Welcome to RunSubmit.com

Question

1

Hi,

I want to do a rolling window autocorrelation with by ID groups. For each ID group, I want to calculate the autocorrelation using 36 months of observations. Is there any easy way to do this?

I checked PROC EXPAND, but it seems to have no such function.

Thank you very much.

flag	asked Oct 31 at 13:20 SAS-L♦♦ 375●5

Sam · Answer 1 · 2009-11-02 15:49:49Z UTC

You can calculate the sample autocorrelations (acf) using the proc arima identify statement like shown[http://support.sas.com/onlinedoc/913/getDoc/en/etsug.hlp/arima_index.htm][1]

proc arima data=sashelp.air;
identify var=air nlag=36 outcov=correlations;
run;quit;

You can accomplish the group processing using the a by statement on your variable that identifies the groups.

Could you explain why you want a rolling window acf?

Sam

Toloc · Answer 2 · 2009-11-02 12:49:17Z UTC

0

Could you provide us a sample of your input dataset ?

answered Nov 2 at 12:49

Toloc
475●6

Rolling Window Autocorrelation

Remember to vote up questions/answers you find interesting or helpful (requires 15 reputation points)

2 Answers