average raster file value in shapefile area returns multiple outputs - interpreting results

Question

So I have my raster file

r <- raster('ras')

and a shapefile

abys <- readShapeSpatial('abys')

I calculated the mean values defined by the shapefile by the following method:

r.vals<- extract(r,abys)
r.mean <- lapply(r.vals,FUN=mean)

However, when using a couple of shapefiles when I return the output I get multiple results, e.g.:

[[1]]
[1] 9321

[[2]]
[1] 6616

[[3]]
[1] 8348

It should just return one which is what I usually get. Is this because of some characterestic of my shapefile or a problem with my methodology?

Thanks for your input

I know nothing about this method but do you perchance have three polygons in your shapefile? Ie what's length(abys)? — Ari B. Friedman
yes! so length(abys) shows there are 3 files (annoyingly simple) so my solution would be to calculate the mean from the 3 polygons — Nick Crouch

Ari B. Friedman Ari B. Friedman · Accepted Answer · 2012-10-19T22:03:45

Your problem is that there are three polygons in abys.

The best solution is not to average the results but to union the polygon first:

library(rgeos)
abys.single <- gUnaryUnion(abys)
r.vals<- extract(r,abys.single)
r.mean <- lapply(r.vals,FUN=mean)