Le jeudi 16 mai 2013 à 14:13 +0000, BLANDENIER Lucien a écrit : > Thank you Vincent and Nikos for your answers, > > Of course I mean v.dissolve and not r.dissolve... > > Maybe I'm not clear enough wiht my problem... > > To explain what I want to obtain : > I've created such a shapfile in qgis using the tool "add part". I have then > one feature with two polygone (see attached files). > Then I've imported this shapfile into GRASS and I also see one feature with > two polygones. > > Is it possible to create such files with GRASS and in my case to merge > different non-adjacent polygone into one polygone?
As Nikos said, grass vector topology does not allow this; no problem for the hole in a polygon (an area inside it without centroid), but non-contiguous areas will necessarily be considered as distinct areas. > > The reason of this is I have to do statistics (v.rast.stats) with this shp > and I want to consider all the features of a category for calculating the > stats and the feature separated. (e.g. I want to have the mean of all pixels > contained in the polygones of the same category and for each polygones > separately) > > regards You could achieve your task with the help of categories. I mean, If you assign a distinct category to each of your "multi-polygon-areas", you can then identify/handle them individually. e.g. run a loop on cat values, which: *extracts polygons with cat=1 to a temporary vector, *uses the temp map to perform your cross-analyses between vector and raster data, *iterates to the next cat value, overwriting the former temp map, *and so on till the last cat. Grass runs this kind of sequential tasks incredibly fast on hundreds of iterations... Yours, Vincent. _______________________________________________ grass-user mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/grass-user
