R/funredun.R

In PlantEcology/funredun: Calculate Community Functional Redundancy

#' Calculate Community Functional Redundancy
#'
#' Calculates functional redundance of a community based on inputs of species diversity (1-Simpson's D) and functional traits.
#'
#'@param spDat Data frame with rows as sites, columns as species, and elements as counts
#'@param funDat Data frame with rows as species (same as spDat column names), columns as functional traits, elements as counts, measures, binary, etc.
#'@param method Available options include "bray", "gower", and "altGower". See [vegan::vegdist] for details. Default is Gower distance.
#'@param redund Redundancy calculation as difference from Simpson's D (R = D - Q) or uniqueness (U = Q/D). Default is difference (TRUE).
#'@param funDiv Functional Diversity as Rao's Q (Botta-Dukát 2005). Default is FALSE.
#'
#'@return A data frame is produced with rows as sites (from spDat) and a column of functional redundancy (includes a column of functional diversity if option selected).
#'
#'@references Botta-Dukát, Z. 2005. Rao's quadratic entropy as a measure of functional diversity based on multiple traits. Journal of Vegetation Science 16:533-540.
#' @export

funredun<-function(spDat,funDat,method='gower',redund=TRUE,funDiv=FALSE){
  #sets progress bar
  pb<-utils::txtProgressBar(min=0,max=nrow(spDat),style=3,width=50,char="=")

  #calculates distance between species based on functional traits
  spDist<-vegan::vegdist(funDat,method=method,upper=T,na.rm=T)

  #creates a vector of Simpson's diversity for the community
  D<-as.vector(vegan::diversity(spDat,index='simpson'))

  #creates empty data frame P with same columns as input community data
  P<-spDat[0,]

  #loop to calculate proportion of each species at each site
  for(i in 1:nrow(spDat)){
    for(j in 1:ncol(spDat)){
      P[i,j]<-spDat[i,j]/sum(spDat[i,])
    }
  }

  #empty vectors of Q and FR
  Q<-as.vector(rep(0,nrow(P)))
  FR<-as.vector(rep(0,nrow(P)))

  #loop to calculate Q as product of dissimilarity between species i and j, proportion of species i, proportion of species j
  for(i in 1:nrow(P)){
    for(j in 1:ncol(P)){
      for(k in 2:ncol(P)){
        Q[i]<-Q[i]+(usedist::dist_get(spDist,names(P)[j],names(P)[k])*P[i,j]*P[i,k])
      }
    }

    #calculates function redundance as D-Q or functional uniqueness Q/D
    ifelse(redund==TRUE,FR[i]<-(D[i]-Q[i]),ifelse(D[i]==0,FR[i]<-NA,FR[i]<-Q[i]/D[i]))
    utils::setTxtProgressBar(pb,i)
  }

  #builds output
  if (funDiv==FALSE) {
    FRoutput<-data.frame(FR)
    names(FRoutput)<-'Func_Redun'
    rownames(FRoutput)<-rownames(spDat)
    return(FRoutput)
  } else {
    FRoutput<-data.frame(FR,Q)
    names(FRoutput)<-c('Func_Redun','Func_Div')
    rownames(FRoutput)<-rownames(spDat)
    return(FRoutput)
  }
}