boot_all_models: Refit all models several times on bootstrapped data

In courtiol/SileR: Survival Analysis of Elephants Using an Extension of the William Siler's Framework

Description

This is a main function of this package. It refits each model multiple times after performing a non-parametric bootstrap on the data. This is used to ultimately derive confidence intervals on parameter estimates.

Usage

boot_all_models(models_list, data, n_boot = 40L, xtol_rel = 1e-07,
  maxtime = Inf, n_CPU = 2L, save_rda = FALSE, seed = 1L)

Arguments

models_list	The list of fitted models
data	The dataset to be used
n_boot	The number of bootstrapped dataset to generate and fit
xtol_rel	Tolerance parameter for nloptr
maxtime	Maximum time (in seconds) allowed for nloptr to fit; use Inf for best fit!
n_CPU	The number of CPU to use for fitting the bootstrapped datasets
save_rda	A boolean indicating whether saving the model as a *.rda object or not
seed	An integer giving the seed for the (pseudo-)random generator

Value

The list containing all the fits.

See Also

fit_all_models

Examples

## Warning: increase n_boot and maxtime for real fits!
Models_boot_test <- boot_all_models(models_list = Models[1:2],
                                    data = SurvEles_small,
                                    n_boot = 2L, maxtime = 1)
Models_boot_test

courtiol/SileR documentation built on May 16, 2020, 8:10 p.m.