R: [R] Re: Computation of space-time empirical crosscovariances

[giovanna jona lasinio]

Thanks, I'm looking at gstat and trying to understand in which way it is
appropriate to add time as the third dimension... Indeed the "simple"
attribute is not for space-time problems :-)

-----Messaggio originale-----
Da: Edzer J. Pebesma [mailto:e.pebesma at geog.uu.nl] 
Inviato: venerdì, 3. dicembre 2004 11:50
A: r-help at stat.math.ethz.ch; giovanna.jonalasinio at uniroma1.it
Oggetto: [R] Re: Computation of space-time empirical crosscovariances


Giovanna,

space-time cross covariance has never seemed to me something that can be
done "simple", but package gstat at least offers you: 1. to include time
as a third dimension, and model 3D anisotropy, 2. to calculate cross
variograms between different moments in time, and proceed with
cokriging.
--
Edzer