[R-sig-Geo] variogram using gstat

[ONKELINX, Thierry]

Hi Edzer,

Well, I wrote similar because I don't now the details by heart.
Furthermore the we analysed the same data independently in R/gstat and
variowin/gslib. The emphasis of our comparisons was rather on what was
the most userfriendly workflow. So the question was not realy if gslib
yields exactly the same results as gstat. But rather if R/gstat is
usefull to teach geostatistics when compared to variowin/gslib. Given
that most of the students use R (or S+) in other courses, we think that
using R/gstat is easier for the students.

I don't know whether the professor will change from variowin/gslib to
R/gstat. We were probabily just a bunch of R-addicted students who
perferred their own toys ;-)

Cheers,

Thierry 


------------------------------------------------------------------------
----
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek / Research Institute for Nature
and Forest
Cel biometrie, methodologie en kwaliteitszorg / Section biometrics,
methodology and quality assurance
Gaverstraat 4
9500 Geraardsbergen
Belgium 
tel. + 32 54/436 185
Thierry.Onkelinx at inbo.be 
www.inbo.be 

To call in the statistician after the experiment is done may be no more
than asking him to perform a post-mortem examination: he may be able to
say what the experiment died of.
~ Sir Ronald Aylmer Fisher

The plural of anecdote is not data.
~ Roger Brinner

The combination of some data and an aching desire for an answer does not
ensure that a reasonable answer can be extracted from a given body of
data.
~ John Tukey

-----Oorspronkelijk bericht-----
Van: Edzer Pebesma [mailto:edzer.pebesma at uni-muenster.de] 
Verzonden: woensdag 7 mei 2008 16:00
Aan: ONKELINX, Thierry
CC: PUJAN RAJ REGMI; r-sig-geo at stat.math.ethz.ch
Onderwerp: Re: [R-sig-Geo] variogram using gstat

ONKELINX, Thierry wrote:
> Dear Pujan,
>  
> The helpfile (?variogram) is a wonderfull thing because it answers you
> question about dist. It reads: 
> dist	 the average distance of all point pairs considered for this
> estimate	
>  
> I think that gstat is based on gslib (Edzer, please correct me if I'm
> wrong). 
No, the gstat "core" code is written in C rather than in fortran. It was

written in a similar time, i.e. the early 90s.
> Last year we compared gslib with R and gstat. Both yielded
> simmilar results. We prefered R and gstat over gslib because R is a
more
> flexible and versatile tool.
>  
I'm always interested in hearing comparison results, and trying to 
figure out where differences come from. In theory, results should be 
identical rather than similar.
--
Edzer