Date:         Mon, 2 Jul 2007 21:43:30 -0400
Reply-To:     Wensui Liu <liuwensui@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Wensui Liu <liuwensui@GMAIL.COM>
Subject:      Re: simultaneous equation model questions
Comments: To: rong.guo@gmail.com
In-Reply-To:  <1183386311.779863.323810@w5g2000hsg.googlegroups.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

yes, I think you can still use proc model, or proc nlmixed.

On 7/2/07, rong.guo@gmail.com <rong.guo@gmail.com> wrote:
> On Jun 28, 4:10 pm, shilin...@yahoo.com wrote:
> > On Jun 28, 12:20 pm, rong....@gmail.com wrote:
> >
> >
> >
> > > Greetings!
> >
> > > I am doing a project that needs to run asimultaneousequation model.
> > > Since this is the first time I deal withsimultaneousequation, I have
> > > some questions and any help on this is highly appreciated!
> >
> > > Let me briefly introduce what I am doing here. There are two decision
> > > variables: "percentage to buy" and "premium", which will be the
> > > dependent variable in theequations, so there are two linearequations
> > > in thissimultaneousmodel. "percentage to buy" and "premium" each
> > > depends on some variables, and at the same time, they also depend on
> > > each other.
> >
> > > I listed all the variables in theequationsto illustrate the
> > > relationship. As you can see, premium and pct_buy are the dependent
> > > variable in one equation, but they are also independent variable in
> > > the other equation. And the 2equationsshare a common independent
> > > variable "offer", and they each also has its own independent
> > > variable.
> >
> > > equation 1:
> > > dependent variable: pct_buy
> > > independent variables: offer, liquidity, premium
> >
> > > equation 2:
> > > dependent variable: premium
> > > independent variables: offer, size, pct_buy
> >
> > > I need to run asimultaneousequation to reflect this scenario. I am
> > > thinking that PROC CALIS might be able to help me, and tried to set up
> > > a model, but I am sure it is not complete. Can anyone please take a
> > > look and let me know if I am on the right track, and give me some
> > > hints on improving the model setup?
> >
> > > proc calis data= bank.offer;
> > > var premium pct_buy offer size liquidity;
> > > lineqs
> > > premium = b1 offer + b2 size + b3 pct_buy + e1,
> > > pct_buy = b4 offer + b5 liquidity + b6 premium + e2;
> > > run;
> >
> > > Thanks so much!
> >
> > > Rong
> >
> > I am not sure for proc calis , but I am sure the proc model can do it
> > -- wthich is in SAS/ETS pakage.
> >
> > HTH
>
> Thanks for your reply. I looked over PROC MODEL, and looks like it
> handles nonlinear equations. The 2 equations in my model are either
> both OLS models, or one is OLS, and the other is PROBIT. In such case,
> can I still use PROC MODEL?
>
> THANKS!
>


--
WenSui Liu
A lousy statistician who happens to know a little programming
(http://spaces.msn.com/statcompute/blog)