| whitening {plotKML} | R Documentation |
Derives a ‘whitenned’ color based on the Hue-Saturation-Intensity color model. This method can be used to visualize uncertainty: the original color is leached proportionally to the uncertainty (white color indicates maximum uncertainty).
whitening(z, zvar, zlim = c(min(z, na.rm=TRUE), max(z, na.rm=TRUE)), elim = c(.4,1), global.var = var(z, na.rm=TRUE), col.type = "RGB")
z |
numeric; target variable (e.g. predicted values) |
zvar |
numeric; prediction error (variance) |
zlim |
upper and lower limits for target variable |
elim |
upper and lower limits for the normalized error |
global.var |
global variance (either estimated from the data or specified) |
col.type |
characted; |
The HSI is a psychologically appealing color model for visualization of uncertainty: hue is used to visualize values and whitening (paleness or leaching percentage) is used to visualize the uncertainty, or in other words the map is incomplete in the areas of high uncertainty. Unlike standard legends for continuous variables, this legend has two axis — one for value range and one for uncertainty range (see also kml_legend.whitening).
The standard range for elim is 0.4 and 1.0 (maximum). This assumes that a satisfactory prediction is when the model explains more than 85% of the total variation (normalized error = 40%). Otherwise, if the value of the normalized error get above 80%, the model accounts for less than 50% of variability.
Whitening is of special interest for visualization of the prediction errors in geostatistics. Formulas to derive the whitening color are explained in Hengl et al. (2004).
Tomislav Hengl and Pierre Roudier
Hengl, T., Heuvelink, G.M.B., Stein, A., (2004) A generic framework for spatial prediction of soil variables based on regression-kriging. Geoderma 122 (1-2): 75-93.
Hue-Saturation-Intensity color model (http://en.wikipedia.org/wiki/HSL_and_HSV)
whitening(z=15, zvar=5, zlim=c(10,20), global.var=7) # significant color; whitening(z=15, zvar=5, zlim=c(10,20), global.var=4) # error exceeds global.var -> totally white;