[R-sig-Geo] raster::zonal with more than 1 zonal layer

[Alexander.Herr at csiro.au]

Is this what you are after?


require(raster)
require(spatial)
require(sp)
dat=raster(matrix(runif(64),nrow=8))
z1=raster(matrix( sample(1:4, 64, replace=T),nrow=8))#represents elevation bands 
z2=raster(matrix(sample(1401:1408,64,replace=T),nrow=8))#represents
watersheds


#assign unique id for each elevation x watershed
z1[]->elv
z2[]->ws
elv*1000000+ws->nd
z1->z3
z3[]<-nd
sort(unique(nd))

# do zonal
zonal(dat,z3, 'mean')


Cheers
Herry

-----Original Message-----
From: R-sig-Geo [mailto:r-sig-geo-bounces at r-project.org] On Behalf Of Dominik Schneider
Sent: Wednesday, 2 March 2016 11:48 AM
To: Help R-Sig_Geo
Subject: [R-sig-Geo] raster::zonal with more than 1 zonal layer

I'd like to summarise a raster using elevation and watershed. I was originally using extract() with a shape file and then each elevation band
within each polygon but it's very slow.   zonal() is much faster and I can
rasterize my polygons to use it.  But how do I robustly combine multiple rasterized shapefiles?
e.g.
dat=raster(matrix(runif(64),nrow=8))
z1=raster(matrix( sample(1:4, 64, replace=T),nrow=8))#represents elevation bands z2=raster(matrix(sample(1401:1408,64,replace=T),nrow=8))#represents
watersheds
zonal(dat,z1*z2,'mean')

this works well if you are certain that each combination of the values in
z1 and z2 are unique and each combination is present. otherwise it gets messy.  are there any suggestions for this use case?
ideally one could do: zonal(dat,stack(z1,z2),'mean')  and all the bookkeeping would be taken care of. my other thought is to extract all the values into data frames and use dplyr but I was wondering if there was a raster way to do this.
Thanks
Dominik

	[[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
R-sig-Geo at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-geo