R/getClosestPosition.R

In ferrettigs/OIB: Ocean Intelligence Bureau Toolbox

#' Get Closest Positions between Fish and Ships
#'
#' 
#' @param out output object from \code{searchVessel} function
#' @export 
getClosestPosition = function(out){

out$drfdat = unique(out$drfdat)
out$locs = unique(out$locs)

dm = geosphere::distm(out$drfdat[,c("lon","lat")], out$locs[,c("lon","lat")])/1000
# matrix of distances in km

tm = outer(out$drfdat[,"datetime"], out$locs[,"timestamp"], "-")
# matrix of differences in time in days
# now I can try to find the euclidean distance
edist = dist(cbind(as.vector(dm), as.vector(tm))) # but it sucks all the memory

#library(pdist)
dists <- pdist(out$drfdat[,c("lon","lat","datetime")], out$locs[,c("lon","lat", "timestamp")])
edist = as.matrix(dists) # matrix of euclidean distances
tag = out$drfdat[which(edist == min(edist), arr.ind = TRUE)[,1],][,c("lat","lon", "datetime")]
boat = out$locs[which(edist == min(edist), arr.ind = TRUE)[,2],][,c("lat","lon","timestamp")]
names(boat)<-c("lat","lon","datetime")

segment = rbind(tag,boat)
segment$distance = (geosphere::distm(segment[,c("lon","lat")])/1000)[1,]
segment$timelag = c(0,diff(segment[,c("datetime")]))
segment

}