如何运行 OLS 回归 pandas 日期时间对象序列是独立值 (x)

Question

我想，我缺少一些基本的东西。但是，这是我的问题。我有 25 天的时间序列数据，我想运行 OLS 回归，Y 值是时间序列，X 值是日期时间索引。这是我的代码

import pandas as pd
from pandas.stats.api import ols

indx = [Timestamp('2015-06-01 00:00:00'), Timestamp('2015-06-02 00:00:00'), Timestamp('2015-06-03 00:00:00'), Timestamp('2015-06-04 00:00:00'), Timestamp('2015-06-05 00:00:00'), Timestamp('2015-06-06 00:00:00'), Timestamp('2015-06-07 00:00:00'), Timestamp('2015-06-08 00:00:00'), Timestamp('2015-06-09 00:00:00'), Timestamp('2015-06-10 00:00:00'), Timestamp('2015-06-11 00:00:00'), Timestamp('2015-06-12 00:00:00'), Timestamp('2015-06-13 00:00:00'), Timestamp('2015-06-14 00:00:00'), Timestamp('2015-06-15 00:00:00'), Timestamp('2015-06-16 00:00:00'), Timestamp('2015-06-17 00:00:00'), Timestamp('2015-06-18 00:00:00'), Timestamp('2015-06-19 00:00:00'), Timestamp('2015-06-20 00:00:00'), Timestamp('2015-06-21 00:00:00'), Timestamp('2015-06-22 00:00:00'), Timestamp('2015-06-23 00:00:00'), Timestamp('2015-06-24 00:00:00'), Timestamp('2015-06-25 00:00:00')]
col = [51.219999999999999, 51.189999999999998, 51.210000000000001, 51.229999999999997, 51.219999999999999, 51.219999999999999, 51.219999999999999, 51.229999999999997, 51.240000000000002, 51.219999999999999, 51.200000000000003, 51.200000000000003, 51.200000000000003, 51.219999999999999, 51.219999999999999, 51.219999999999999, 51.219999999999999, 51.270000000000003, 51.280000000000001, 51.280000000000001, 51.299999999999997, 51.299999999999997, 51.280000000000001, 51.280000000000001, 51.270000000000003]
df = pd.DataFrame(col,index=indx,columns=['abc'])
sumstat = ols(y=df['abc'],x=df.index)

但是，我收到以下错误，因为索引是日期时间对象

Exception: Invalid RHS type: <class 'pandas.tseries.index.DatetimeIndex'>

Answer 1

将此添加到您的代码中：

df['jDate'] = df.index.to_julian_date()
ols(x=df.jDate, y=df.abc)

完整的样子：

import pandas as pd
from pandas.stats.api import ols

indx = [Timestamp('2015-06-01 00:00:00'), Timestamp('2015-06-02 00:00:00'), Timestamp('2015-06-03 00:00:00'), Timestamp('2015-06-04 00:00:00'), Timestamp('2015-06-05 00:00:00'), Timestamp('2015-06-06 00:00:00'), Timestamp('2015-06-07 00:00:00'), Timestamp('2015-06-08 00:00:00'), Timestamp('2015-06-09 00:00:00'), Timestamp('2015-06-10 00:00:00'), Timestamp('2015-06-11 00:00:00'), Timestamp('2015-06-12 00:00:00'), Timestamp('2015-06-13 00:00:00'), Timestamp('2015-06-14 00:00:00'), Timestamp('2015-06-15 00:00:00'), Timestamp('2015-06-16 00:00:00'), Timestamp('2015-06-17 00:00:00'), Timestamp('2015-06-18 00:00:00'), Timestamp('2015-06-19 00:00:00'), Timestamp('2015-06-20 00:00:00'), Timestamp('2015-06-21 00:00:00'), Timestamp('2015-06-22 00:00:00'), Timestamp('2015-06-23 00:00:00'), Timestamp('2015-06-24 00:00:00'), Timestamp('2015-06-25 00:00:00')]
col = [51.219999999999999, 51.189999999999998, 51.210000000000001, 51.229999999999997, 51.219999999999999, 51.219999999999999, 51.219999999999999, 51.229999999999997, 51.240000000000002, 51.219999999999999, 51.200000000000003, 51.200000000000003, 51.200000000000003, 51.219999999999999, 51.219999999999999, 51.219999999999999, 51.219999999999999, 51.270000000000003, 51.280000000000001, 51.280000000000001, 51.299999999999997, 51.299999999999997, 51.280000000000001, 51.280000000000001, 51.270000000000003]
df = pd.DataFrame(col,index=indx,columns=['abc'])
df['jDate'] = df.index.to_julian_date()
sumstat = ols(x=df.jDate, y=df.abc)

如何运行 OLS 回归 pandas 日期时间对象序列是独立值 (x)

how to run OLS regression with pandas datetime object series being independent value (x)

python

pandas

statsmodels

如何 运行 OLS 回归 pandas 日期时间对象序列是独立值 (x)

how to run OLS regression with pandas datetime object series being independent value (x)

python

pandas

statsmodels

如何运行 OLS 回归 pandas 日期时间对象序列是独立值 (x)