R/dataOverlap.r

In Beothuk/bio.utilities: Package that interacts with bio providing base-level generic routines to make your life easier.

#' @export
#' @title dataOverlap
#' @description compare 2 datasets and returns values that overlap
#' @param \code{data1} dataset #1
#' @param \code{data2} dataset #2 
#' @param \code{distance.Theshold} defaults to 0.1
#' @param \code{dnames} defaults to \code{c("EID","X","Y","date","salinity","temperature","depth","source")}
#' @return \code{Data} 
#' @author unknown, \email{<unknown>@@dfo-mpo.gc.ca}
#' @family overly specific
#' @note should not the return object should probably be Data1, not Data?
#' @export
dataOverlap<-function(data1,data2,distance.Theshold=0.1,dnames=c("EID","X","Y","date","salinity","temperature","depth","source")){

	data1$EID<-1:nrow(data1)
	data2$EID<-(1:nrow(data2))+nrow(data1)

	data1$source<-1
	data2$source<-2

	days<-sort(unique(data1$date))

	Add.lst<-list()
	Reject.lst<-list()

	for(i in 1:length(days)){

		x<-subset(data1,date==days[i],dnames)
		y<-subset(data2,date==days[i],dnames)
		z<-rbind(x,y)
		z$Overlap<-NA
		attr(z,'projection')<-"LL"
		z<-convUL(z)
		W<-owin(range(z$X),range(z$Y))
		ppp<-as.ppp(subset(z,select=c('X','Y')),W)
		z$nndist<-nndist(ppp)
		z<-convUL(z)
		z$Overlap[z$nndist>=distance.Theshold]<-F
		z$Overlap[z$nndist<distance.Theshold]<-T
		Add.lst[[i]]<-subset(z,source==1)
		}

	Data1<-do.call("rbind",Add.lst)
	return(Data)
}