Hello everyone, I am experiencng some problems in producing forecasts and backcast from an ARIMA(1,0,0) model. I need to produce an insample backcast and a seasonal normal backcast and forecast. I have a seasonal consumption function. By using actual data I get actual demand. By passing Seasonal Normal regressors I need to obtain seasonal normal demand.
I tested backcasting in arima by passing data used in the estimation back to the model through predict. This didnt work as the output was different from the fitted value I could see for the estimated model. In other words, in arima I am not able to produce my seasonal normal backcast. I then started looking at Arima in Forecast. By passing the estimated model to Arima I can reproduce the fitted value from the model. My problem comes when I try to apply that model to different dependent (y) data, as the y is supposed to be passed to Arima, so essentially I can do only an in sample backcast for same dependent variable. Is there a way I can get a Seasonal Normal backcast based on the model estimated on actual data but by passing Seasonal Normal regressors? Thanks, Paolo [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.