[Gstat-info] Initial exponential modal prior to fitting
Mach Nife
machnife at yahoo.com
Tue Feb 28 13:55:22 CET 2006
Hi Edzer,
As you suggested I used 10% of 1000m = 100m for the
range and for the sill: (Standard Deviation)^2=
0.245
Here's the formula:
0 nug()+ 0.245 exp(100)
After fitting using fit=1 the result is this:
Warning: singular model in variogram
fitvariogram(planting): 'planting.sem': 0.615988
Nug(0) + 1.79057 Exp(7949.84);
In ArcGIS Geostatistical Analyst, using the same
dataset and by letting the software calculate the
optimal settings, we get this autmatically optimized
formula (OK): 0.14372 Nugget + 0.16158
Exponential(952.21).
How come? Is there other resources (open source /
freeware) available that can be used as tools to
calculate optimal models?
Sincerely,
machnife
> --- "Edzer J. Pebesma" <e.pebesma at geo.uu.nl> wrote:
>
> > Mach Nife wrote:
> >
> > >I'm wondering, if I want to use the Exponential
> > >model, in a totally non-interactive way, what
> would
> > be
> > >a good initial formula to use prior to fitting? I
> > >noticed that with different initial formulas, the
> > >result of the fitting is different...
> > >
> > >Please escuse me for my lack of knowledge on this
> > >subject.
> > >
> > >machnife
> > >
> >
> >__________________________________________________
> > >Do You Yahoo!?
> > >Tired of spam? Yahoo! Mail has the best spam
> > protection around
> > >http://mail.yahoo.com
> > >_______________________________________________
> > >Gstat-info mailing list
> > >Gstat-info at geo.uu.nl
> >
>
>http://mailman.geo.uu.nl/mailman/listinfo/gstat-info
> > >
> > >
> > I would choose a range parameter of 10% of the
> > spatial extent of the data area or volume, a zero
> > nugget and a sill equal to the sample variance.
> > --
> > Edzer
> >
> >
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam
> protection around
> http://mail.yahoo.com
>
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
More information about the Gstat-info
mailing list