I am conducting rolling window forecast using Thailand inflation data for the periods between January 2003 and December 2014 where the length of the rolling forecast window is 36, the length of the out of sample forecast is 4 horizons and number of rolling samples is 50. The last date in the first estimation period should be December 2008. As its a training project, i am using an AR(1) model.
I have tried using the greybox package
x <- thailand
ourCall <- "predict(arima(x=data,order=c(1,0,0)),n.ahead=h)"
returnedValues2 <- ro(x, h=4, origins=50, call=ourCall, value=ourValue, ci=TRUE, co=TRUE)
the question I face is how do you select the origins , in my case i want the length of rolling window to be 36 and conduct 50 rolling samples estimations to forecast 4 horizons ahead
