An extension of the R
package geoR
that works with cost-based distances
Basically, three functions are adapted in order to admit custom distance matrices as additional arguments.
Namely variog
, to produce empirical variograms; likfit
, to fit theoretical variograms and krige.conv
, to preform kriging.
Check the documentation on these functions for examples of usage.
The cost-based distance matrix need to be computed elsewhere.
I have used GRASS GIS
, with the help of a script.
However, currently there is an R-package gdistance which could help in computing cost-based distances.
install.packages('devtools') # package devtools needed
devtools::install_github('famuvie/geoRcb')
library(geoRcb)
example('variog')
example('likfit')
example('krige.conv')
Antonio López-Quílez and Facundo Muñoz (2009). Geostatistical computing of acoustic maps in the presence of barriers. Mathematical and Computer Modelling 50(5-6):929-938. preprint | journal