Date:         Fri, 22 Jun 2012 08:47:35 -0400
Reply-To:     Rick Wicklin <Rick.Wicklin@SAS.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Rick Wicklin <Rick.Wicklin@SAS.COM>
Subject:      Re: Compute a running total

Thanks, Nat.

Just to clarify, this is not about a running MEAN. This is about a
variation of a running COUNT. However, the count only "runs" over a
certain window of length L, so it isn't a true running total.

The ARIMA idea is a good one. Can I tell ARIMA to compute a running mean
over an interval length L? If so, I could then multiple the means by L to
recover the counts. Presumably I'd need to use PROC EXPAND first?

Incidentally, I've previously written about running means and discovered a
nice little formula for a running variance. If anyone is interested, see
http://blogs.sas.com/content/iml/2012/01/18/compute-a-running-mean-and-
variance/ or short URL http://bit.ly/zdkgWK

Rick Wicklin