How to interpret the Output of the following ARIMA model

by Sam Gladio   Last Updated April 16, 2018 07:19 AM

How to interpret the output of the below ARIMA(with Fourier Terms) code

y <- msts(ts(dataAR$Total[57:331]), seasonal.periods=c(30.4375,91.3125))

Model

fit <- auto.arima(y, xreg=cbind(fourier(y, K=c(2,1)),
                                dataAR$V1[57:331],
                                dataAR$V2[57:331],
                                dataAR$V3[57:331],
                                dataAR$V4[57:331]),
                  lambda = 0,
                  seasonal=FALSE)

summary(fit)

Output:

Series: y
Regression with ARIMA(1,1,2) errors
Box Cox transformation: lambda= 0

Coefficients: ar1 ma1 ma2 S1-30 C1-30 S2-30 C2-30 S1-91 C1-91 -0.2001 -0.4897 -0.4525 -0.0955 0.2572 0.0039 0.1151 -0.1259 0.1926 0.0035 -3.9984 -1.8375 0.0354
s.e. 0.1978 0.1773 0.1655 0.0834 0.0902 0.0850 0.0776 0.0917 0.0869 0.2699 0.1146 0.2829 0.2098

sigma^2 estimated as 0.5642: log likelihood=-304.84
AIC=637.68 AICc=639.3 BIC=688.26

Training set error measures:
ME RMSE MAE MPE MAPE MASE ACF1
Training set -2.644095 130.8606 81.77106 -62.16161 91.01262 0.7943549 0.06562703



Related Questions


Predict energy/heating load via temperature

Updated June 17, 2017 15:19 PM



Problem defining ARIMA order

Updated March 22, 2017 15:19 PM

tbats() model not capturing seasonality (weekly data)

Updated February 02, 2018 21:19 PM