Missing configuration is resolved automatically. Repeated calls start from
the RTMB objective's env$last.par.best; each pass begins at the best state
retained by the preceding pass.
Usage
sbt_optimise(
fit,
n_passes = 2L,
control = NULL,
check = TRUE,
check_args = list()
)Arguments
- fit
An
sbt_fit()object.- n_passes
Positive number of sequential optimizer passes.
- control
Optional
nlminbcontrol list. The stored control is used when omitted.- check
Run
check_mle()after optimization.- check_args
Named arguments passed to
check_mle().
Value
The optimised sbt_fit, containing the accepted nlminb result,
complete fitted parameter state, diagnostics, and, when check = TRUE, a
payload-bound MLE validation record.
See also
Other fit workflow:
check_mcmc(),
check_mle(),
sbt_add_bounds(),
sbt_add_map(),
sbt_add_msy(),
sbt_add_parameters(),
sbt_add_priors(),
sbt_build_object(),
sbt_example_fit(),
sbt_fit(),
sbt_fit_compatibility(),
sbt_fit_io,
sbt_fit_rebuild(),
sbt_fit_report(),
sbt_fit_validate(),
sbt_fit_validation(),
sbt_mcmc(),
sbt_obj()
Examples
if (FALSE) { # \dontrun{
fit <- sbt_fit(sbt_example_data())
fit <- sbt_optimise(
fit,
n_passes = 2L,
control = list(eval.max = 10000L, iter.max = 10000L)
)
fit
} # }