*
*  Example 20.5
*  J. Wooldridge, Econometrics of Cross Section and Panel Data
*
*  Note: The sample data files include all required data transformations.
*        However, in many cases we show how these data transformations would
*        be created in RATS from the basic data, as this is something you will
*        need to know how to do in practice.
*
open data recid.raw
data(format=free,org=columns) 1 1445 black alcohol drugs super married felon workprg property $
  person priors educ rules age tserved follow durat cens ldurat
*
* Define the linear parametric part of the hazard function. The free parameters
* will be in the vector b.
*
frml(regressors,vector=b) zfrml
# workprg priors tserved felon alcohol drugs black married educ age constant
*
nonlin b alpha
compute b=%const(0.0),alpha=1.0
*
* The data are censored at varying durations. The "cens" variable is a dummy
* which shows whether a particular observation was censored. The likelihood
* element takes different forms for censored and uncensored observations.
*
* Note that the log likelihood value doesn't quite match with that shown in the
* text. If the (alpha-1) in the second line of the formula is replaced by just
* alpha, you'll get the same results as shown in the text.
*
frml weibull = (z=zfrml),(w=-exp(z)*durat**alpha),$
  %if(cens,w,z+log(alpha)+(alpha-1)*log(durat)+w)
maximize(method=bfgs,iters=200) weibull