On 05/25/2014 05:36 PM, Moshood Agba Bakare wrote:
> Hi Roger,
> Thanks for the script. I will surely use it. I could not see an option for
> "grid". Please remember I am trying to generate interpolated values on a
> regular grid.
>
Moshood,
Please remember you cannot cross validate on a grid
Hi Roger,
Thanks for the script. I will surely use it. I could not see an option for
"grid". Please remember I am trying to generate interpolated values on a
regular grid.
Thank you
Moshood
On Sun, May 25, 2014 at 2:17 AM, Roger Bivand wrote:
> On Sun, 25 May 2014, rubenfcasal wrote:
>
> Hi M
On Sun, 25 May 2014, rubenfcasal wrote:
Hi Moshood,
I think your approach is wrong.
Cross-validation is a way to diagnose if your fitted
model/procedure is correct (you cross-validate the model, not the
predictions). You can use cross-validation to estimate the prediction
error. The pr
Hi rubenfcasal,
Thanks a lot for your response. having fitted spherical, Gaussian, and
exponential model, I had already used krige.cv to determine which of the
three functions gave highest prediction accuracy. I have four different set
of data from the same field Being irregularly space data, the
Hi Moshood,
I think your approach is wrong.
Cross-validation is a way to diagnose if your fitted
model/procedure is correct (you cross-validate the model, not the
predictions). You can use cross-validation to estimate the prediction
error. The prediction error is the expected error w
sig-geo-boun...@r-project.org] on
behalf of Moshood Agba Bakare [bak...@ualberta.ca]
Sent: Thursday, May 22, 2014 10:52 AM
To: Michele Fiori
Cc: r-sig-geo@r-project.org
Subject: Re: [R-sig-Geo] gstat cross validation for accuracy ordinary kriging
vs IDW
Hi Michele,
I tried the script as advised
un...@r-project.org]
> Per conto di Moshood Agba Bakare
> Inviato: mercoledì 21 maggio 2014 21:16
> A: r-sig-geo@r-project.org
> Oggetto: [R-sig-Geo] gstat cross validation for accuracy ordinary kriging
> vs
> IDW
>
> Hi all,
>
> I have been having a couple of challenge
Hi all,
I have been having a couple of challenge with my analysis. I have
irregularly space spatial yield monitor data over four years. Pooling this
data together is not feasible because of misalignment. That is, the
coordinates of data point varies from one year to the other.
I created a common