[R] how to interpolate multidimensional (3D) spatial data

[Georg Roth]

  Hello,

I am an archaeologist who wonders wether a
multidimensional (3D) spatial interpolation
is possible in R.
The data consists of measurements in 3D space
with coordinates X-Y-Z and a measured variable
U, that is, U is measured at points located
at X-Y-Z.
e.g. a data structure like the following:

X    	Y     	Z      	U
5.10 	5.02 	16.03  	6.674
5.41 	4.32 	14.03  	6.668
1.72 	3.59 	12.28  	7.521
.       .       .       .
.       .       .       .

I am used to interpolation in GIS with 2D-data
(x-y, u). The interpolation method should possible
be any I am familiar with like kriging, minimum
curvature (splines), inverse distance or something
like this.

Having done 3D kernel density estimation I think
of a visualization of the interpolated 3D field
of U like the density in 3D rgl-plots {function
'plot(kde(xyz,hpi(xyz)))', package="ks"}, that is,
having the interpolated values shown as a kind of
semitransparent cloud in 3D with the colour and/or
the transparency representing the values of U.

Is there any function or package that does this
kind of computation and visualization?

Please excuse my bumpy English.

Thank You for any help.

And, thanks to the volunteer moderators allowing
mails for non-subscribers

Georg

************************************************
                                                *
"Any technology sufficiently advanced          *
  is indistinguishable from magic."             *
  (Arthur C. Clarke)                            *
                                                *
************************************************

Georg Roth M.A. (prom.)
Historisches Seminar
Professur für Ur- und Frühgeschichte
Universität Leipzig
Ritterstraße 14
04109 Leipzig
Tel: 0049 - (0)341 - 97 37069
Fax: 0049 - (0)341 - 97 37046