Dear all,
I have some soil moisture measurements
distributed in the basin and Iâm trying to find the best simulation method for
my data through the cross validation. I started with the tutorial of Meuse
dataset that implemented in gstat. In the case of
predictions, the cros
Edzer Pebesma
Verzonden: maandag 23 februari 2009 20:32
Aan: dde...@sciborg.uwaterloo.ca
CC: r-sig-geo@stat.math.ethz.ch
Onderwerp: Re: [R-sig-Geo] cross validation gstat
dde...@sciborg.uwaterloo.ca wrote:
> Hi list,
> A quick question regarding n-fold validation...
> I've seen several
...@sciborg.uwaterloo.ca
Cc: r-sig-geo@stat.math.ethz.ch
Subject: Re: [R-sig-Geo] cross validation gstat
dde...@sciborg.uwaterloo.ca wrote:
> Hi list,
> A quick question regarding n-fold validation...
> I've seen several papers suggest the LOOCV is too optimistic. Is
> n-fold
Thanks Edzer,
for some reason I had it in my head that n-fold was a variant of what
you describe; an independent randomly selected set to "check" the fit
if the model. I guess that's where I was heading with CV, sore form of
relative assessment of how "good" the fitted variogram model was/is
dde...@sciborg.uwaterloo.ca wrote:
> Hi list,
> A quick question regarding n-fold validation...
> I've seen several papers suggest the LOOCV is too optimistic. Is
> n-fold closer to a "true" validation?
I don't think "true" validation exists; could you explain what it is? If
you mean having a compl
Hi list,
A quick question regarding n-fold validation...
I've seen several papers suggest the LOOCV is too optimistic. Is
n-fold closer to a "true" validation?
I am assuming that it uses the variogram that is constructed using ALL
data, so my assumption is that the variogram is not re-fit for
Excellent!
It was exactly this what I meant.
I also agree that it is not needed a function to calculate such simple
stuff, but certainly would be desirable (and effective) to have it in
the help file from krige.cv (to clarify the less statistically minded
like myself :-))
I totally agree with
Marta Rufino wrote:
> Great!
Hi Marta,
I now have in the example section of gstat.cv help:
# multivariable; thanks to M. Rufino:
meuse.g <- gstat(id = "zn", formula = log(zinc) ~ 1, data = meuse)
meuse.g <- gstat(meuse.g, "cu", log(copper) ~ 1, meuse)
meuse.g <- gstat(meuse.g, model = vgm(1, "Sph
Great!
This works wonderfully...maybe would be nice if you add it to the
example in the help page :-)
Further comments in /CV/... from the gstat.cv output, which
cross-validation measures should be considered when establishing the
performance of kriging, in relation to other methods, for examp
That was not the problem, the problem was that you used meuse.g instead
of meuse.fit to pass on to gstat.cv. For meuse.g, you have perfect
correlation between Cu and Zn, so that collocated observations (meaning
a Zn and a Cu observation at each obs location) act as a duplicate in
univarite krig
Hi,
You should check if you have duplicate observations, duplicate
observations lead to a singular matrix. Use the function zerodist() to
check where the observations are and remove.duplicates() to remove them.
cheers,
Paul
Marta Rufino schreef:
> Hello,
>
> yes, I know it is suppose to do it,
Hello,
yes, I know it is suppose to do it, but I could not find how, because it
gives me an error... for example:
require(gstat); require(lattice)
data(meuse)
coordinates(meuse) = ~x + y
data(meuse.grid)
gridded(meuse.grid) = ~x + y
meuse.g <- gstat(id = "zn", formula = log(zinc) ~ 1, data = me
Yes, Martha; function gstat.cv works on multivariable objects; do read
it's documentation.
Best regards,
--
Edzer
Marta Rufino wrote:
> Dear list members,
>
> Is it possible to do cross-validation on multivariate kriging?
> i.e. apply krige.cv to multivariate kriging in gstat?
>
> thank you very
Dear list members,
Is it possible to do cross-validation on multivariate kriging?
i.e. apply krige.cv to multivariate kriging in gstat?
thank you very much,
Best wishes,
Marta
___
R-sig-Geo mailing list
R-sig-Geo@stat.math.ethz.ch
https://stat.ethz.ch/
14 matches
Mail list logo
