[GRASS5] [bug #2637] (grass) (better) implementation of Spatial Models

[Martin Wegmann]

Hello,

sorry, it was rather as "wish" than a "bug"  report -- must have mixed 
up the menus.

Concerning GRASPER, I did not intend to say that this program is not 
sophisticated and I know that a lot of work has been put into this 
program, sorry if you got me wrong.

I was rather interested to simplify the data exchange e.g. 
import of raster data into GRASPER (equivalent to the prediction map 
export from GRASPER to GRASS).
 
Simplifying data exchange by implementing various spatial prediction 
models in GRASS should not result in a point-and-click analysis (via 
GUI) or a one-command-does-all process at all, but would drop the 
tedious import-export routine.

In the end the process of spatial modeling (commands, preprocessing, data 
cleaning etc.) would be the same but the exchange of data would be more easy. 

and concerning scripting: I am on the way to learn it ,-)

cheers Martin


On Monday 27 September 2004 15:26, Roger Bivand wrote:
> Martin:
>
> This is not a bug. GRASPER is quite sophisticated and involved a lot of
> hard work. Any data analysis involves a lot of data collection, data
> cleaning, and then analysis (with lots of possible alternatives to choose
> between). Nobody, not even the big proprietary GIS provide
> point-and-click analysis - and I would not trust their results if they
> did. If you choose to contribute dressed-up TclTk interfaces to scripts
> for GRASPER etc., feel free, but I think you'll find that you are into at
> least a 10K lines project just to start with. The bottom line is on Frank
> Harrell's "Can one be a good data analyst without being a half-good
> programmer? The short answer to that is,`No.' The long answer to that is,
> `No.'"
>
> (http://www.georeference.org/Forums/random-quotes/view-random-quotes.asp)
>
> which means scripting anyway.
>
> Roger Bivand
>
> On Mon, 27 Sep 2004, Request Tracker wrote:
> > this bug's URL: http://intevation.de/rt/webrt?serial_num=2637
> > -------------------------------------------------------------------------
> >
> > Subject: (better) implementation of Spatial Models
> >
> > Platform: GNU/Linux/i386
> > grass obtained from: Trento Italy site
> > grass binary for platform: Compiled from Sources
> >
> > It would be great if spatial prediction models like Environmental
> > Envelope, ANN, Generalized Additive Models (GAM), Tree etc. would be
> > implemented in GRASS and could be used straight forward using a bunch of
> > commands.
> >
> > At the moment a few can be used e.g. GRASP (GAM) via R but this approach
> > is quite cumbersome. (http://www.cscf.ch/grasp/)
> >
> > More infos concerning the different models can be found in these papers
> > (pdf download at http://ecospat.unil.ch/index.pl/publications):
> >
> > Guisan, A. & Zimmermann, N.E. 2000. Predictive habitat distribution
> > models in ecology. Ecol. Modelling 135(2-3): 147-186.
> >
> > Guisan, A., Edwards, T.C. and Hastie, T. 2002. Generalized linear and
> > generalized additive models in studies of species distributions: setting
> > the scene. Ecol. Modelling 157(2-3): 89-100
> > )
> >
> > thanks Martin
> >
> > -------------------------------------------- Managed by Request Tracker
> >
> > _______________________________________________
> > grass5 mailing list
> > grass5 at grass.itc.it
> > http://grass.itc.it/mailman/listinfo/grass5