[STATSGRASS] different kriging results with the same parameters

[Edzer J. Pebesma]

Carlos "Guâno" Grohmann wrote:
> Tryed. didn't work..
>
> OK_pred_md250 <- krige(band1 ~ 1, srtm2, mask_SG, model=vrg.eye, 
> maxdist=250);
>
> Error in model.frame(formula, rownames, variables, varnames, extras,
> extranames,  :
>        invalid variable type for 'band1'
>
> Why do you think that the nugget variance would change anything?
Because the Gaussian variogram model is known to often lead to 
near-singular covariance matrices, meaning numerical instabilities, huge 
kriging weights etc.
--
Edzer
>
> tks
>
>
>
> On 11/3/06, Edzer J. Pebesma <e.pebesma at geo.uu.nl> wrote:
>> My only thought is try increasing the nugget variance; the Gaussian
>> variogram model always triggers some kind of alarm. Still, output
>> shouldn't vary.
>>
>> Have you tried replacing srtm$band1~1 with band1~1 (in which case it is
>> taken from the second argument, srtm2)? Something like
>>
>> krige(band1 ~ 1, srtm2, mask_SG, model=vrg.eye, maxdist=250);
>> -- 
>> Edzer
>>
>> Carlos "Guâno" Grohmann wrote:
>> > I was kriging some SRTM DEMs, and I noticed that I get different
>> > results with the same parameters!!
>> > check out:
>> >
>> > summary(srtm) - original data
>> > Data attributes:
>> >     band1
>> > Min.   : 634.0
>> > 1st Qu.: 759.0
>> > Median : 814.0
>> > Mean   : 856.5
>> > 3rd Qu.: 923.0
>> > Max.   :1312.0
>> >
>> > vrg.eye<-(vgm(psill=950, model="Gau", range=245, nugget=6.2));
>> >
>> > OK_pred_md250 <- krige(srtm$band1 ~ 1, loc=srtm2, newdata=mask_SG,
>> > model=vrg.eye, maxdist=250);
>> >
>> >
>> > summary(OK_pred_md250);
>> > Data attributes:
>> >   var1.pred         var1.var
>> > Min.   : 632.8   Min.   :  8.054
>> > 1st Qu.: 759.2   1st Qu.:  8.576
>> > Median : 814.3   Median :  8.712
>> > Mean   : 856.4   Mean   :  9.052
>> > 3rd Qu.: 923.6   3rd Qu.:  8.865
>> > Max.   :1307.6   Max.   :395.609
>> >
>> > Data attributes:
>> >   var1.pred         var1.var
>> > Min.   : 632.4   Min.   :  8.040
>> > 1st Qu.: 759.2   1st Qu.:  8.577
>> > Median : 814.3   Median :  8.711
>> > Mean   : 856.4   Mean   :  9.048
>> > 3rd Qu.: 923.6   3rd Qu.:  8.863
>> > Max.   :1306.2   Max.   :341.567
>> >
>> > Data attributes:
>> >   var1.pred         var1.var
>> > Min.   : 632.9   Min.   :  8.039
>> > 1st Qu.: 759.2   1st Qu.:  8.576
>> > Median : 814.3   Median :  8.712
>> > Mean   : 856.4   Mean   :  9.052
>> > 3rd Qu.: 923.6   3rd Qu.:  8.865
>> > Max.   :1306.9   Max.   :345.084
>> >>
>> >
>> >
>> > So, the minimum and maximum have changed each time. Why?
>> >
>> > thanks
>> >
>>
>>
>
>