thin multiple species based on spec.col

[PhillRob]

Is thin() looping through multiple species if unique(spec.col) >1? Can regions be used to separate out species instead?

[mlammens]

Hi @PhillRob. Both your questions point to good ideas for future advances. No, thin() doesn't loop through the different species right now. Though you could do this as a loop (or a subsetting) before running thin(). As for the region idea, that's something we meant to do, but haven't gotten to yet. Sorry.

[PhillRob]

Thanks @mlammens. I managed to loop through species using ddplyr on my 300k records gbif dataset, but it's too slow if you remove the timeout.

th<- function(x) {
    print(unique(x$species))
    results<-spThin(
        x,
        x.col = "decimalLatitude",
        y.col = "decimalLongitude",
        dist=1000,
        method="lpsolve",
        great.circle.distance=TRUE,
        timeout= 1000
    ) 
    return(as.data.frame(results[[1]]))
}
a<-ddply(.data = ISO.d.y.l.l.15,.(name),.fun = function(x) th(x))

To loop through the spatial dataframe using rarify and return usable results i put the data transformation steps inside a high level function.

rarefy.sp <- function(x) {
    print(unique(x$species))
    spDF <- SpatialPointsDataFrame(
        coords=as.matrix(x[,c("decimalLongitude", "decimalLatitude")]),
        data=x,
        proj4string=CRS('+proj=longlat +ellps=WGS84 +datum=WGS84 +no_def')
    )
    spDF <- spTransform(
        spDF,
        CRS('+proj=lcc +lat_1=9 +lat_2=3 +lat_0=6
            +lon_0=-66 +x_0=1000000 +y_0=1000000 
            +ellps=intl +towgs84=-288,175,-376,0,0,0,0 +units=m +no_defs')
        )
    # rarefy data
    rarefy <- spRarefy(
        spDF, 
        grid = 10000,
        cellsize = 1000, 
        nrep = 1000
    )
    rarefy.WSG <- spTransform(
        rarefy[[1]],
        CRS('+proj=longlat +ellps=WGS84 +datum=WGS84 +no_def')
    )
    
    return(as.data.frame(rarefy.WSG))
}

[Sandhyamutnali]

What is the difference between Spatially thin and Spatially Rerefy