import numpy as np import pandas as pd import scipy.stats as stats #统计模块 import scipy

import pymysql #导入数据库模块

from datetime import datetime #时间模块 import statsmodels.formula.api as smf # OLS regression

import pyreadr # read RDS file

from matplotlib import style import matplotlib.pyplot as plt #画图模块 import matplotlib.dates as mdates

from matplotlib.font_manager import FontProperties #作图中文 from pylab import mpl

输出矢量图 渲染矢量图

%matplotlib inline %config InlineBackend.figure_format='svg'

from IPython.core.interactiveshell import InteractiveShell #jupyter运行输出的模块 #显示每一个运行结果 InteractiveShell.ast_node_interactivity='all'

#设置列不限制数量 pd.set_option('display.max_columns',None) data = pd.read_csv("D:/python-homework/python-homework/000001.csv")#导入数据 data['Day']=pd.to_datetime(data['Day'],format='%Y/%m/%d')#变成时间格式 data.set_index('Day',inplace=True)#生成索引 data.sort_values(by=['Day'],axis=0,ascending=True)#升序排序 data_new=data['1995-01':'2024-09'].copy()#用copy来进行深度复制 data_new['Close']=pd.to_numeric(data_new['Close'])#变成数值型 data_new['Preclose']=pd.to_numeric(data_new['Preclose'])#变成数值型 #计算回报率 data_new['Raw_return']=data_new['Close']/data_new['Preclose']-1 data_new Month_data=data_new.resample('ME')['Raw_return'].apply(lambda x:(1+x).prod()-1).to_frame() Month_data inflation=pd.read_csv("D:/python-homework/python-homework/国内生产总值GDP累计同比.csv") inflation.dropna(axis=0, how='any', inplace=True)#删去缺失值 inflation['date']=pd.to_datetime(inflation['date'],format='%Y-%m-%d')#变成时间格式 inflation.set_index('date',inplace=True)#设置成索引 inflation.sort_values(by=['date'],axis=0,ascending=True)#升序排序 market_variance = data_new.resample('ME').apply( { 'Raw_return':lambda x:sum(x**2) } ) market_variance.reset_index(inplace=True)#把索引给去掉 market_variance.rename(columns={'Day':'month','Raw_return':'RV'},inplace=True)#把名字给换掉 market_variance.set_index('month',inplace=True)#用月当作索引 market_variance reg_data=pd.merge(Month_data,inflation,left_index=True,right_index=True,how='inner') reg_data=pd.merge(reg_data,market_variance,left_index=True,right_index=True,how='inner') reg_data reg_data['rate'].describe().round(3)#保留3位小数 #看峰度和偏度 reg_data.skew().round(3)#偏度 reg_data.kurt().round(3)#峰度 fig, ax=plt.subplots(figsize=(14,8))

ax.plot(inflation['rate'],label='Inflation',#图片数据 linestyle='-',#图片类型， color='#D98719',#图片颜色 linewidth=4,#图片线宽 marker='o') ax.set_ylabel('Inflation') #添加图例 ax.legend(loc='upper right',fontsize=12) plt.show();

画多条直线图

fig, ax1=plt.subplots(figsize=(14,8)) ax1.plot(reg_data['Raw_return'].shift(1),label='Return', linestyle='-',#图片类型， color='red',#图片颜色 linewidth=4,#图片线宽 marker='o') ax1.set_ylabel('Return',color='red') data_format=mdates.DateFormatter('%Y') ax1.xaxis.set_major_formatter(data_format) ax1.xaxis.set_major_locator(mdates.YearLocator()) plt.xticks(rotation=90) ax2=ax1.twinx()#把图纸镜像转一下 ax2.plot(reg_data['rate'],label='inflation', linestyle='-',#图片类型， color='blue',#图片颜色 linewidth=4,#图片线宽 marker='o') ax2.set_ylabel('Inflation',color='blue')

plt.title("China's Return and Inflation") #提取图例 lines,labels=ax1.get_legend_handles_labels() lines2,labels2=ax2.get_legend_handles_labels() ax2.legend(lines+lines2,labels+labels2,loc='upper right')#合并到第二张图

plt.show(); reg_data['lRV'] = reg_data['RV'].shift(1) reg_data['lrate']=reg_data['rate'].shift(1)#有滞后性 model_forecast_rv = smf.ols('RV ~ lrate', data=reg_data).fit( cov_type='HAC' , cov_kwds={'maxlags':6} ) #左边的是被解释变量，右边的是解释变量 print(model_forecast_rv.summary()) reg_data['lRV'] = reg_data['RV'].shift(1) reg_data['lrate']=reg_data['rate'].shift(1) model_forecast_rv = smf.ols('RV ~ lrate', data=reg_data['2000':]).fit( cov_type='HAC' , cov_kwds={'maxlags':6} )
print(model_forecast_rv.summary()) reg_data['lRV'] = reg_data['RV'].shift(1) model_forecast_rv = smf.ols('RV ~ lRV', data=reg_data['2000':]).fit( cov_type='HAC' , cov_kwds={'maxlags':6} ) print(model_forecast_rv.summary()) model_forecast_rv = smf.ols('RV ~ lRV+lrate', data=reg_data['2000':]).fit( cov_type='HAC' , cov_kwds={'maxlags':6} ) print(model_forecast_rv.summary()) reg_data_new=reg_data['2000':].copy() reg_data_new['fitted_RV']=model_forecast_rv.fittedvalues

model_forecast_return=smf.ols('Raw_returnfitted_RV',data=reg_data_new).fit( cov_type='HAC' , cov_kwds={'maxlags':6} ) print(model_forecast_return.summary()) model_rate=smf.ols('Raw_returnlrate',data=reg_data['2000':]).fit( cov_type='HAC' , cov_kwds={'maxlags':6} ) print(model_rate.summary()) print('hello world')