R/phyto_ts_aggregate.r

In algaeClassify: Tools to Standardize Phytoplankton Taxonomic Data and Perform Functional Group Classifications

#' Aggregate phytoplankton timeseries based on abundance. Up to 3 grouping variables can be given:
#' e.g. genus, species, stationid, depth range.
#' If no abundance var is given, will aggregate to presence/absence of grouping vars.
#'
#' @param phyto.data data.frame
#' @param AbundanceVar character string with field name containing abundance data
#'	Can be NA if data is only a species list and aggregated presence/absence is desired.
#' @param SummaryType 'abundance' for a matrix of aggregated abundance,'presence.absence'
#'	for 1 (present) and 0 (absent).
#' @param DateVar      character string: field name for date variable. character or POSIX data.
#' @param GroupingVar1 character string: field name for first grouping variable. defaults to spp.
#' @param GroupingVar2 character string: name of additional grouping var field
#' @param GroupingVar3 character string: name of additional grouping var field
#' @param remove.rare  TRUE/FALSE. If TRUE, removes all instances of GroupingVar1 that occur < 5%
#'                     of time periods.
#' @param fun          function used to aggregate abundance based on grouping variables
#' @param format       character string: format for DateVar POSIXct conversion
#'
#' @export phyto_ts_aggregate
#'
#' @return a data.frame with grouping vars, date_dd_mm_yy, and abundance or presence/absence
#'
#' @examples
#' data(lakegeneva)
#' lakegeneva<-genus_species_extract(lakegeneva,'phyto_name')
#' lg.genera=phyto_ts_aggregate(lakegeneva,SummaryType='presence.absence',
#'                              GroupingVar1='genus')
#' head(lg.genera)


phyto_ts_aggregate=function(phyto.data,DateVar='date_dd_mm_yy',
							SummaryType=c('abundance','presence.absence'),
							AbundanceVar='biovol_um3_ml',GroupingVar1='phyto_name',GroupingVar2=NA,
							GroupingVar3=NA,remove.rare=FALSE,fun=sum,format='%d-%m-%y')
{

  SummaryType=SummaryType[1]

  phyto.data$date_dd_mm_yy=as.POSIXct(phyto.data[[DateVar]],format=format)

  if(remove.rare)
  {
    rare.tab=table(phyto.data[[GroupingVar1]],phyto.data$date_dd_mm_yy)
    rare.tab=ifelse(rare.tab>0,1,0)
    num.occur=dim(rare.tab)[2]
    freq.occur=apply(rare.tab,1,sum)
    rare.spp=row.names(rare.tab)[freq.occur<(num.occur*.05)]
    phyto.data=phyto.data[phyto.data[[GroupingVar1]] %in% rare.spp==FALSE,]
  }

  groupingvars=c(GroupingVar1,GroupingVar2,GroupingVar3)
  groupingvars=stats::na.omit(groupingvars)

  #option to create a presence/absence matrix even if abundances are known
  if(SummaryType=='presence.absence' & !is.na(AbundanceVar))
  {
	phyto.data$presence=0
	phyto.data$presence[!is.na(phyto.data[[AbundanceVar]]) & phyto.data[[AbundanceVar]]>0]=1
	AbundanceVar='presence'
	fun=max
  }

  if(is.na(AbundanceVar))
  {
    phyto.data$presence=1
    AbundanceVar='presence'
    fun=max ##presence absence if no abundance var- this assumes every row is a detected species at that time
  }


  for(i in 1:length(groupingvars))
  {
    phyto.data[[groupingvars[i]]][is.na(phyto.data[[groupingvars[i]]])]=-9999
  }

  groupingvars=paste(groupingvars,collapse=' + ')

  agg.func=stats::formula(paste(AbundanceVar,"~ date_dd_mm_yy + ",groupingvars))

  ts.agg=stats::aggregate(agg.func,data=phyto.data,FUN=fun)

  return(ts.agg)



}