boot_coxrfx: Bootstrap confidence intervals for regression coefficients

Description

This function computes 95% highest density bootstrap confidence intervals (non-parametric) for the regression coefficients estimated by CoxRFX.

Usage

boot_coxrfx(
  mstate_data_expanded,
  which_group,
  min_nr_samples = 100,
  output = "CIs",
  ...
)

Value

For each regression coefficient, the confidence intervals and the number of bootstrap samples on which they are based, if the `output` argument is equal to `CIs`; if `output` is equal to `CIs_and_coxrfx_fits`, also the CoxRFX objects for each bootstrap sample.

Arguments

mstate_data_expanded: Data in `long format`, possibly with `expanded` covariates (as obtained by running mstate::expand.covs).
which_group: A character vector with the same meaning as the `groups` argument of the function CoxRFX but named (with the covariate names).
min_nr_samples: The confidence interval of any coefficient is based on a number of bootstrap samples at least as high as this argument. See details.
output: Determines the sort of output. See value.
...: Further arguments to the CoxRFX function.

Author

Rui Costa

Details

In a given bootstrap sample there might not be enough information to generate estimates for all coefficients. If a covariate has little or no variation in a given bootstrap sample, no estimate of its coefficient will be computed. The present function will keep taking bootstrap samples until every coefficient has been estimated at least min_nr_samples times.