Proc ARIMA - by groups
in [SAS]
|
From: rsermer on 21 Jan 2010 14:03 I have a question regarding Proc ARIMA - I would like to forecast for many by groups and was wondering if there is a way to automate the model fitting process. Maybe by outputting the autocorrelations to see if they are decreasing towards zero and then outputting the p+d and q to fit into the model for each by group. Or am I oversimplifying this and each iteration needs to be reviewed for stationarity and the model fitted by hand? This is what I have for code: PROC ARIMA DATA = forecast; BY on_off region; I VAR = losses SCAN; RUN; Thanks for your help.
From: Wensui Liu on 21 Jan 2010 20:17 in your situation, you should take a look at HPF procedure. On Thu, Jan 21, 2010 at 2:03 PM, rsermer(a)gmail.com <rsermer(a)gmail.com> wrote: > I have a question regarding Proc ARIMA - I would like to forecast for > many by groups and was wondering if there is a way to automate the > model fitting process. Maybe by outputting the autocorrelations to see > if they are decreasing towards zero and then outputting the p+d and q > to fit into the model for each by group. Or am I oversimplifying this > and each iteration needs to be reviewed for stationarity and the model > fitted by hand? > > This is what I have for code: > > PROC ARIMA DATA = forecast; > BY on_off region; > I VAR = losses SCAN; > RUN; > > > Thanks for your help. > -- ============================== WenSui Liu Blog : statcompute.spaces.live.com Tough Times Never Last. But Tough People Do. - Robert Schuller ==============================
|
Pages: 1 Prev: SASware ballot #18 Proc Transpose preserving variable Next: keep text together |