[R-sig-Geo] Directional variograms with 3D data

Waichler, Scott R Scott.Waichler at pnl.gov
Wed Aug 11 06:47:46 CEST 2010


Hello,

I have a dataset with coordinates in x, y, z.  I would like to generate directional variograms, including ones for distinct horizontal and vertical directions, and fit models to them.  Apparently geoR does not accept 3D coordinates, so I am left with (?) gstat.  I have trouble understanding and applying the 5-parameter version of anis in gstat's vgm(), so I thought I would try generating one direction at a time, as below with a data frame called "han":

th.vg1 <- variogram(th ~ 1, han, alpha=0, tol.hor = 90, beta=0, tol.vert=30) # within 30 deg of horizontal  
th.vg2 <- variogram(th ~ 1, han, alpha=0, tol.hor = 90, beta=90, tol.vert=30) # within 30 deg of vertical

th.vg1 and th.vg2 are identical; both are omnidirectional.  What do I need to do to to get distinct horizontal and vertical directions?

Scott Waichler
Pacific Northwest National Laboratory
P.O. Box 999, Richland, WA  99352
scott.waichler at pnl.gov
509-372-4423, 509-341-4051 (cell)



More information about the R-sig-Geo mailing list