R/rankabund.R

In PaulESantos/diversity.tool: Tools for biodiversity data manipulation

#' Rankabun: Rank abundance
#' @description
#'   Calculating rank-abundance curves data.
#'
#' @param comm Community data.
#' @param group Index for changing the output:
#'                "none" the default output.
#'                "sites" the rank abundance indexes are estimated for each row (plot, site).
#'
#' @return a tibble
#'
#' @export
#'
#' @examples
#'
#' require(vegan)
#' data(dune)
#' rankabund(dune)
#' rankabund(dune, group = "sites")
rankabund <- function(comm, group = "none"){

  SPLIT <- c("none", "sites")
  if (is.na(pmatch(group, SPLIT)) | pmatch(group,
                                           SPLIT) == -1)
    stop("invalid group variable")
  group <- match.arg(group, SPLIT)

  name <- rownames(comm)
  #all communities
  df <- comm|>
    dplyr::as_tibble()|>
    dplyr::mutate(sites = name)|>
    tidyr::gather(species, abundance, -sites)


  output <- df|>
    dplyr::group_by(species)|>
    dplyr::summarise(abun = sum(abundance),
                     .groups = "drop")|>
    dplyr::arrange(desc(abun))|>
    dplyr::mutate(
      rank = seq(1, length(species)),
      proportion = (abun / sum(abun)) * 100,
      acumfreq = cumsum(proportion),
      logabun = log10(abun),
      rel.abund = abun/sum(abun))|>
    dplyr::ungroup()


  #by sites
  if (group == "none"){
    return(output)
  }
  else(group == "sites")
  {
    return(df|>
             dplyr::group_by(sites)|>
             dplyr::arrange(desc(abundance))|>
             dplyr::mutate(
               rank = seq(1, length(sites)),
               proportion = (abundance / sum(abundance)) * 100,
               acumfreq = cumsum(proportion),
               logabun = log10(abundance),
               rel.abund = abundance / sum(abundance))|>
             dplyr::filter(proportion != 0)|>
             dplyr::arrange(sites))|>
      dplyr::ungroup()
  }
}