Skip to content
Prev 15179 / 29559 Next

GSTAT and anisotropy

This example

 > library(gstat)
 > data(meuse)
 > coordinates(meuse) <- 1:2
 > sum(variogram(zinc~1, meuse)$np)
[1] 6883
 > sum(variogram(zinc~1, meuse, alpha=45)$np)
[1] 6883
 > sum(variogram(zinc~1, meuse, alpha=45, tol.hor=10)$np)
[1] 1450

shows the effect of the default value of the argument tol.hor 
(documented in the "Usage"-section of ?variogram). Specifying a single 
value for alpha thus gives an isotropic variogram with the default for 
tol.hor.

Am 29.05.2012 06:39, schrieb Michael Hewson:
I assume you plotted xv.aeVin with plot.gstatVariogram. From the 
following example I would conclude you have to provide your own panel 
function or use something else than simply the plot method, as both 
values given to scales yield the same result. But it might well be that 
I missed something here. There are many screws you can screw in the 
world of lattice.

g <- gstat(NULL, "z", zinc~1, meuse)
g <- gstat(g, "l", lead~1, meuse)
plot(variogram(g, alpha=c(45,90,135,0)), scales=list(relation="same"))
plot(variogram(g, alpha=c(45,90,135,0)), scales=list(relation="free"))

HTH,
Tom