An implementation of the method described by Pontius et al. (2011), which compares a reference map at time 1, a reference map at time 2 and a simulated map at time 2 to evaluate allocation performance at multiple resolutions while taking into account persistence. The method quantifies disagreement within coarse squares (minor allocation disagreement), disagreement between coarse squares (major allocation disagreement), disagreement about the quantity of land use change and agreement.
1 2 3 4 5 6 7 8  ThreeMapComparison(x, x1, y1, ...)
## S4 method for signature 'Model,ANY,ANY'
ThreeMapComparison(x, x1, y1, factors, timestep, ...)
## S4 method for signature 'RasterLayer,RasterLayer,RasterLayer'
ThreeMapComparison(x, x1, y1,
factors, categories, labels, ...)

x 
either a RasterLayer of observed land use at time 0 or an object
inheriting from class 
x1 
a RasterLayer of observed land use at a subsequent time. Only
required if 
y1 
a RasterLayer of simulated land use corresponding to 
factors 
numeric vector of aggregation factors (equivalent to the 'fact'
argument to 
timestep 
numeric value indicating the timestep of the simulated land use
map. Only required if 
categories 
numeric vector of land use categories in observed maps. Only
required if 
labels 
character vector (optional) with labels corresponding to

... 
additional arguments to 
A ThreeMapComparison
object.
Pontius Jr, R.G., Peethambaram, S., Castella, J.C. (2011). Comparison of three maps at multiple resol utions: a case study of land change simulation in Cho Don District, Vietnam. Annals of the Association of American Geographers 101(1): 4562.
AgreementBudget
, FigureOfMerit
,
raster::aggregate
1  ## see lulccpackage examples

Questions? Problems? Suggestions? Tweet to @rdrrHQ or email at ian@mutexlabs.com.
All documentation is copyright its authors; we didn't write any of that.