[R] modeling outliers in arfima

[Koopmans, Matthijs]

Hello,

I am analyzing a time series trajectory with many outliers.

When I model the outliers using arima (TSA library),  the following program statement runs like a charm:

m3.m495<-arima(Rate,order=c(1,0,1),seasonal=list(order=c(1,0,1),period=5),
  xreg=data.frame(t74=1*(seq(Rate)==74),t107=1*(seq(Rate)==107)))

'Rate' is the outcome variable. The program generates the regular arima estimates as well as those associated with the outliers (here t=74 and 107).

The corresponding arfima statement (arfima library) follows:

m495.m4a<-arfima(Rate,order=c(1,0,1),lmodel="g", numeach=c(1,1),
 seasonal=list(order=c(1,0,1), period=5,numeach=c(1,1),
 xreg=data.frame(t74=1*(seq(Rate)==74),t107=1*(seq(Rate)==107)))

The program accepts this language and produces the regular arfima estimates, but not those associated with the outliers.

I' be greatly helped by any insights as to why this might be the case.

Regards,

Matt Koopmans