Estimate IPCW Kaplan-Meier survival probabilities by covariate

Fits a weighted Kaplan-Meier estimator stratified by the specified covariate and returns survival probabilities evaluated at a pre-specified set of times.

Usage

get_ipcw_km_prob_x_se(
  data,
  covariate = "x",
  weight_var = "wgt",
  event_var = "delta",
  pre_times = seq(0, 50, 1)
)

Arguments

data

A data frame in long (counting-process) format with an IPCW weight column, as returned by get_ipcw_wgt_se(). Must contain time columns named tstart and tstop for the counting process time intervals.

covariate

Character string. Name of the stratification covariate column. Default is "x".

weight_var

Character string. Name of the IPCW weight column. Default is "wgt".

event_var

Character string. Name of the event indicator column (1 = event, 0 = censored). Default is "delta".

pre_times

Numeric vector of times at which to evaluate survival probabilities. Defaults to seq(0, 50, 1). Choose the second number to cover the entire range of times observed in data

Value

A tibble with columns time, surv (survival probability), and a column named after covariate containing the stratum labels.

Examples

set.seed(20240429)
dat <- sim_data_se(n = 500)
dat_long <- get_ipcw_wgt_se(dat)
get_ipcw_km_prob_x_se(dat_long, pre_times = seq(0, 2429, 1))
#> # A tibble: 3,724 × 3
#>     time  surv x    
#>    <dbl> <dbl> <fct>
#>  1     0     1 x=0  
#>  2     1     1 x=0  
#>  3     2     1 x=0  
#>  4     3     1 x=0  
#>  5     4     1 x=0  
#>  6     5     1 x=0  
#>  7     6     1 x=0  
#>  8     7     1 x=0  
#>  9     8     1 x=0  
#> 10     9     1 x=0  
#> # ℹ 3,714 more rows