I would like some help with programming moving window comparison of spatial point data vs raster data. To account for georeferencing errors in the remote sensing data r I'd like to calculate error statistics such as MAE in 250 m increments given a 500m product. So in the example below, a 1km x 1km grid, I would like to the calculate MAE for the 9 different possible pixel locations given a 250m shift (4 corners + the 5 locations corresponding to the 'center plus sign')
require(raster)require(sp) x=runif(100,0,1000) y=runif(100,0,1000) z=runif(100) df=data.frame(x,y,z) coordinates(df)<-~x+y projection(df)<-'+proj=utm +zone=13 +datum=WGS84' r=raster(matrix(runif(4),nrow=2)) extent(r)=c(0,1000,0,1000) projection(r)<-'+proj=utm +zone=13 +datum=WGS84' I was looking at focal in the raster package but that is for a window on raster data. Googling also brought up caTools and runmean in particular but I wasn't sure how to make that work since it works on vectors and isn't inherently spatial. Thanks in advance P.S. This is cross posted at http://stackoverflow.com/questions/18424103/calculate-moving-window-mae-for-point-and-raster-databut haven't gotten any responses there. [[alternative HTML version deleted]] _______________________________________________ R-sig-Geo mailing list R-sig-Geo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-geo