.-
help for ^lncure^
.-

Estimate lognormal cure model (Beta version)
------------------------------------------------------------------

     ^lncure^  depvar [varlist] [weight] [^if^ exp] [^in^ range]
			[^,^ ^tr^ ^d^ead^(^varname^)^ ^t0(^varname^)^ ^r^obust ^cl^uster^(^varname^)^
			   ^sc^ore^(^newvars^)^ ^nocons^tant ^l^evel^(^#^)^ ^nocoef^ ^nohead^er
			   maximize_options ]


^fweight^s, ^pweight^s, and ^iweight^s are allowed; see help @weights@.

This commands share the features of all estimation commands; see help @est@.

^lncure^ may be used with ^sw^ to perform stepwise estimation; see help @sw@.


^!! Predict not working at this time^
The syntax of @predict@ following ^lncure^ is

	^predict^ [type] newvarname [^if^ exp] [^in^ range] [^,^ statistic]

where statistic is

	^t^ime        predicted survival time; the default
	^lnt^ime      predicted ln(survival time)
	^hazard^      predicted hazard
	^xb^          linear prediction
	^stdp^        standard error of the linear prediction
	^s^urv        predicted S(depvar) or S(depvar|t0)
	^csn^ell      (partial) Cox-Snell residuals
	^mg^ale       (partial) martingale residuals

These statistics are available both in and out of sample; type ^predict^ ...
^if e(sample)^ ... if wanted only for the estimation sample.


Description
-----------

^lncure^ estimates maximum-likelihood log-normal distribution (survival time)
cure models. And estimates 

The dependent variable depvar represents the time of failure or censoring and
varlist represents the independent variables.  These commands allow estimation
with fixed or time-varying covariates, allow for left truncation (delayed
entry) and gaps, and may be used with single- or multiple- failure data.

Options
-------

^tr^ reports the estimated coefficients transformed to time ratio, i.e., exp(b)
    rather than b, and may not be combined with ^hazard^.  Standard errors and
    confidence intervals are similarly transformed.  Expected time estimates
    may be redisplayed with or without ^tr^.

^dead(^varname^)^ specifies the name of a variable recording 0 if the observation
    is censored and a value other than 0 -- typically 1 -- if the observation
    represents a failure.

^t0(^varname^)^ specifies the variable that records when the observation became at
    risk.  ^t0()^ can be used to handle left truncation, gaps, time-varying
    covariates, and recurring failures.  If ^t0()^ is not specified, all observa-
    tions are assumed to have become at risk at time 0.

^robust^ specifies that the Huber/White/sandwich estimator of variance is to be
    used in place of the traditional calculation.  ^robust^ combined with
    ^cluster()^ further allows observations which are not independent within
    cluster (although they must be independent between clusters).  See
    ^[U] 23.11 Obtaining robust variance estimates^.

^cluster(^varname^)^ specifies that the observations are independent across groups
    (clusters) but not necessarily independent within groups.  varname speci-
    fies to which group each observation belongs; e.g., ^cluster(personid)^ in
    data with repeated observations on individuals.  ^cluster()^ can be used with
    @pweight@s to produce estimates for unstratified cluster-sampled data.  Spec-
    ifying ^cluster()^ implies ^robust^.

^score(^newvars^)^ creates new variables containing the contributions to the
    scores; see ^[R] weibull^.  Two new variables need to be specified.  The
    first new variable will contain d(ln L-j)/d(x_jb) and the second will
    contain d(ln L_j)/d(ln_sigma).

^noconstant^ suppresses the constant term (intercept) in the model.

^level(^#^)^ specifies the confidence level for the confidence intervals of the
    coefficients; see help @level@.

^nocoef^ and ^noheader^ are for use by Stata programmers.  See ^[R] lnormal^.

maximize_options control the maximization process; see help @maximize@.  You
    should never have to specify them.


Options for @predict@
-------------------

^time^, the default, calculates the predicted survival time.  Note that this is
    the prediction from time 0 conditional on constant covariates.

^lntime^ calculates the ln() of what ^time^ produces.

^hazard^ calculates the predicted hazard.

^xb^ calculates the linear prediction.

^stdp^ calculates the standard error of the linear prediction.

^surv^ calculates the predicted S(t|t0).  If you did not specify ^t0()^ when you
    estimated the model, t0=0 and thus ^surv^ calculates the predicted survivor
    function at the time of failure or censoring, S(t).  Otherwise, it is the
    probability of surviving through t given survival through t0.  In such
    cases, you may wish to also see help for @streg@.

^csnell^ calculates the (partial) Cox-Snell residual.  If you have single obser-
    vations per subject, then ^csnell^ calculates the usual Cox-Snell residual.
    Otherwise, ^csnell^ calculates the additive contribution of this observation
    to the subject's overall Cox-Snell residual.  In such cases, you may wish
    to see help for @streg@.

^mgale^ calculates the (partial) martingale residual.  The issues are the same as
    with ^csnell^ above.  In such cases, you may wish to also see help for @streg@.


Examples
--------

	. ^lncure time drug age, dead(dead)^
	. ^lncure, tr^
	. ^lncure failtime load bearings^
	. ^lncure failtime load bearings, dead(failed) robust^
	. ^lncure tod treat bp, dead(dead) robust cluster(patid)^

Also see
--------

 Manual:  ^[U] 23 Estimation and post-estimation commands^,
	  ^[U] 29 Overview of model estimation in Stata^,
	  ^[R] streg^, ^[R] weibull^, ^[R] lnormal

On-line:  help for @est@, @postest@; @cox@, @glm@, @ltable@, @stcox@, @streg@, @sw@, @weibull@,
@lnormal@