[STATSGRASS] v.stats.rast for category raster or vector cross statistics
On Mon, 22 Oct 2007, Matthew Perry wrote:
Agustin, I would suggest taking a look at starspan (http://starspan.casil.ucdavis.edu/). It specializes in vector-on-raster analyses and has a --count-by-class flag for categorical rasters. I wrote a bit more about it at http://www.perrygeo.net/wordpress/?p=30 . - matt On 10/22/07, Agustin Lobo <Agustin.Lobo at ija.csic.es> wrote:
Dear lists, Given Av (vector of polygons) and Br (raster of a categorical variable, i.e. a classification or a soil map) I want to calculate for each polygon in Av, the surface of each category from Br. v.stats.rast almost does it, but calculates lumped stats for each polygon (i.e, assumes that Br is a continuous variable).
Alternatively, I could have Br as another vector Bv, but don't find the way of calculating cross stats as r.stats -a in=Ar,Br would do for raster layers. Converting Av into raster Ar is inconvenient, as the polygons are few and small and would have to use a very samll cell size (Br is large) Any easier way through R?
StarSpan is a good suggestion for heavy lifting. Within R, you could look at the overlay methods in sp to create a vector for polygon membership of raster size, then write a tapply function to accumulate tables with zero counts on categories by the polygon index. Roger
-- Dr. Agustin Lobo Institut de Ciencies de la Terra "Jaume Almera" (CSIC) LLuis Sole Sabaris s/n 08028 Barcelona Spain Tel. 34 934095410 Fax. 34 934110012 email: Agustin.Lobo at ija.csic.es http://www.ija.csic.es/gt/obster
_______________________________________________ statsgrass mailing list statsgrass at grass.itc.it http://grass.itc.it/mailman/listinfo/statsgrass
Roger Bivand Economic Geography Section, Department of Economics, Norwegian School of Economics and Business Administration, Helleveien 30, N-5045 Bergen, Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43 e-mail: Roger.Bivand at nhh.no