[Gstat-info] help with Gstat

[Tomislav Hengl]

Dear Ana,
 
Most of the things you are asking for ARE available in the gstat user manual
or on the website. There is no problem to use multiple or predictors for UK
or to do indicator kriging in gstat. It might get, however, very
time-consuming to produce predictions for large images (>1M pixels) or large
number of predictors.
 
This guide might also help you - http://spatial-analyst.net/RKguide.php
 
cheers,
 
Tom Hengl
http://spatial-analyst.net 
 
 

  _____  

From: gstat-info-bounces at geo.uu.nl [mailto:gstat-info-bounces at geo.uu.nl] On
Behalf Of Ana Amorim
Sent: 08 March 2006 20:21
To: gstat-info at geo.uu.nl
Subject: [Gstat-info] help with Gstat



Dear friends

 

I'm developing a project on Gis and Geostatistical modelling of temperature
and I'm starting to discover the Gstat software. 

 

I'm trying to do Kriging with external drift of temperature, with other data
such as altitude, distance to the sea, and latitude and longitude.... I
would like to know if the Gstat allows more then one variable for external
drift, and if yes, what's the command file I should use. I've seen the
manual and the example 13 as dr. Pebesma recommends on an e-mail, but still
not enough clear. 

 

I'm trying also to do some modelling of variability and uncertain with
Indicator kriging. I would like to define several cutoff/threshold values
(ex. for temperature, 5º, 10º, 15º C, etc) and know what's the variance
probability around this values. My question is, if Gstat allows doing this
indicator kriging all at the same time, knowing it will do the order
relation violation correction for the indicators. Once again, what's the
command file to process indicator kriging? 

 

I think the manual doesn't have any examples of command files for this
kriging I'm asking for. If you have some illustrative examples it would be
great for my understanding. 

 

Thanks in advance for the attention.

 

Ana Maria Torres Amorim

Geo Systems Center

Technical University of Lisbon

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.geo.uu.nl/pipermail/gstat-info/attachments/20060309/f4c20e1b/attachment.html