BOOTSTRAPPING PMETRICS RESULTS

[mat194]

Hi everyone! A very naive question. I wanted to bootstrap the data to get the mean estimate + 95%CI for the parameters in my model (I am just screwing around with the example dataset...)

This is the code where I am trying to aggregate the results (median and mean) of my parameter for each run of the bootstrapped data set. Do you have any more elegant solutions to improve the efficiency of this process? How can I optimize the code to reduce computation time? The problem I am facing is the very high number of run to reach even 100 bootstrapped results .
Also, the command report = FALSE doesn't seem to work since I still get the pop-up.
What has been your experience with bootstrapping? Is it advisable to rely on bootstrapping only when a validation dataset is unavailable?

n_boot <- 100 #number of bootstrap
data_bs <- pk_data[["data"]] #initial data
model_bs <- run8$model
bootstrap_estimates_mean <- list()
bootstrap_estimates_median <- list()
successful_bootstrap_count <- 0

while (successful_bootstrap_count < n_boot) {
  tryCatch({
    bootstrap_data <- data_bs[sample(1:nrow(data_bs), replace = TRUE),]
    bootstrap_data <- bootstrap_data %>% 
      group_by(id) %>% 
      arrange(id, time) %>% 
      mutate(time = time - min(time))
    bootstrap_data <- PM_data$new(data = bootstrap_data)
    
    run_bs <- PM_fit$new(model = model_bs, data = bootstrap_data)
    run_bs$run(engine = "NPAG", cycle = 100, report = FALSE, quiet = FALSE)
    bootstrap_results <- PM_load(successful_bootstrap_count + 1)
  
    if (!is.null(bootstrap_results[["final"]][["popMean"]]) && !is.null(bootstrap_results[["final"]][["popMedian"]])) {
      bootstrap_estimates_mean[[successful_bootstrap_count + 1]] <- bootstrap_results[["final"]][["popMean"]]
      bootstrap_estimates_median[[successful_bootstrap_count + 1]] <- bootstrap_results[["final"]][["popMedian"]]
      
      successful_bootstrap_count <- successful_bootstrap_count + 1
    } else {
      stop("Bootstrap results are NULL")
    }
  }, error = function(e) {
    message("Error encountered: ", e$message, ". Trying another bootstrap sample.")
   
  })
}