[R-sig-Geo] experimental covariogram

[Nick Perpinias]

Dear Edzer,

Thank you very much for your prompt answer.

Well, my thinking (and please forgive me if it is wrong) is that there is
no free lunch. What I mean is that covariogram needs a second order
stationarity, but semivariogram allows for a trend, i.e. the field has to
fulfill intrinsic stationarity. That makes semivariogram more general. So
what do you have to "pay" for this generality, especially in cases where
there is no trend in the field but still one uses the semivariogram. I
would intuitively expect that there is something to lose.

Moreover a follow up question is: is there a fitting function for
covariogram?

Thank you again for your help.

Best,

Nikos

2016-06-22 16:27 GMT+02:00 Edzer Pebesma <edzer.pebesma at uni-muenster.de>:

>
>
> On 22/06/16 16:15, Nick Perpinias wrote:
> > Dear R-GIS users,
> >
> > I am trying to do a performance evaluation/comparison between the
> > covariogram and semivariogram for certain spatially correlated data.
> > In particular, I would like to investigate if there is a certain trade
> off
> > using either of these two tools when for example there is no trend on the
> > spatial field.
> >
> > Although checking the main packages for spatial statistics in R (sp,
> > spatstat, geoR, geoRglm, gstat) I wasn't able to find a function
> > calculating the experimental covariogram of a given spatial dataset.
>
> I don't think that sp or spatstat bring you anything in this respect.
>
> gstat::variogram computes the covariogram when asked; the code that
> (finally) does it is here:
>
> https://github.com/edzer/gstat/blob/master/src/sem.c#L342
>
> >
> > In geoRglm there is only a very specific implementation for covariogram
> > estimation (covariog function) and in gstat packet the variogram function
> > has the option of covariogram without making clear if the covariogram is
> > computed or extracted from the semivariogram.
> >
> > I would be grateful if you could point out an existing function of a plan
> > of attack for my question.
>
> I'd be interested in learning why you think this is important.
>
> >
> > Thank you in advance for your time and help.
> >
> > Best Regards,
> >
> > Nikos
> >
> >       [[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
> >
>
> --
> Edzer Pebesma
> Institute for Geoinformatics  (ifgi),  University of Münster
> Heisenbergstraße 2, 48149 Münster, Germany; +49 251 83 33081
> Journal of Statistical Software:   http://www.jstatsoft.org/
> Computers & Geosciences:   http://elsevier.com/locate/cageo/
> Spatial Statistics Society http://www.spatialstatistics.info
>
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>

	[[alternative HTML version deleted]]