[R-sig-Geo] extract function - return the ID of the polygon
Brian Oney
zenlines at gmail.com
Mon Jan 24 19:26:09 CET 2011
Hello Roger,
Using the extract function with a stack is exactly what I am doing. It
is a handly little tool. Essentially one of the few ways (that I know)
to handle large files from which a (much smaller) specific subset is
required, while remaining within the limits of R and a laptop.
I would have used the poly$FID or 1:N(polygons) and did at first but
found that the numbers of my polygons did not match up. Looking back now
it was probably operator error. Sometimes a foolproof method is better
when the computer is working better than the brain.
thanks for the handy package.
cheers,
Brian
On 1/22/2011 5:40 PM, Brian Oney wrote:
> Hello list,
> I would like to know how to also reurn the ID of the polygon when
> using the 'extract' function to extract raster cell values.
> Honestly, I am just extracting and using the cell numbers to extract
> values for these polygons in other raster layers for the same extent,
> because it is so much faster.
> Right now the one idea I have for the raster package would be to
> rasterize the polygons and then extract using the same polygons and
> then take the cell numbers and their values so that I would have a
> cell with a value for the polygon to which it corresponds.
>
> poly.rast <- rasterize(poly, rast, field=c(poly$FID))
> ex <- extract(poly.rast, poly, cellnumbers=T)
>
> I don't know how this would work with much larger files. My current
> operation corresponds to ~12000 cells and could be faster.
> Could be the problem that it is saturday and my brain is refusing to
> work...
>
> A function that might be handy (and fast) in the future might be of
> the cellFrom* function family (with an argument to carry over a
> polygon field ID etc.).
> Thanks for the help!
> Best,
> Brian
>
More information about the R-sig-Geo
mailing list