R/assignNetworkMembership.R

In ksauby/ACS: Adaptive Cluster Sampling

#' Assign network membership to adaptive cluster sample units.

#' @template popdata
#' @param plot.size The length and width of plots, in coordinate units. Defaults to 1.

#' @description This function assigns network membership to units in an adaptive cluster sample; if units are neighbors, they are assigned the same network ID. 

#' @return Returns the popdata data.frame with a new column for Network ID as well as a column for m, the number of units in each network.
#' @references 
#' \insertRef{saubyadaptive}{ACSampling}



#' @examples
#' data(Thompson1990Fig1Pop)
#' 
#' # plot species abundance
#' library(magrittr)
#' library(ggplot2)
#' library(dplyr)
#' dat <- Thompson1990Fig1Pop %>% 
#' 	filter(y_value > 0)
#' ggplot(dat, aes(x,y, size=y_value)) + geom_point()
#' 
#' # assign network membership of units containing the species of interest
#' P_networks <- assignNetworkMembership(Thompson1990Fig1Pop %>%
#' 	 filter(y_value > 0))
#' 
#' # plot networks
#' ggplot(
#' 	data=P_networks, 
#' 	aes(x,y, size=y_value, shape=factor(NetworkID))
#' ) + geom_point()
#' # coordinates should be given as x and y

#' @export
#' @importFrom dplyr group_by mutate summarise_each filter arrange select
#' @importFrom igraph clusters
#' @importFrom intergraph asIgraph
#' @importFrom magrittr %$% %>% %<>%
#' @importFrom network network
#' @importFrom stats dist

assignNetworkMembership <- function(popdata, plot.size=1) {
	NetworkID <- x <- NULL
	D <- as.matrix(dist(cbind(popdata$x, popdata$y), method="euclidian"))
	D = ifelse(D > plot.size, 0, D)
	D %<>% as.matrix
	G <- network(D, directed=FALSE) %>% asIgraph()
	popdata$NetworkID <- clusters(G)$membership
	popdata %<>%
		group_by(NetworkID) %>%
		mutate(m = length(x)) %>%
		as.data.frame
	return(popdata)
}