[R-sig-Geo] Unexpected behavior of raster "mask" function

[Robert J. Hijmans]

Alex,
Thanks for the clear example. I had not considered that case.
I have added an argument 'updateNA' to the mask function (version
2.2-43) such that you can do:

masked_image <- mask(img, msk, updatevalue=2, updateNA=TRUE)

That is, if updateNA is TRUE, NA cells outside the mask are also updated.

Robert


On Fri, Jul 18, 2014 at 11:18 AM, Alex Zvoleff
<azvoleff at conservation.org> wrote:
> I am using the mask function in the raster package (2.2-38) to mask out
> areas within an image that are outside an area of interest (AOI). There are
> NAs within this AOI that are meaningful - after masking the image I am
> using freq to tabulate these NAs along with the other values in my AOI. For
> this reason, I am using the "updatevalue" option in the mask function to
> recode all areas outside of my AOI to a value (say 99) so that I can ignore
> these areas in subsequent analysis, without having ignored areas share a
> value (NA) with areas inside my AOI.
>
> However, the mask function does not operate as I expected - areas in the
> image that should be masked but that are NA prior to masking are not set to
> "updatevalue". Is there an way to tell the mask function to recode all
> areas (regardless of their initial value) to updatevalue? The below example
> shows my problem:
>
> library(raster)
> # Setup an image with two NA values. Make one NA value inside
> # the AOI, and make the other outside the AOI
> img <- matrix(1, nrow=5, ncol=5)
> img[3, 2] <- NA # Inside AOI
> img[1, 2] <- NA # Outside AOI
> img <- raster(img)
> plot(img)
>
> # Setup a mask where only the 3x3 region in the center of img is retained
> msk <- matrix(1, nrow=5, ncol=5)
> msk[1,] <- NA
> msk[5,] <- NA
> msk[,1] <- NA
> msk[,5] <- NA
> msk <- raster(msk)
> plot(msk)
>
> # Apply mask
> masked_image <- mask(img, msk, updatevalue=2)
>
> # Notice that cell (1, 2) of masked_image is NA, instead of the updatevalue
> (2).
> # This cell was not affected by the mask because it was NA prior to calling
> mask.
> # Is there a way to tell the mask function to set all masked areas to
> updatevalue,
> # regardless of whether a cell is NA prior to masking?
> plot(masked_image)
>
> Thanks,
> Alex
>
> --
> Alex Zvoleff
> Postdoctoral Associate
> Tropical Ecology Assessment and Monitoring (TEAM) Network
> Conservation International
> 2011 Crystal Dr. Suite 500, Arlington, Virginia 22202, USA
> Tel: +1-703-341-2749, Fax: +1-703-979-0953, Skype: azvoleff
> http://www.teamnetwork.org | http://www.conservation.org
>
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo