Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: RE: st: Constraints for simultaneous equations model


From   Beat Hintermann <bhintermann@arec.umd.edu>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   RE: RE: st: Constraints for simultaneous equations model
Date   Thu, 17 Apr 2008 08:55:45 -0400

This is a good point.

I'm trying to estimate the dependency of prices in 350 small markets on one another and on exogenous variables. I'm using a distance matrix to weigh the impact of each market on all others (a 350x350 matrix). Which is the systematic dependency you may be referring to. In the end I will probably only estimate one single parameter for the inter-market dependencies, as all coefficients could be a linear function of distance.

Naturally, I can cut off the dependency if the distance is greater than some arbitrary threshold. This will introduce more zeros in my coefficients, but not reduce the number of equations. I don't even know if Stata can handle such large systems. So what would you suggest I do?




________________________________________
From: owner-statalist@hsphsun2.harvard.edu [owner-statalist@hsphsun2.harvard.edu] On Behalf Of Maarten buis [maartenbuis@yahoo.co.uk]
Sent: Thursday, April 17, 2008 2:04 AM
To: statalist@hsphsun2.harvard.edu
Subject: RE: RE: st: Constraints for simultaneous equations model

I thought it was possible to define constraints in matrix format,
though I doubt it would be more convenient. The easiest way to define
that many constraints is in a -forvalues- loop. If you have that many
constraints than they probably follow some sort of system, and you can
make use of that in a loop. Are you very very sure that you have 300
equation? My first reaction to that is that you are doing something
very wrong, or you are trying to do something that can be achieved much
much more easily in another way.

-- Maarten

--- Beat Hintermann <bhintermann@arec.umd.edu> wrote:

> Bob
>
> Imagine you have to specify several hundred constraints like that.
> Instead of writing them down one by one I'm looking for a way to do
> this in matrix notation. But maybe that's just not possible in Stata.
> I can always program it by hand in Matlab.
>
> Beat
> ________________________________________
> From: owner-statalist@hsphsun2.harvard.edu
> [owner-statalist@hsphsun2.harvard.edu] On Behalf Of Robert A Yaffee
> [bob.yaffee@nyu.edu]
> Sent: Wednesday, April 16, 2008 9:49 PM
> To: statalist@hsphsun2.harvard.edu
> Subject: Re: RE: st: Constraints for simultaneous equations model
>
> Beat,
>    Suppose you have reg3 (y1 y2 x1 x2 x3) (y2 y1 x1 x2 x3) and you
> want to drop out x1 from the first y1 equation, you could specify
> this as follows:
>
> constraint 1 [x1] y1 = 0
> reg3 (y1 y2 x1 x2 x3) (y2 y1 x1 x2 x3) , 3sls constr(1)
>
> This sets up the constraint of x1 = 0.
>
>         Bob Yaffee
>
>
> Robert A. Yaffee, Ph.D.
> Research Scientist
> Silver School of Social Work
> New York University
>
>
> ----- Original Message -----
> From: Beat Hintermann <bhintermann@arec.umd.edu>
> Date: Wednesday, April 16, 2008 9:33 pm
> Subject: RE: st: Constraints for simultaneous equations model
> To: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
>
>
> > I wasn't using anything specific yet, but probably will end up
> using reg3.
> >
> > In any case, I'm aware of the conditions necessary for
> identification.
> > My system will end up being over-identified because I have a whole
> lot
> > of exclusion restrictions. My problem is how to enter these
> > restrictions in a reasonably efficient way. My system actually
> > contains over 300 equations, not the 10 I used in the example, so
> > typing them in individually is simply unfeasible.
> >
> >
> > ________________________________________
> > From: owner-statalist@hsphsun2.harvard.edu
> > [owner-statalist@hsphsun2.harvard.edu] On Behalf Of Robert A Yaffee
> [bob.yaffee@nyu.edu]
> > Sent: Wednesday, April 16, 2008 9:26 PM
> > To: statalist@hsphsun2.harvard.edu
> > Subject: Re: st: Constraints for simultaneous equations model
> >
> > Beat,
> >     Are you using reg3?   You need to check to be sure the rank
> condition
> > is satisfied.
> >           Regards,
> >               Bob Yaffee
> >
> >
> > Robert A. Yaffee, Ph.D.
> > Research Scientist
> > Silver School of Social Work
> > New York University
> >
> >
> > ----- Original Message -----
> > From: Beat Hintermann <bhintermann@arec.umd.edu>
> > Date: Wednesday, April 16, 2008 6:22 pm
> > Subject: st: Constraints for simultaneous equations model
> > To: "statalist@hsphsun2.harvard.edu"
> <statalist@hsphsun2.harvard.edu>
> >
> >
> > > I wonder whether there is an efficient way to define constrains
> for
> > a
> > > simultaneous equation model.
> > >
> > > For example, assume I have 10 endogenous variables. For
> identification
> > > purposes I need to place at least 9*10/2=45 restrictions on the
> > > parameters. Is there a way to do this in matrix form, as opposed
> to
> > > defining 45 individual constraints?
> > >
> > > Any help much appreciated
> > >
> > > Beat
> > >
> > > *
> > > *   For searches and help try:
> > > *   http://www.stata.com/support/faqs/res/findit.html
> > > *   http://www.stata.com/support/statalist/faq
> > > *   http://www.ats.ucla.edu/stat/stata/
> > *
> > *   For searches and help try:
> > *   http://www.stata.com/support/faqs/res/findit.html
> > *   http://www.stata.com/support/statalist/faq
> > *   http://www.ats.ucla.edu/stat/stata/
> >
> > *
> > *   For searches and help try:
> > *   http://www.stata.com/support/faqs/res/findit.html
> > *   http://www.stata.com/support/statalist/faq
> > *   http://www.ats.ucla.edu/stat/stata/
> *
> *   For searches and help try:
> *   http://www.stata.com/support/faqs/res/findit.html
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>
> *
> *   For searches and help try:
> *   http://www.stata.com/support/faqs/res/findit.html
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>


-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

http://home.fsw.vu.nl/m.buis/
-----------------------------------------


      ___________________________________________________________
Yahoo! For Good helps you make a difference

http://uk.promotions.yahoo.com/forgood/
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index