Prev: adstocking GRP
Next: Proc compile
From: Eli Y. Kling on 7 Oct 2008 08:35
Hi,

In my search for a modeling approach to poison time series I came
accros a paper written by Bent Jorgensen et al (1998) with the title
"A State Space Model for Multivariate Longitudinal Count Data". their
approach is very interesting and I was wondering if someone has
developed a sas code to match?

The reason I want this is that I am tasked with comming up with an SPC
to monitor counts reflecting the amount of information flowing into
our DWH.

Any thought appriciated.

Eli
From: Oliver.Kuss on 8 Oct 2008 03:24
On 7 Okt., 14:35, "Eli Y. Kling" <Eli.Kl...(a)gmail.com> wrote:
> Hi,
>
> In my search for a modeling approach to poison time series I came
> accros a paper written by Bent Jorgensen et al (1998) with the title
> "A State Space Model for Multivariate Longitudinal Count Data". their
> approach is very interesting and I was wondering if someone has
> developed a sas code to match?
>
> The reason I want this is that I am tasked with comming up with an SPC
> to monitor counts reflecting the amount of information flowing into
> our DWH.
>
> Any thought appriciated.
>
> Eli

Dear Eli,
there is a paper of Piepho and OgutuI (The American Statistician,
August 2007, Vol. 61, No. 3) where they explain how state-space models
can be interpreted as mixed models. Respective PROC MIXED code is
also given. You might be especially interested in their example 4.3
which is modelling count data. I think the Jorgensen model is somewhat
more complicated but the paper might be a good place to start.

Hope that helps,
Oliver
 | 
Pages: 1
Prev: adstocking GRP
Next: Proc compile