[R-sig-Geo] Mapping cells from dataframe constructed in raster

[Robert J. Hijmans]

Hi Janice,

That works but seems much more complicated than need be. I think you
should read the vignette that comes with the raster package:
http://cran.r-project.org/web/packages/raster/vignettes/Raster.pdf

Here is a simpler approach

library(raster)
s <- stack(system.file("external/rlogo.grd", package="raster"))
# create a matrix
m <- as.matrix(s)

# add a column with the cell numbers
m <- cbind(cell=1:ncell(s), m)

# it seems that in your case this is not needed because you used
# write.csv (which adds the cellsnumbers (rows in matrix) and then
read.csv to accomplish
# the same

# for this example, I take a random sample of 100 cells
set.seed(0)
x <- sample(m[,1], 100)

# create a RasterLayer template from the RasterStack
r <- raster(s)

# assign values to cells 'x'
r[x] <- 1:length(x)
plot(r)


Next time you ask a question provide a self-contained example like
Andrew's or  this one, which makes it much easier to answer.

Robert


On Thu, May 2, 2013 at 9:20 AM, Janice Lee <janice.jlsh at gmail.com> wrote:
> Dear all,
>
> I would appreciate anyone's help on this.
> I constructed a dataframe from several raster layers and compiled these
> layers using the 'stack' and 'brick' function in 'raster' package
>
> s <- stack(r1,r2,r3,r4,r5)
> b <- brick(s)
>
> I converted this into a matrix and stored this as a dataframe:
>
> mb <- as.matrix(b)
> write.csv(mb,"C:/Source/mb.csv")
>
> This new dataframe (mb) lists all the cells as rows and all the attributes
> in my raster layers (r1-r5) as columns
> The first column in mb also provides a number to each cell which I assume
> acts as a form of identification number to every grid cell
> I used mb in an R-coded model which simulates land cover change, in which I
> selected the most suitable cells for land use conversion
> My final dataframe (df) has a lower number of rows compared to mb but has
> preserved the identification number of each grid cell.
>
> My question is: Can I map out the cells in df by tracing its identification
> number back to the original dataframe mb?
> I have been trying to figure this out and will appreciate any suggestions.
>
> Many thanks and best regards,
> Janice Lee
>
>         [[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