[GRASS-user] Weighted v.kernel calculation

[Jarosław Jasiewicz]

Ted Rosenbaum pisze:
> I have a dataset of manufacturing plants with employment levels at 
> each one.  I am looking to generate a kernel density map to show 
> changing geographic distribution of employment over time.  I am 
> looking to generate a map similar to that generated by v.kernel, but 
> just with weights on each point given by the employment of the plant.
> Is there a way to do this is grass (or other software...).
> Thanks,
> Ted Rosenbaum
>
> -------------------------------
> Graduate Student
> Department of Economics
> Yale University
> ------------------------------------------------------------------------
>
> _______________________________________________
> grass-user mailing list
> grass-user at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/grass-user
>   
weighted kernel density you can do with R spatstat package and next put 
it back to the GRASS:

in R:
library(maptools)
library(spgrass6)
library(spatstat)

points=readVECT6("points")
ppp_v=as("points["weight_attribute"], "ppp")
dens_ppp_v=density(ppp_v, sigma=500, weights=ppp_v$marks)
sp_dens=as(dens_ppp_v,"SpatialGridDataFrame")
writeGRASS6(sp_dens,"name_of_result_raster_map")


HTH
J.