[R-sig-Geo] semivariogram + coordinate units

Scionforbai scionforbai at gmail.com
Fri Apr 11 04:46:18 CEST 2008


> What could be the effect of using the  coordinates in the given units
> (ie without transforming them to the same units) on the analysis?

It would be a strange kind of "anisotropy" if you want to calculate
the empirical omni-directional variogram. Indeed, the semivariogram is
a vectorial function. In the expression:
gamma(h)=E[Z(x+h)-Z(x)]²
h is a vector; of course it is supposed that both coordinates have the
same units. If not, you are working on a transformed space, not on a
cartesian one.

Such things are done to filter anisotropy of data (i.e.
stratification, "zonal anisotropy"); but they are meaningful only if
you compute the directional variograms (one in the x-direction, one in
the y-dir, each with its own unit).

Do the transformation.

ScionForbai


More information about the R-sig-Geo mailing list