mergeTrackObs: Count the nearest observations to points corresponding to...

In pgirmess: Spatial Analysis and Data Mining for Field Ecologists

Count the nearest observations to points corresponding to track intervals

Description

Count the nearest observations to points corresponding to track intervals (e.g. observation counts along a road discretized into points).

Usage

mergeTrackObs(sppdfInt,sppdfObs,obscol=NULL)

Arguments

sppdfInt	sfc object containing points (POINT) (the track)
sppdfObs	sfc object containing points (POINT) (the observations)
obscol	The column number in which the number of observations at this point can be found in sppdfObs, if any (default=NULL, no such data)

Details

Road side counts and faeces collections are often carried out along tracks (paths, roads, transects, trails, etc.). Tracks can be discretized in regular intervals e.g. with transLines2pix or st_thintrack, each point being the center of a track interval. mergeTrackobs computes the number of observations that are the closest to each interval (compare to the other intervals) and give it in the column 'nObs'. if the argument 'obscol' is NULL, observations are considered presence/absence. If the number of observations is a true count (0 or any positive number) the argument 'obscol' can be used to identify the column of 'sppdfObs' where those counts must be found. Coordinate reference systems must be the identical.

Value

A sfc object, of POINT geometry, with the following columns:

• ID, ID number

• nObs, The number of observations in the interval

• geometry, POINT geometry

See Also

transLines2pix, st_thintrack

Examples

if(require(sf)){
# track

sl<-st_sfc(st_linestring(cbind(c(1,2,3),c(1,1.5,1))))
plot(sl, col = "blue")

    #observations
     obs <- structure(list(ID = 1:15, long = c(1.04609377280342, 1.0890625305741, 
1.08125002916125, 1.24921880953755, 1.34687507719818, 1.50312510545521, 
1.88984392539134, 2.37812526369453, 2.39375026652023, 2.36640651157525, 
2.38593776510738, 2.62031280749291, 2.69843782162142, 2.85078159917202, 
2.90546910906198), lat = c(1.04062476682306, 1.05624976964876, 
1.03671851611663, 1.13828103448369, 1.16562478942867, 1.26718730779574, 
1.43124983746561, 1.32968731909855, 1.32187481768569, 1.30624981485999, 
1.28281231062144, 1.20468729649293, 1.13828103448369, 1.08749977530016, 
1.03671851611663)), .Names = c("ID", "long", "lat"), row.names = c(NA, 
-15L), class = "data.frame")
points(obs[,2:3],col="red")

obs<-st_as_sf(obs,coords=c(2:3))
# possibly a count (number of individuals observed) on each location
obs$n<-c(3,4,0,1,1,5,6,4,3,4,4,7,2,2,1) 

# examples

# Presence/absence on each observation
track<-transLines2pix(sl,0.1)
trackObs<-mergeTrackObs(track,obs)

par(mfrow=c(1,2))
plot(sl,reset=FALSE)
plot(track,add=TRUE,col="blue")
plot(st_geometry(obs),add=TRUE,col="red",pch=1)

plot(sl,reset=FALSE)
plot(track,add=TRUE,col="blue")
plot(trackObs,cex=trackObs$nObs,pch=19, col="red",add=TRUE)


# 0 or more observations on each location
head(obs)
st_drop_geometry(obs) # the counts are in column "n" (column 2 in the data.frame)

trackObs<-mergeTrackObs(track,obs,obscol="n")

par(mfrow=c(1,2))
plot(sl,reset=TRUE)
plot(track,add=TRUE,col="blue")
plot(st_geometry(obs),add=TRUE,col="red",pch=1)

plot(sl)
plot(track,add=TRUE,col="blue")
plot(st_geometry(trackObs),cex=trackObs$nObs/3,pch=19, col="red",add=TRUE)


}
     

pgirmess documentation built on May 29, 2024, 4:26 a.m.