以上证指数为例学习pandas数据分析_python中如何应用上证180指数为基准数据分析-CSDN博客

本文链接：https://blog.csdn.net/cowardserge/article/details/104954846

本文通过分析上证指数，探讨盘中交易、每日回报和隔夜回报策略，利用支持向量回归构建预测模型。数据显示盘中交易策略表现优于其他策略，模型虽然有一定预测作用，但效果有限。文章还讨论了模型升级、翻转模型以及不同时间段的测试结果。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

进来学习人民邮电出版社出版的《Python机器学习指南》，第7章为股票市场预测。特参考预测上证指数，发现与文章中标普500趋势差别很大，市场不同，股票波动也不一样。

通过分析可知：
1、自2010-01-01至今，上证指数波动较大；
2、盘中交易模型策略，明显优于每日回报和隔夜交易回报；
3、策略构建模型有一定预测作用，但效果不明显；

本文章仅供学习DataFrame数据操作，以及预测模型构建。未完待续！

一、主要股指

import tushare as ts
ts.get_index()

	code	name	change	open	preclose	close	high	low	volume	amount
0	000001	上证指数	-1.83	2792.3221	2779.6407	2728.7563	2815.8737	2728.7563	291159852	3186.1139
1	000002	Ａ股指数	-1.83	2926.1598	2912.8667	2859.4864	2950.8481	2859.4864	290911924	3184.7548
2	000003	Ｂ股指数	-0.91	223.8562	222.9788	220.9449	226.0923	220.8130	247928	1.3591
3	000008	综合指数	-1.89	2624.3784	2615.6766	2566.3254	2638.1342	2565.1782	58775863	618.0305
4	000009	上证380	-1.61	4635.3816	4601.6229	4527.5488	4690.9859	4527.5407	64660678	688.4413
5	000010	上证180	-2.15	7979.1711	7944.9118	7774.3879	8059.9412	7774.3879	91047723	1314.0773
6	000011	基金指数	-1.12	6096.8121	6078.7292	6010.9243	6157.6370	5996.7194	176965959	484.9572
7	000012	国债指数	0.00	180.7271	180.7106	180.7158	180.7271	180.6389	710175	7.1898
8	000016	上证50	-2.23	2695.4241	2685.5907	2625.7470	2724.3019	2625.6003	37681392	661.8735
9	000017	新综指	-1.83	2359.4264	2348.7081	2305.6608	2379.3315	2305.6608	287887091	3025.1833
10	000300	沪深300	-1.98	3729.6853	3709.6822	3636.2565	3775.8506	3636.2565	155037547	2441.2716
11	000905	中证500	-1.78	5257.9438	5212.0619	5119.4716	5333.1204	5119.4716	164286604	1605.3606
12	399001	深证成指	-1.70	10294.9550	10202.7530	10029.5680	10479.7830	10029.5680	45258080567	5183.7388
13	399002	深成指R	-1.70	12629.8200	12516.7070	12304.2440	12856.5660	12304.2440	19749137662	2811.7789
14	399003	成份Ｂ指	-2.42	4914.2920	4888.0810	4769.6800	5006.8100	4762.6150	10605034	0.4797
15	399004	深证100R	-1.68	5715.5180	5671.0870	5575.9920	5819.3920	5574.7890	6297467580	1187.9062
16	399005	中小板指	-1.05	6692.9000	6616.3500	6546.8130	6850.9000	6546.8130	21684891350	2288.3061
17	399006	创业板指	-1.60	1941.3010	1917.7020	1887.0420	1980.7890	1887.0420	11913560022	1752.0276
18	399008	中小300	-1.35	1291.8000	1277.8930	1260.6520	1318.8290	1260.6520	11579259839	1484.3824
19	399100	新指数	-1.64	7856.6660	7787.0980	7659.3310	7990.9410	7659.3310	44510130185	5157.1885
20	399101	中小板综	-1.31	9754.3890	9659.9810	9533.4390	9939.2100	9533.4390	21684891350	2288.3061
21	399106	深证综指	-1.55	1720.5270	1704.7360	1678.2480	1750.6060	1678.2480	45258080567	5183.7388
22	399107	深证Ａ指	-1.55	1800.1780	1783.6340	1755.9270	1831.6840	1755.9270	45239025191	5183.0058
23	399108	深证Ｂ指	-1.73	858.8660	856.0570	841.2740	870.7270	840.2250	19055376	0.7330
24	399333	中小板R	-1.05	7534.8670	7448.6870	7370.4010	7712.7430	7370.4010	4507355897	814.0477
25	399606	创业板R	-1.60	2048.0590	2023.1620	1990.8160	2089.7190	1990.8160	3025419853	647.9224

二、导入上证指数历史

import pandas as pd
import numpy as np
import pandas_datareader as pdr
cons = ts.get_apis()#建立连接

start_date = pd.to_datetime('2010-01-01')
stop_date = pd.to_datetime('2020-03-12')
shx = ts.bar('000001',conn=cons,asset='INDEX',start_date='2010-01-01',end_date="")
shx = shx.dropna()#删除所有null行
shx.sort_values(['datetime'],inplace=True)#日期顺序
shx

	code	open	close	high	low	vol	amount	p_change
datetime
2010-01-05	000001	3254.47	3282.18	3290.51	3221.46	1422749.0	1.653394e+11	1.18
2010-01-06	000001	3277.52	3254.21	3295.87	3253.04	1351312.0	1.602844e+11	-0.85
2010-01-07	000001	3253.99	3192.78	3268.82	3176.71	1452287.0	1.605338e+11	-1.89
2010-01-08	000001	3177.26	3196.00	3198.92	3149.02	1146025.0	1.248136e+11	0.10
2010-01-11	000001	3301.61	3212.75	3306.75	3197.33	1615021.0	1.825149e+11	0.52
...	...	...	...	...	...	...	...	...
2020-03-12	000001	2936.02	2923.49	2944.47	2906.28	3077784.0	3.282092e+11	-1.52
2020-03-13	000001	2804.23	2887.43	2910.88	2799.98	3664504.0	3.930197e+11	-1.23
2020-03-16	000001	2897.30	2789.25	2898.03	2784.66	3518786.0	3.756270e+11	-3.40
2020-03-17	000001	2796.28	2779.64	2826.91	2715.22	3061496.0	3.230128e+11	-0.34
2020-03-18	000001	2792.32	2728.76	2815.87	2728.76	2911598.0	3.186114e+11	-1.83

2479 rows × 8 columns

三、收盘价绘制数据

import matplotlib.pyplot as plt
%matplotlib inline

fig, ax = plt.subplots(figsize=(15,10))
shx['close'].plot(color='black')
plt.title('SHX', fontsize=20)

在这里插入图片描述

四、数据分析

1.收盘价每日变动

shx['daily change'] = pd.Series(shx['close']- shx['open'])
shx['daily change']

datetime
2010-01-05     27.71
2010-01-06    -23.31
2010-01-07    -61.21
2010-01-08     18.74
2010-01-11    -88.86
               ...  
2020-03-12    -12.53
2020-03-13     83.20
2020-03-16   -108.05
2020-03-17    -16.64
2020-03-18    -63.56
Name: daily change, Length: 2479, dtype: float64

shx

	code	open	close	high	low	vol	amount	p_change	daily change
datetime
2010-01-05	000001	3254.47	3282.18	3290.51	3221.46	1422749.0	1.653394e+11	1.18	27.71
2010-01-06	000001	3277.52	3254.21	3295.87	3253.04	1351313.0	1.602844e+11	-0.85	-23.31
2010-01-07	000001	3253.99	3192.78	3268.82	3176.71	1452287.0	1.605338e+11	-1.89	-61.21
2010-01-08	000001	3177.26	3196.00	3198.92	3149.02	1146025.0	1.248136e+11	0.10	18.74
2010-01-11	000001	3301.61	3212.75	3306.75	3197.33	1615021.0	1.825149e+11	0.52	-88.86
...	...	...	...	...	...	...	...	...	...
2020-03-12	000001	2936.02	2923.49	2944.47	2906.28	3077784.0	3.282092e+11	-1.52	-12.53
2020-03-13	000001	2804.23	2887.43	2910.88	2799.98	3664504.0	3.930197e+11	-1.23	83.20
2020-03-16	000001	2897.30	2789.25	2898.04	2784.66	3518786.0	3.756270e+11	-3.40	-108.05
2020-03-17	000001	2796.28	2779.64	2826.91	2715.22	3061496.0	3.230128e+11	-0.34	-16.64
2020-03-18	000001	2792.32	2728.76	2815.87	2728.76	2911598.0	3.186114e+11	-1.83	-63.56

2479 rows × 9 columns

shx['daily change'].sum()#历史收盘价变动总数

7698.730000000007

shx['close'].sum()-shx['open'].sum()#每天收盘价减去开盘价，每日波动，求和

7698.730000000447

shx[shx['daily change']>200]['daily change'].sum()#日浮动大于200点日期，求和

547.3500000000004

shx[shx['daily change']>200]#日浮动大于200点日期

	code	open	close	high	low	vol	amount	p_change	daily change
datetime
2015-06-30	000001	4006.75	4277.22	4279.97	3847.88	7091766.0	9.415246e+11	5.53	270.47
2015-07-09	000001	3432.45	3709.33	3748.48	3373.54	6569146.0	6.733110e+11	5.76	276.88

shx[shx['daily change']<-200]['daily change'].sum()#日浮动大于-200点日期，求和

-1730.150000000001

shx[shx['daily change']<-200]#日浮动大于-200点日期

	code	open	close	high	low	vol	amount	p_change	daily change
datetime
2015-05-28	000001	4943.74	4620.27	4986.50	4614.24	7829646.0	1.247926e+12	-6.50	-323.47
2015-06-19	000001	4689.93	4478.36	4744.08	4476.50	4526896.0	6.854582e+11	-6.42	-211.57
2015-06-26	000001	4399.93	4192.87	4456.90	4139.53	5652178.0	7.878357e+11	-7.40	-207.06
2015-06-29	000001	4289.77	4053.03	4297.47	3875.05	6737863.0	9.042714e+11	-3.34	-236.74
2015-07-27	000001	3985.57	3725.56	4051.16	3720.44	5560032.0	7.212981e+11	-8.48	-260.01
2015-08-18	000001	3999.13	3748.16	4006.34	3743.39	5437708.0	7.224672e+11	-6.15	-250.97