE'].mean()-5*taxiorder2019['GETOFF_LONGITUDE'].std()  
ub = taxiorder2019['GETOFF_LONGITUDE'].mean()+5*taxiorder2019['GETOFF_LONGITUDE'].std()  
# 找到异常值
taxiorder2019[(taxiorder2019['GETOFF_LONGITUDE']>ub) | (taxiorder2019['GETOFF_LONGITUDE']<lb)]['GETOFF_LONGITUDE'] = np.nan 
# 进行插值
taxiorder2019['GETOFF_LONGITUDE'] = taxiorder2019['GETOFF_LONGITUDE'].interpolate()

A城市巡游车与网约车运营特征对比分析—数据读取及统计

初始化

数据读取

巡游车GPS数据读取

巡游车订单读取

网约车GPS数据读取

网约车订单数据读取

数据统计

悦读