Hi all,
for S or R users, GeoR is worth a look for auto-fitting procedures. This
R package allows fitting of variograms (with the option of trend
removal) via least squares (equal weights, n_pairs weights, or 'Cressie'
weights), or computationally using maximum liklihood or REML. If you go
for the latter you can produce profile liklihood plots for parameters
etc. The package also will do most of the conventional flavors of
kriging, but the main focus of the package is model based geostatistics,
and is well demonstrated in:
DIGGLE, P.J., RIBEIRO Jr, P.J. & CHRISTENSEN, O.F. (2003) An
introduction to model based geostatistics. /In/ Möller, J. (ed) *Spatial
statistics and computational methods*. Lecture notes in statistics, vol.
173, p. 43-86, Springer.
The package website is here:
http://www.est.ufpr.br/geoR/
and it can also be obtained from here:
http://cran.r-project.org/
Hope that helps
Luke
begin:vcard
fn:Luke Spadavecchia
n:Spadavecchia;Luke
org:University of Edinburgh;Biosphere Atmosphere Modelling Group
adr;dom:Mayfield Rd.;;Crew Labs; Kings Buildings;Edinburgh
email;internet:[EMAIL PROTECTED]
tel;home:0131 5166243
x-mozilla-html:FALSE
version:2.1
end:vcard
* By using the ai-geostats mailing list you agree to follow its rules
( see http://www.ai-geostats.org/help_ai-geostats.htm )
* To unsubscribe to ai-geostats, send the following in the subject or in the
body (plain text format) of an email message to [EMAIL PROTECTED]
Signoff ai-geostats
