Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: RE: Stata equivalent of SAS "Proc Traj"?


From   Stas Kolenikov <skolenik@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: RE: Stata equivalent of SAS "Proc Traj"?
Date   Fri, 13 Feb 2009 14:23:53 -0600

The more I learn SAS, the more I like Stata... I too doubt there would
be anybody who'd be interested in rewriting any of SAS IML stuff in
Mata, unless an automatic translator is available. Bill Gould showed
on a number of occasions how easy it is to convert Numerical Recipes
Fortran or C code to Mata. I imagine similar conversion might be
doable with PROC IML or with R.

Some of the functionality you are looking for might be available
through GLLAMM. The mixture component of PROC TRAJ is doable through
the discrete support of the latent variable GLLAMM feature, and of
course all the multilevel/latent coefficients power is out there
immediately. Learning GLLAMM is an undertaking though. You'd have to
study the manuals for a day before your first command produces any
reasonable output, and another three days to figure out how that
discrete mixture model can be specified, with all the identification
hurdles. If you are good in Stata/Mata programming, you might be able
to write your own -proctraj- command by then :))

There is also a finite mixture model -fmm- command by Partha Deb. You
should be able to stick linear regression inside of it, but not a
random coefficients model.

On 2/13/09, David Bell <dcbell@iupui.edu> wrote:
> ***
>  When I took Nagin's Developmental Trajectory workshop, I asked Jones to
> convert the program for Stata.  He did not show the requisite enthusiasm for
> redoing something he had already done for SAS.
>
>  I would be extremely enthusiastic if someone would take up the task for
> Stata.
>
>
> > I don't think an equivalent Stata procedure has been written, although I
> could be wrong. Jones, one of the developers of Proc Traj, has a website
> that might be useful
> (http://www.andrew.cmu.edu/user/bjones/index.htm). Also,
> given that the procedure is implemented as a SAS macro, it shouldn't be too
> hard to write equivalent code for Stata.
> >
> > JH
> >
> >
> > -----Original Message-----
> > From: owner-statalist@hsphsun2.harvard.edu
> [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of
> Ariel Linden
> > Sent: Thursday, February 12, 2009 8:37 PM
> > To: statalist@hsphsun2.harvard.edu
> > Subject: st: Stata equivalent of SAS "Proc Traj"?
> >
> > Hi all,
> >
> > There is a user-written program for SAS called "Proc Traj" which uses a
> > semi-parametric approach to categorize individuals into groups based on
> > their longitudinal trajectories.
> >
> > Is there similar procedure for us Stata folk?
> >
> > Here are the references:
> >
> > Jones, B. L., & Nagin, D. S. (2007). Advances in group-based trajectory
> > modeling and an SAS procedure for estimating them. Sociological Methods &
> > Research, 35, 542-571.
> >
> > Jones, B., Nagin, D. S., & Roeder, K. (2001). A SAS procedure based on
> > mixture models for estimating developmental trajectories. Sociological
> > Research and Methods, 29, 374-393.
> >


-- 
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   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