R/assess_bathing_quality_eu.R

In KWB-R/kwb.flusshygiene: Functions used within FLUSSHYGIENE project (BMBF)

#' Bathing quality assessment (EU)
#'
#' Computes the quality assessment according to european bathing directive
#' 2006/7/EC from E.Coli values. The four possible quality levels are:
#' excellent, good, sufficient and poor
#'
#' @param e.coli A numeric vector with e.coli values
#' @param log logical. Are the values log-values?
#'
#'
#' @return Returns a single factor with all quality levels
#' @export

assess_bathing_quality_eu <- function(e.coli, log = TRUE)
{
  e.coli <- stats::na.omit(e.coli)

  if (length(e.coli) < 16) {

    warning("Less than 16 e.coli elements. EU directive does need 16 at least.")
  }

  if (15 < mean(e.coli)) {

    log <- FALSE
  }

  if (! log) {

    e.coli <- log10(e.coli)
  }

  mu <- mean(e.coli)

  # sd uses nominator n-1 which leads to higher sigmas
  # Set sigma to 0.5 if sd() returns NA (if there is only one e.coli value)
  sigma <- kwb.utils::defaultIfNA(stats::sd(e.coli), 0.5)

  q90 <- 1.282 * sigma + mu
  q95 <- 1.65  * sigma + mu

  levels <- c("excellent", "good", "sufficient", "poor")

  if (q95 < log10(500)) {

    return(ordered("excellent", levels = levels))
  }

  if (q95 < log10(1000)) {

    return(ordered("good", levels = levels))
  }

  if (q90 < log10(900)) {

    return(ordered("sufficient", levels = levels))
  }

  return(ordered("poor", levels = levels))
}