[R-sig-Geo] vgmST parameters

Dr. Benedikt Gräler b.graeler at 52north.org
Tue Sep 26 10:27:37 CEST 2017

Admittedly, the choice of the initial parameter is not an easy task. 
Different models combine the components in a different way. In the 
separable model for instance, the spatial and temporal variogram are set 
to a joint sill of 1 each where the sill of the entire variogram is a 
separate parameter. Contrary, the metric covariance model only has a 
single variogram model as input that represents the entire joint sill 
where others are a product or the sum of two or three models (see [1] 
for details). I wrote a shiny app that is available as source from [2] 
and a link to a running instance is listed at [3]. Running the app 
locally for one owns empirical variogram might help to identify the 
right starting values.



[2] https://github.com/BenGraeler/copulatheque/tree/master/spacetime
[3] http://ben.graeler.org/shiny-apps/

On 25/09/2017 04:44, Mahsa Sameti wrote:
> Dear all
> When working with vgmST function, how we choose starting values for space,
> time anf joint vgm?
> If we should choose them from sample stvariogram, why are they different in
> different covariance models? I mean, initial values of space and time vgm in
> separable model is different from sum-metric model. how we choose them?
> Thanks alot.
> 	[[alternative HTML version deleted]]
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo

Dr. Benedikt Gräler
52°North Initiative for Geospatial Open Source Software GmbH
Martin-Luther-King-Weg 24
48155 Muenster, Germany

E-Mail: b.graeler at 52north.org
Fon: +49-(0)-251/396371-39
Fax: +49-(0)-251/396371-11

Twitter: @FiveTwoN

General Managers: Dr. Albert Remke, Dr. Andreas Wytzisk
Local Court Muenster HRB 10849
-------------- next part --------------
A non-text attachment was scrubbed...
Name: b_graeler.vcf
Type: text/x-vcard
Size: 445 bytes
Desc: not available
URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20170926/c2c0b06f/attachment.vcf>

More information about the R-sig-Geo mailing list