Create censoring variables                                      [ssa6: STB-22]
--------------------------

  ^censor^ ident_var time_var [ order_var ] ^=^ exp [ ^if^ exp ]  [ ^in^ range ] ^,^

	 	^gen^erate^(^new_var^)^ [ ^bef^ore^(^exp^)^ ]


creates two new variables useful in survival analysis with time-varying
regressors.  The first new variable, ^new_var^, is a zero/one variable that is
set to one in the period at the end of which the individual changes state (as
defined by the ^= exp^, which is interpreted as a logical expression).  The
second new variable, ^Tnew_var^, is the time at censoring.  Both variables are
set to missing in periods after the individual changes states.  The data are
assumed to consist of multiple observations on several individuals. Within each
individual, observations are assumed to be sorted by ^time_var^ and (optionally)
^order_var^.





Options
-------

^gen^erate^(^new_var^)^ is not optional.  It specifies the names of the censoring
    variable (^new_var^) and the time-to-censoring variable (^Tnew_var^).

^bef^ore^(^exp^)^ defines an additional condition that may censor the observations
    on an individual prior to the last observation on that individual.



Examples
--------

 . censor id time = married==1, gen(firstmar) before(out_migr==1)

 . censor id t_diagn = illness==1, gen(cancer) before(illness==2 | illness==3)


Author
------

        Dr. Philippe Bocquier, CERPOD
        bocquier@@orstom.orstom.fr


Also see
--------

 Manual:  [5d] survival
    STB:  ssa6 (STB-22)
On-line:  ^help^ for ^firstocc^, ^slice^, and ^tmerge^