3.3.1 时间序列¶

index横坐标为日期数据
数据导入：pandas reader

3.3.2 Resample函数¶

计数、均值、方差、累加、累乘
周期转换
数据验证: for循环vs内置函数

In [ ]:

pip install pandas_datareader

Collecting pandas_datareader
  Downloading pandas_datareader-0.10.0-py3-none-any.whl (109 kB)
     ---------------------------------------- 0.0/109.5 kB ? eta -:--:--
     --- ------------------------------------ 10.2/109.5 kB ? eta -:--:--
     ---------- -------------------------- 30.7/109.5 kB 262.6 kB/s eta 0:00:01
     ------------------------ ------------ 71.7/109.5 kB 435.7 kB/s eta 0:00:01
     ------------------------------------ 109.5/109.5 kB 576.9 kB/s eta 0:00:00
Collecting lxml (from pandas_datareader)
  Obtaining dependency information for lxml from https://files.pythonhosted.org/packages/80/2e/49751104148b03ad880aaf381cc24d67b7d8f401f7d074ad7db4f6d95597/lxml-4.9.3-cp39-cp39-win_amd64.whl.metadata
  Downloading lxml-4.9.3-cp39-cp39-win_amd64.whl.metadata (3.9 kB)
Requirement already satisfied: pandas>=0.23 in c:\users\dengzl\.conda\envs\data_analysis\lib\site-packages (from pandas_datareader) (2.1.1)
Collecting requests>=2.19.0 (from pandas_datareader)
  Obtaining dependency information for requests>=2.19.0 from https://files.pythonhosted.org/packages/70/8e/0e2d847013cb52cd35b38c009bb167a1a26b2ce6cd6965bf26b47bc0bf44/requests-2.31.0-py3-none-any.whl.metadata
  Using cached requests-2.31.0-py3-none-any.whl.metadata (4.6 kB)
Requirement already satisfied: numpy>=1.22.4 in c:\users\dengzl\.conda\envs\data_analysis\lib\site-packages (from pandas>=0.23->pandas_datareader) (1.26.0)
Requirement already satisfied: python-dateutil>=2.8.2 in c:\users\dengzl\.conda\envs\data_analysis\lib\site-packages (from pandas>=0.23->pandas_datareader) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in c:\users\dengzl\.conda\envs\data_analysis\lib\site-packages (from pandas>=0.23->pandas_datareader) (2023.3.post1)
Requirement already satisfied: tzdata>=2022.1 in c:\users\dengzl\.conda\envs\data_analysis\lib\site-packages (from pandas>=0.23->pandas_datareader) (2023.3)
Collecting charset-normalizer<4,>=2 (from requests>=2.19.0->pandas_datareader)
  Obtaining dependency information for charset-normalizer<4,>=2 from https://files.pythonhosted.org/packages/b3/c5/edc62435a27b017a5826d215f25ef3ab02b8b68d37b6e64cf5b602f1b55d/charset_normalizer-3.3.0-cp39-cp39-win_amd64.whl.metadata
  Downloading charset_normalizer-3.3.0-cp39-cp39-win_amd64.whl.metadata (33 kB)
Collecting idna<4,>=2.5 (from requests>=2.19.0->pandas_datareader)
  Using cached idna-3.4-py3-none-any.whl (61 kB)
Collecting urllib3<3,>=1.21.1 (from requests>=2.19.0->pandas_datareader)
  Obtaining dependency information for urllib3<3,>=1.21.1 from https://files.pythonhosted.org/packages/26/40/9957270221b6d3e9a3b92fdfba80dd5c9661ff45a664b47edd5d00f707f5/urllib3-2.0.6-py3-none-any.whl.metadata
  Downloading urllib3-2.0.6-py3-none-any.whl.metadata (6.6 kB)
Collecting certifi>=2017.4.17 (from requests>=2.19.0->pandas_datareader)
  Obtaining dependency information for certifi>=2017.4.17 from https://files.pythonhosted.org/packages/4c/dd/2234eab22353ffc7d94e8d13177aaa050113286e93e7b40eae01fbf7c3d9/certifi-2023.7.22-py3-none-any.whl.metadata
  Using cached certifi-2023.7.22-py3-none-any.whl.metadata (2.2 kB)
Requirement already satisfied: six>=1.5 in c:\users\dengzl\.conda\envs\data_analysis\lib\site-packages (from python-dateutil>=2.8.2->pandas>=0.23->pandas_datareader) (1.16.0)
Using cached requests-2.31.0-py3-none-any.whl (62 kB)
Using cached lxml-4.9.3-cp39-cp39-win_amd64.whl (3.9 MB)
Using cached certifi-2023.7.22-py3-none-any.whl (158 kB)
Downloading charset_normalizer-3.3.0-cp39-cp39-win_amd64.whl (98 kB)
   ---------------------------------------- 0.0/98.2 kB ? eta -:--:--
   ---------------------------------------- 98.2/98.2 kB 2.7 MB/s eta 0:00:00
Downloading urllib3-2.0.6-py3-none-any.whl (123 kB)
   ---------------------------------------- 0.0/123.8 kB ? eta -:--:--
   ---------------------------------------- 123.8/123.8 kB 3.7 MB/s eta 0:00:00
Installing collected packages: urllib3, lxml, idna, charset-normalizer, certifi, requests, pandas_datareader
Successfully installed certifi-2023.7.22 charset-normalizer-3.3.0 idna-3.4 lxml-4.9.3 pandas_datareader-0.10.0 requests-2.31.0 urllib3-2.0.6
Note: you may need to restart the kernel to use updated packages.

In [ ]:

import pandas as pd
import pandas_datareader as pdr

In [ ]:

data=pdr.get_data_fred('GS10')   # 官方提示这个是用来获取美国政府5年来10年期债券的汇率
data

Out[ ]:

	GS10
DATE
2018-11-01	3.12
2018-12-01	2.83
2019-01-01	2.71
2019-02-01	2.68
2019-03-01	2.57
2019-04-01	2.53
2019-05-01	2.40
2019-06-01	2.07
2019-07-01	2.06
2019-08-01	1.63
2019-09-01	1.70
2019-10-01	1.71
2019-11-01	1.81
2019-12-01	1.86
2020-01-01	1.76
2020-02-01	1.50
2020-03-01	0.87
2020-04-01	0.66
2020-05-01	0.67
2020-06-01	0.73
2020-07-01	0.62
2020-08-01	0.65
2020-09-01	0.68
2020-10-01	0.79
2020-11-01	0.87
2020-12-01	0.93
2021-01-01	1.08
2021-02-01	1.26
2021-03-01	1.61
2021-04-01	1.64
2021-05-01	1.62
2021-06-01	1.52
2021-07-01	1.32
2021-08-01	1.28
2021-09-01	1.37
2021-10-01	1.58
2021-11-01	1.56
2021-12-01	1.47
2022-01-01	1.76
2022-02-01	1.93
2022-03-01	2.13
2022-04-01	2.75
2022-05-01	2.90
2022-06-01	3.14
2022-07-01	2.90
2022-08-01	2.90
2022-09-01	3.52
2022-10-01	3.98
2022-11-01	3.89
2022-12-01	3.62
2023-01-01	3.53
2023-02-01	3.75
2023-03-01	3.66
2023-04-01	3.46
2023-05-01	3.57
2023-06-01	3.75
2023-07-01	3.90
2023-08-01	4.17
2023-09-01	4.38

In [ ]:

type(data)

Out[ ]:

pandas.core.frame.DataFrame

In [ ]:

# 均值
type(data.index)
data.resample(rule='Y').mean()   # 每年的均值

Out[ ]:

	GS10
DATE
2018-12-31	2.975000
2019-12-31	2.144167
2020-12-31	0.894167
2021-12-31	1.442500
2022-12-31	2.951667
2023-12-31	3.796667

In [ ]:

# 计数
count = data.resample(rule='Y').count()
count

Out[ ]:

	GS10
DATE
2018-12-31	2
2019-12-31	12
2020-12-31	12
2021-12-31	12
2022-12-31	12
2023-12-31	9

In [ ]:

# 累加
sum1 = data.resample(rule='Y').sum()
sum1

Out[ ]:

	GS10
DATE
2018-12-31	5.95
2019-12-31	25.73
2020-12-31	10.73
2021-12-31	17.31
2022-12-31	35.42
2023-12-31	34.17

In [ ]:

sum1/count

Out[ ]:

	GS10
DATE
2018-12-31	2.975000
2019-12-31	2.144167
2020-12-31	0.894167
2021-12-31	1.442500
2022-12-31	2.951667
2023-12-31	3.796667

In [ ]:

# 标准差
data.resample(rule='Y').std()   # 数据样本内部的差异

Out[ ]:

	GS10
DATE
2018-12-31	0.205061
2019-12-31	0.410996
2020-12-31	0.362152
2021-12-31	0.178230
2022-12-31	0.734276
2023-12-31	0.306023

In [ ]:

# 最大
data.resample(rule='Y').std().max()

Out[ ]:

GS10    0.734276
dtype: float64

In [ ]:

# 累乘
data.resample(rule='Y').prod()

Out[ ]:

	GS10
DATE
2018-12-31	8.829600
2019-12-31	7709.579518
2020-12-31	0.129870
2021-12-31	74.203795
2022-12-31	300600.586272
2023-12-31	159859.118429

In [ ]:

# 加载一份新的数据
data2 = pdr.get_data_fred('GS5')
data2

Out[ ]:

	GS5
DATE
2018-11-01	2.95
2018-12-01	2.68
2019-01-01	2.54
2019-02-01	2.49
2019-03-01	2.37
2019-04-01	2.33
2019-05-01	2.19
2019-06-01	1.83
2019-07-01	1.83
2019-08-01	1.49
2019-09-01	1.57
2019-10-01	1.53
2019-11-01	1.64
2019-12-01	1.68
2020-01-01	1.56
2020-02-01	1.32
2020-03-01	0.59
2020-04-01	0.39
2020-05-01	0.34
2020-06-01	0.34
2020-07-01	0.28
2020-08-01	0.27
2020-09-01	0.27
2020-10-01	0.34
2020-11-01	0.39
2020-12-01	0.39
2021-01-01	0.45
2021-02-01	0.54
2021-03-01	0.82
2021-04-01	0.86
2021-05-01	0.82
2021-06-01	0.84
2021-07-01	0.76
2021-08-01	0.77
2021-09-01	0.86
2021-10-01	1.11
2021-11-01	1.20
2021-12-01	1.23
2022-01-01	1.54
2022-02-01	1.81
2022-03-01	2.11
2022-04-01	2.78
2022-05-01	2.87
2022-06-01	3.19
2022-07-01	2.96
2022-08-01	3.03
2022-09-01	3.70
2022-10-01	4.18
2022-11-01	4.06
2022-12-01	3.76
2023-01-01	3.64
2023-02-01	3.94
2023-03-01	3.82
2023-04-01	3.54
2023-05-01	3.59
2023-06-01	3.95
2023-07-01	4.14
2023-08-01	4.31
2023-09-01	4.49

In [ ]:

# 合并两个样本
data['GS5'] = data2

In [ ]:

data

Out[ ]:

	GS10	GS5
DATE
2018-11-01	3.12	2.95
2018-12-01	2.83	2.68
2019-01-01	2.71	2.54
2019-02-01	2.68	2.49
2019-03-01	2.57	2.37
2019-04-01	2.53	2.33
2019-05-01	2.40	2.19
2019-06-01	2.07	1.83
2019-07-01	2.06	1.83
2019-08-01	1.63	1.49
2019-09-01	1.70	1.57
2019-10-01	1.71	1.53
2019-11-01	1.81	1.64
2019-12-01	1.86	1.68
2020-01-01	1.76	1.56
2020-02-01	1.50	1.32
2020-03-01	0.87	0.59
2020-04-01	0.66	0.39
2020-05-01	0.67	0.34
2020-06-01	0.73	0.34
2020-07-01	0.62	0.28
2020-08-01	0.65	0.27
2020-09-01	0.68	0.27
2020-10-01	0.79	0.34
2020-11-01	0.87	0.39
2020-12-01	0.93	0.39
2021-01-01	1.08	0.45
2021-02-01	1.26	0.54
2021-03-01	1.61	0.82
2021-04-01	1.64	0.86
2021-05-01	1.62	0.82
2021-06-01	1.52	0.84
2021-07-01	1.32	0.76
2021-08-01	1.28	0.77
2021-09-01	1.37	0.86
2021-10-01	1.58	1.11
2021-11-01	1.56	1.20
2021-12-01	1.47	1.23
2022-01-01	1.76	1.54
2022-02-01	1.93	1.81
2022-03-01	2.13	2.11
2022-04-01	2.75	2.78
2022-05-01	2.90	2.87
2022-06-01	3.14	3.19
2022-07-01	2.90	2.96
2022-08-01	2.90	3.03
2022-09-01	3.52	3.70
2022-10-01	3.98	4.18
2022-11-01	3.89	4.06
2022-12-01	3.62	3.76
2023-01-01	3.53	3.64
2023-02-01	3.75	3.94
2023-03-01	3.66	3.82
2023-04-01	3.46	3.54
2023-05-01	3.57	3.59
2023-06-01	3.75	3.95
2023-07-01	3.90	4.14
2023-08-01	4.17	4.31
2023-09-01	4.38	4.49

In [ ]:

data['mean'] = (data['GS10'] + data['GS5']) / 2   # 计算10年期和5年期两个债券的均值并放进样本里
data

Out[ ]:

	GS10	GS5	mean
DATE
2018-11-01	3.12	2.95	3.035
2018-12-01	2.83	2.68	2.755
2019-01-01	2.71	2.54	2.625
2019-02-01	2.68	2.49	2.585
2019-03-01	2.57	2.37	2.470
2019-04-01	2.53	2.33	2.430
2019-05-01	2.40	2.19	2.295
2019-06-01	2.07	1.83	1.950
2019-07-01	2.06	1.83	1.945
2019-08-01	1.63	1.49	1.560
2019-09-01	1.70	1.57	1.635
2019-10-01	1.71	1.53	1.620
2019-11-01	1.81	1.64	1.725
2019-12-01	1.86	1.68	1.770
2020-01-01	1.76	1.56	1.660
2020-02-01	1.50	1.32	1.410
2020-03-01	0.87	0.59	0.730
2020-04-01	0.66	0.39	0.525
2020-05-01	0.67	0.34	0.505
2020-06-01	0.73	0.34	0.535
2020-07-01	0.62	0.28	0.450
2020-08-01	0.65	0.27	0.460
2020-09-01	0.68	0.27	0.475
2020-10-01	0.79	0.34	0.565
2020-11-01	0.87	0.39	0.630
2020-12-01	0.93	0.39	0.660
2021-01-01	1.08	0.45	0.765
2021-02-01	1.26	0.54	0.900
2021-03-01	1.61	0.82	1.215
2021-04-01	1.64	0.86	1.250
2021-05-01	1.62	0.82	1.220
2021-06-01	1.52	0.84	1.180
2021-07-01	1.32	0.76	1.040
2021-08-01	1.28	0.77	1.025
2021-09-01	1.37	0.86	1.115
2021-10-01	1.58	1.11	1.345
2021-11-01	1.56	1.20	1.380
2021-12-01	1.47	1.23	1.350
2022-01-01	1.76	1.54	1.650
2022-02-01	1.93	1.81	1.870
2022-03-01	2.13	2.11	2.120
2022-04-01	2.75	2.78	2.765
2022-05-01	2.90	2.87	2.885
2022-06-01	3.14	3.19	3.165
2022-07-01	2.90	2.96	2.930
2022-08-01	2.90	3.03	2.965
2022-09-01	3.52	3.70	3.610
2022-10-01	3.98	4.18	4.080
2022-11-01	3.89	4.06	3.975
2022-12-01	3.62	3.76	3.690
2023-01-01	3.53	3.64	3.585
2023-02-01	3.75	3.94	3.845
2023-03-01	3.66	3.82	3.740
2023-04-01	3.46	3.54	3.500
2023-05-01	3.57	3.59	3.580
2023-06-01	3.75	3.95	3.850
2023-07-01	3.90	4.14	4.020
2023-08-01	4.17	4.31	4.240
2023-09-01	4.38	4.49	4.435

In [ ]:

# 周期转换
data.resample(rule='Y')

Out[ ]:

<pandas.core.resample.DatetimeIndexResampler object at 0x000002642B876B20>

In [ ]:

data.resample(rule='Y').first()

Out[ ]:

	GS10	GS5	mean
DATE
2018-12-31	3.12	2.95	3.035
2019-12-31	2.71	2.54	2.625
2020-12-31	1.76	1.56	1.660
2021-12-31	1.08	0.45	0.765
2022-12-31	1.76	1.54	1.650
2023-12-31	3.53	3.64	3.585

In [ ]:

526互联

3.3-时间序列和Resample函数

3.3.1 时间序列¶

3.3.2 Resample函数¶