[R-sig-Geo] Displaying temporal variation of raster data

[Oscar Perpiñán Lamigueiro]

Thiago Veloso <thi_veloso at yahoo.com.br> writes:

>  Dear R colleagues,
>
>  I have a netcdf file containing ten time slices, where each slice corresponds to yearly deforestation rates over the Amazon forest. Values range from 0 (no deforest) to 1 (maximum deforest). Please take a look in a sample of the data: http://imageshack.us/photo/my-images/23/deforestamazon.png/
>
>  My objective is to show the temporal variation of this data. How can I produce categorical histograms for each one of the ten years? What I have in mind is something like the lower panel of this figure: http://www.crwr.utexas.edu/gis/gishydro05/Time/DHI_Temporal_Analyst/TemporalAnalystForArcGIS.htm, but plotting bars relative to three or four classes of deforesting rates instead of lines.
>
>  Is it possible to produce such a figure using any SIG-related R package?
Hello,

The combination of 'setZ' from raster with 'histogram' (or 'densityplot') from
rasterVis is a suitable solution:

library(raster)
library(rasterVis)

r <- raster(ncol=10, nrow=10)
s <- stack(lapply(1:12, function(x)setValues(r, runif(ncell(r)))))
idx <- seq(as.Date('2000-01-15'), length.out=12, by='month')
s <- setZ(s, idx)
layerNames(s) <- format(idx, '%b') 
s

histogram(s)
densityplot(s)

And if you use 'zoo':

library(zoo)
histogram(s, FUN=as.yearqtr)
densityplot(s, FUN=as.yearqtr)

Best,

Oscar.

-- 
Oscar Perpiñán Lamigueiro
Dpto. de Ingeniería Eléctrica
EUITI-UPM

http://procomun.wordpress.com