我需要估计最适合我的数据的ARMA(p,q)模型的顺序。我计划使用auto.arima()函数。在实际将此功能应用于实际数据之前,我在模拟的ARMA(2,2)模型上对其进行了测试,但是auto.arima告诉我ARMA(0,0)是最佳拟合模型。我在做错什么吗?
我的R代码是:
arma.sim<-arima.sim(model=list(ar=c(.9,-.2),ma=c(-.7,.1)),n=100)
auto.arima(arma.sim, ic = "bic", stepwise = FALSE, stationary = TRUE,
trace=TRUE, approximation = FALSE, d=0)
,输出为:
ARIMA(0,0,0) with zero mean : 285.6369
ARIMA(0,0,0) with non-zero mean : 290.2392
ARIMA(0,0,1) with zero mean : 290.211
ARIMA(0,0,1) with non-zero mean : 294.8129
ARIMA(0,0,2) with zero mean : 294.5108
ARIMA(0,0,2) with non-zero mean : 299.1143
ARIMA(0,0,3) with zero mean : 299.1079
ARIMA(0,0,3) with non-zero mean : 303.7112
ARIMA(0,0,4) with zero mean : 303.4891
ARIMA(0,0,4) with non-zero mean : 308.0936
ARIMA(0,0,5) with zero mean : 306.2664
ARIMA(0,0,5) with non-zero mean : 310.8631
ARIMA(1,0,0) with zero mean : 290.2075
ARIMA(1,0,0) with non-zero mean : 294.8092
ARIMA(1,0,1) with zero mean : Inf
ARIMA(1,0,1) with non-zero mean : Inf
ARIMA(1,0,2) with zero mean : 298.9777
ARIMA(1,0,2) with non-zero mean : 303.581
ARIMA(1,0,3) with zero mean : Inf
ARIMA(1,0,3) with non-zero mean : Inf
ARIMA(1,0,4) with zero mean : 307.8321
ARIMA(1,0,4) with non-zero mean : 312.4367
ARIMA(2,0,0) with zero mean : 294.4807
ARIMA(2,0,0) with non-zero mean : 299.0844
ARIMA(2,0,1) with zero mean : 298.9845
ARIMA(2,0,1) with non-zero mean : 303.5879
ARIMA(2,0,2) with zero mean : Inf
ARIMA(2,0,2) with non-zero mean : Inf
ARIMA(2,0,3) with zero mean : Inf
ARIMA(2,0,3) with non-zero mean : Inf
ARIMA(3,0,0) with zero mean : 299.0557
ARIMA(3,0,0) with non-zero mean : 303.659
ARIMA(3,0,1) with zero mean : 303.5876
ARIMA(3,0,1) with non-zero mean : 308.1908
ARIMA(3,0,2) with zero mean : Inf
ARIMA(3,0,2) with non-zero mean : Inf
ARIMA(4,0,0) with zero mean : 303.5521
ARIMA(4,0,0) with non-zero mean : 308.1561
ARIMA(4,0,1) with zero mean : 308.0647
ARIMA(4,0,1) with non-zero mean : 312.6687
ARIMA(5,0,0) with zero mean : 307.3978
ARIMA(5,0,0) with non-zero mean : 311.999
Best model: ARIMA(0,0,0) with zero mean
Series: arma.sim
ARIMA(0,0,0) with zero mean
sigma^2 estimated as 0.9728: log likelihood=-140.52
AIC=283.03 AICc=283.07 BIC=285.64