Call:
arima(x = dpgs, order = c(6, 0, 0), xreg = dpus, include.mean = F)
Coefficients:
ar1 ar2 ar3 ar4 ar5 ar6dpus
0.3953 0.1634 0.0946 0.0297 -0.0873 -0.0525 0.1927
s.e. 0.0389 0.0400 0.0404 0.0405 0.0400 0.0373 0.0136
sigma^2 estimated as 0.0002524: log likelihood = 1949.61, aic = -3883.21
>tratio=m4coef/sqrt(diag(m4var.coef))
>tratio
ar1ar2ar3ar4ar5ar6
10.1665214 4.0831152 2.34037
这个是自动适应参数估计的结果。模型估计为ARIMA(4,0,2),即ARMA(4,2)
系数为:
ar1 ar2 ar3 ar4 ma1 ma2
-0.5505 0.2316 0.0880 -0.4325 -0.1944 -0.5977
s.e. 0.1657 0.1428 0.1402 0.1270 0.1766 0.1732
s.e.是系数的标准差,系数显著性要自己算,|系数/se| >1.96 即 95%的置信度
sigma^2 estimated 估计值方差
log likelihood 对数似然值
(这个不用解释了吧)
AIC=709.13 AICc=710.73 BIC=725.63
再就是下面一堆误差计算
ME Mean Error
RMSE Root Mean Squared Error
MAE Mean Absolute Error
MPE Mean Percentage Error
MAPE Mean Absolute Percentage
MASE Mean Absolute Scaled Error