st: Update to -stjm- on SSC

["CROWTHER, Michael" <mjc76@leicester.ac.uk>]

Thanks to Kit Baum a major update to -stjm- is available from SSC. This can be installed using -ssc- or updated using -adoupdate-.

-stjm- fits joint models for longitudinal and survival data under maximum likelihood.

This is a version 12 update, with a number of improvements and additions:

- computation time has been approximately halved
- adaptive quadratre routines are greatly improved. The adaptive section is executed at the beginning of each full Newton-Raphson iteration, under a convergence scheme.
- 2 new survival submodels are available, a 2-component mixture Weibull model and a mixture Weibull-exponential model
- fixed and random restricted cubic splines can now be used to model time in the longitudinal submodel through options frcs() and rrcs(), where the user specifies the degrees of freedom for the spline function.

Many of these improvements were implemented when I undertook an internship at StataCorp. I'd like to thank everyone at StataCorp for their hospitality, and in particular Yulia Marchenko, Alan Riley, Jeff Pitblado and Vince Wiggins.

The previous -stjm- for Stata 11 is now available as -stjm11- on SSC. It's important to note that the adaptive routines in -stjm11- are only implemented once at the beginning of each iteration, compared to the convergence scheme in the Stata 12 version, therefore estimates between -stjm- and -stjm11- may differ slightly.

Best wishes,
Michael


Michael Crowther
NIHR Methods Fellow in Medical Statistics
Centre for Biostatistics & Genetic Epidemiology
Department of Health Sciences
University of Leicester
Adrian Building
University Road
Leicester LE1 7RH

Tel: 0116 229 7278
E-mail: michael.crowther@le.ac.uk
Homepage: http://www2.le.ac.uk/departments/health-sciences/research/biostats/staff-pages/mjc76/
*
*   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/