R/nada.sum.R

In fabregithub/r4jecs: Collection of Functions for Japan Environment and Children's Study

#' NADA summary
#'
#' This is a function to calculate NADA summary statistics
#'
#' @author Shoji F. Nakayama
#'
#' @param data data only containing measurement data and censoring information
#'
#' @export
#'
#'

nada.sum <- function (data) {
  N <- ncol(data)/2
  res <- list()
  for (i in 1:N) {
    clname <- names(data)[2 * i - 1]
    d1 <- na.omit(data[, 2 * i - 1]) # remove NA
    d2 <- na.omit(data[, 2 * i]) # remove NA
    cf.res <- NADA::cenfit(d1, d2)
    cf.summary <- NADA::censummary(d1, d2)
    l.cf.res <- NADA::cenfit(log10(d1), d2)
    result <- list(summary = cf.summary,
                   mean = mean(cf.res),
                   sd = sd(cf.res),
                   quantile = quantile(cf.res),
                   gm = 10^mean(l.cf.res)[1],
                   gsd = 10^sd(l.cf.res))
    res.0 <- data.frame(N = result$summary$all[1],
                        N.censored = result$summary$all[2],
                        Percent.censored = round(result$summary$all[3], 1),
                        Percent.detected = round(100 - result$summary$all[3], 1),
                        Min = nsf(result$summary$all[4], 3, 2),
                        P25 = nsf(result$quantile[3], 3, 2),
                        Median = nsf(result$quantile[4], 3, 2),
                        P75 = nsf(result$quantile[5], 3, 2),
                        P95 = nsf(result$quantile[7], 3, 2),
                        Max = nsf(result$summary$all[5], 3, 2),
                        Mean = nsf(result$mean[1], 3, 2),
                        SD = nsf(result$sd, 3, 2),
                        GM = nsf(result$gm, 3, 2),
                        GSD = nsf(result$gsd, 3, 2)
    )
    res[[i]] <- t(res.0)
    colnames(res[[i]]) <- clname
  }
  result <- data.frame(res)
  result <- tibble::rownames_to_column(result, var = 'Summary')
  return(result)
}