st: Plotting estimated survival functions from two different models

[Dami Olajide <damiolajide@gmail.com>]

Dear Statalist,

I need help to plot estimated survival functions from two different
models (weibull and weibull+selection (dursel)) on a single graph,
with the latter using the approach proposed by Boehmke et al (2005). I
have estimated the two models and the respective predicted hazards
saved respectively. I tried to use a sytax from Gutierrez (2002)
comparing 'parametric fraility and shared fraility survival models,
but it did not work. I have 23,901 observations. I followed the idea
of 'set obs'

set obs 47000
replace _t = _n - 23901 in 23902/47000

then

from the weibull model;
predict hazard_weibull in 23902/47000, hazard

and from the duration with selection model;
predict hazard_dursel  in 23902/47000, hazard

then

graph hazard_weibull hazard_dursel _t in 23902/47000, c(11[-]) s(ii) sort ///
    l1title ("pop hazard") ///
    key1(c(1) "weibull") key2(c(1[-]) "log normal")

I received an error message that "class member function not found".

Can more a experienced STATA user please help?

Thanks

Deane

References:

Boehmke, F., Morey, D., and Shannon, M. (2005). Selection bias and
continuous-time duration models: consequences anf proposed solution.
American Journal of Political Science. Vol. 50 (1): 192-207.

Gutierrez, R.G. (2002). Parametric fraility and shared fraility
survival models. The Stata Journal, 2 (1):22-44.

*
*   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/