R/df2vegtable.R

In kamapu/vegtables: Handling Vegetation Data Sets

#' @name df2vegtable
#'
#' @title Convert a data frame into a vegtable object.
#'
#' @description
#' Conversion of a data frame containing a cross table of abundance or cover of
#' species in single plots.
#'
#' This function coerces a data frame containing a vegetation cross table into
#' a [vegtable-class] object. The input data frame `x` may include information
#' on the layers or not.
#'
#' @param x A data frame formatted for a taxlist object.
#' @param species Numeric or integer indicating the position of the column with
#'     species names.
#' @param layer Numeric or integer indicating the position of the column with
#'     layers.
#' @param ... Further arguments passed from or to other methods.
#'
#' @return A [vegtable-class] object.
#'
#' @author Miguel Alvarez \email{kamapu78@@gmail.com}
#'
#' @examples
#' ## Creating data set 'dune_veg'
#' library(vegan)
#'
#' ## Load data from vegan
#' data(dune)
#' data(dune.env)
#'
#' ## Conversion to vegtable
#' dune_veg <- data.frame(
#'   species = colnames(dune), t(dune),
#'   stringsAsFactors = FALSE, check.names = FALSE
#' )
#' dune_veg <- df2vegtable(dune_veg, species = 1)
#'
#' summary(dune_veg)
#'
#' ## Adding environmental variables
#' dune.env$ReleveID <- as.integer(rownames(dune.env))
#' header(dune_veg) <- dune.env
#'
#' summary(dune_veg)
#' @rdname df2vegtable
#'
#' @exportMethod df2vegtable
#'
setGeneric(
  "df2vegtable",
  function(x, species, layer, ...) {
    standardGeneric("df2vegtable")
  }
)

#' @rdname df2vegtable
#'
#' @aliases df2vegtable,data.frame,numeric,numeric-method
setMethod(
  "df2vegtable", signature(
    x = "data.frame", species = "numeric",
    layer = "numeric"
  ),
  function(x, species, layer, ...) {
    # First creates species list
    taxlist <- new("taxlist")
    taxlist <- add_concept(taxlist, TaxonName = unique(paste(x[
      ,
      species
    ])))
    # Some tests previous to run the function
    Cover <- x[, -c(species, layer)]
    for (i in 1:ncol(Cover)) Cover[, i] <- paste(Cover[, i])
    Layer <- as.factor(x[, layer])
    ReleveID <- list()
    for (i in colnames(Cover)) ReleveID[[i]] <- rep(i, nrow(Cover))
    ReleveID <- do.call(c, ReleveID)
    TaxonUsageID <- taxlist@taxonNames$TaxonUsageID[match(
      paste(x[
        ,
        species
      ]),
      taxlist@taxonNames$TaxonName
    )]
    TaxonUsageID <- rep(TaxonUsageID, ncol(Cover))
    Layer <- rep(Layer, ncol(Cover))
    Cover <- do.call(c, Cover)
    x <- new("vegtable",
      samples = data.frame(
        ReleveID = as.integer(ReleveID),
        Layer = Layer,
        TaxonUsageID = TaxonUsageID,
        Cover = Cover,
        stringsAsFactors = FALSE
      ),
      header = data.frame(
        ReleveID = unique(as.integer(ReleveID))
      ),
      species = taxlist
    )
    x@samples <- x@samples[!is.na(Cover), ]
    x@samples <- x@samples[Cover != "NA", ]
    return(x)
  }
)

#' @rdname df2vegtable
#'
#' @aliases df2vegtable,data.frame,numeric,missing-method
setMethod(
  "df2vegtable", signature(
    x = "data.frame", species = "numeric",
    layer = "missing"
  ),
  function(x, species, ...) {
    x[, ncol(x) + 1] <- NA
    df2vegtable(x, species, layer = ncol(x), ...)
  }
)