用Python实现误差修正模型ECM

用R语言的教程比较多，此教程用Python实现，只看误差修正模型的直接看最后

平稳性检验

#导入数据
import pandas as pd
from statsmodels.tsa import stattools
df = pd.read_csv('XXXXXX.csv')
df.set_index('Date',inplace=True)

假设要做平稳性检验的两列，列名为’A’,‘B’，单位根ADF检验

from statsmodels.stats.diagnostic import unitroot_adf
a = unitroot_adf(df['A'])
b = unitroot_adf(df['B'])
print("A平稳性",a)
print("B平稳性",b)
#一阶差分平稳性
a1 = unitroot_adf(df['A'].diff())
b1 = unitroot_adf(df['B'].diff())
print("A一阶差分平稳性",a)
print("B一阶差分平稳性",b)

协整检验

from statsmodels.tsa.stattools import coint
df=df.dropna()
print(coint(df['A']