[SOLVED]Practical Range vs Theoretical Range

Here you can report the problems encountered when using a function relative to the variography:
- experimental variography (vario.calc, vario.grid, vmap.calc, ...)
- model fitting (model.auto)

[SOLVED]Practical Range vs Theoretical Range

Postby Costu » Tue Jan 07, 2014 3:52 pm

RGeoS version 9.1.11

Hi all,

when I write

Code: Select all
model.create(melem.name(2))


the theoretical range is equal to 1/3 times the practical range :

Code: Select all
Covariance Part
---------------
- Exponential
  Range       =      1.000
  Theo. Range =      0.334
  Sill        =      1.000
Total Sill    =      1.000


But when the model is created from model.auto, it is the converse (and it seems false) :

Code: Select all
vario=vario.grid(simtub(dbout=db.create(flag.grid=T,nx=c(100,100)),nbtuba=1000,model=model.create(melem.name(2),ndim=2,range=30)))
model.auto(vario,struct=melem.name(2))


Result :

Code: Select all
Covariance Part
---------------
- Exponential
  Sill        =      0.749
  Anisotropy  =      1.000     1.009
  Ranges      =     22.238    22.036
  Theo. Ranges=     66.619    66.014
Total Sill    =      0.749




Furtheremore, the definition of the range parameter seems to have changed for the K-Bessel function.
Could you create a function allowing to compute the range from the scale and vice-versa for all the melem types.

Thanks a lot for your great package.
Costu
 
Posts: 10
Joined: Sun Oct 14, 2012 8:37 pm

Re: Practical Range vs Theoretical Range

Postby Nicolas Desassis » Wed Jan 08, 2014 4:00 pm

Dear Costu,

this bug has been fixed, it was only a display bug and it didn't change the numerical results.

Concerning the K-Bessel function, the input parameter has been changed in the function model.create. Now the range parameter is actually the practical range as for the other elementary model.
If you want to keep the same behaviour as in the previous versions, you can write

Code: Select all
model=model.create(melem.name(8), scale =12)


if you want for instance the scale parameter to be equal to 12.

For the new functions you require, they will be added in a future version.

Thanks for your interest for RGeoS.
Nicolas Desassis
 
Posts: 198
Joined: Thu Sep 20, 2012 4:22 pm


Return to Variography

Who is online

Users browsing this forum: No registered users and 1 guest

cron