Compute bootstrap percentile confidence intervals for competing risks estimates

Computes the 2.5th and 97.5th percentiles of the bootstrap distribution for each column of a matrix of bootstrap estimates. Intended for competing risks quantities (e.g. cumulative incidence at a set of time points, or Fine-Gray regression terms), where each bootstrap replicate produces more than one estimate, unlike the single scalar log hazard ratio handled by get_boot_pci_se().

Usage

get_boot_pci_cr(boot_mat)

Arguments

boot_mat

A numeric matrix of bootstrap estimates, with one row per bootstrap replicate and one column per estimand (e.g. time point or regression term), as returned by row-binding the results of multiple calls to cuminc_waverage_cr(), cuminc_ipcw_cr(), or fg_weighted_cr().

Value

A 2-row numeric matrix with rows "lower" and "upper" giving the 2.5th and 97.5th percentiles of the bootstrap distribution for each column of boot_mat. Columns with any NA bootstrap estimate return NA for both bounds.

Examples

set.seed(1)
boot_mat <- matrix(rnorm(500 * 3), nrow = 500, ncol = 3)
get_boot_pci_cr(boot_mat)
#>            [,1]      [,2]      [,3]
#> lower -1.995029 -2.242602 -1.840346
#> upper  2.030194  2.003819  2.046006