0

I hope you can help me with a problem that I've had for a while now. I need to make a bias file for MaxEnt, for which I have used this tutorial: https://scottrinnan.wordpress.com/2015/08/31/how-to-construct-a-bias-file-with-r-for-use-in-maxent-modeling/ and altered it to my own situation. However, I'm stuck now...

I need to use the kde2d function to create a 2d kernel density estimation and then turn that into a raster. However, the raster that is created has a different resolution for x and y. This is a problem, since I have to use it in MaxEnt which won't accept unequal x and y resolutions.

This is what I did:

biasraster <- raster("file.tif") #load raster with all the occurrences  
presences <-which(values(biasraster)==1)  
pres.locs<- coordinates(biasraster)[presences,]  
dens <-kde2d(pres.locs[,1],pres.locs[,2],n=c(nrow(biasraster),ncol(biasraster))) #2d kernel density function on the biasraster  
dens.ras<-raster(dens) #create raster from kde2d function

The original resolution of the biasraster is 0.00833333 for both x and y, but the resolution for dens.ras has changed to 0.0104052, 0.00833333 (x,y) (so the y resolution is the correct one).

As can probably be seen from the question, I am a total noob when it comes to coding (in r). I've been trying to figure out what to do for about a week now but I can't find any answers that seem to work, so I hope someone here can help me.

Nadja
  • 3
  • 1
  • Without the example file, one can guess that the indexing will reduce range in one of the directions, leading to different resolutions. Perhaps try to specify `h=` argument in `MASS::kde2d`. – nya Apr 11 '17 at 10:42

1 Answers1

0

I had the same problem. I only could solve it using:

dens.ras <- resample(dens.ras, climdat, method="bilinear") #make bias file in the same resolution of climdat

but additionaly, I need also cut dens.ras.

densmod<-crop(dens.ras,extent(climdat))

With these additional steps, my bias file worked well.

Stela
  • 16
  • 1