tolerance for convergence. Iteration stops once the relative change in
deviance is less than rel.tolerance. Default is 0.00001.
failure:
controls the return value if the algorithm fails to converge.
If failure = 1, a warning message is generated and estimates of
the parameters at the last iteration are returned. If failure = 2, an
error message is returned but no parameter estimates.
Default is failure = 1.
VALUE:
a list is returned, consisting of the three parameters, conveniently packaged
up to supply as the control argument to survreg.
EXAMPLES:
survreg(Surv(days, event) ~ voltage, data = capacitor,
control = survreg.control(maxiter = 40))