
rem ILWIS Script

//Created by T. Hengl (http://spatial-analyst.net)
//Input parameters: %1.mpr - simulated values, %2 - prediction error, %3 - original sampled variance
//Purpose: Visualisation of uncertainty for a given simulation image and prediction error map. 
//Last update: 27. February 2006.

//Normalize the prediction error

err=sqrt(%2)/sqrt(%3)

//Convert to amount of variation explained by the model

rsquare=1-err^2
calc rsquare.mpr

//Randomize values

RND_n=0*rsquare+rnd()

//Visualize amount of pixels proportional to the amount of variation explained

%1_VIS=iff(rsquare>RND_n, %1, ?)

calc %1_vis.mpr
breakdep %1_vis.mpr
setdescr %1_vis.mpr Uncertainty visualized using pixel mixing

del err.mpr -force
del RND_n.mpr -force
del rsquare.mpr -force

open %1_vis.mpr -noask
 