[GRASS5] [bug #3704] (grass) d.rast.arrow does not save a line in the monitor history

[Helena Mitasova]

Michael Barton wrote:
> A negative magnitude is not necessarily an error, but could have real
> meaning.

it actually is an error because the magnitude of gradient is 
sqrt(dx2+dy2) and that cannot be negative. But you can have the arrows
representing something else (e.g. direction of a flux or movement from a 
model that for whatever reason wants to use negative magnitude - e.g. by 
using only angles  up to 180 deg).
So the program should at least give a warning and say what it is going 
to do (skip it, which I believe is the correct approach, or turn it 
around, which may be confusing)

It is great to have this finally available - thanks a lot,

Helena
> 
> Michael
> __________________________________________
> Michael Barton, Professor of Anthropology
> School of Human Evolution and Social Change
> Arizona State University
> Tempe, AZ 85287-2402
> 
> phone: 480-965-6213
> fax: 480-965-7671
> www: http://www.public.asu.edu/~cmbarton
> 
> 
> 
> 
>>From: Daniel Calvelo Aros <dcalvelo at minag.gob.pe>
>>Reply-To: <daniel.calvelo at minag.gob.pe>
>>Date: Wed, 12 Oct 2005 10:02:02 -0500
>>To: Hamish <hamish_nospam at yahoo.com>, Michael Barton <michael.barton at asu.edu>
>>Cc: <grass-bugs at intevation.de>, <david.p.finlayson at gmail.com>,
>><grass5 at grass.itc.it>
>>Subject: Re: [GRASS5] [bug #3704] (grass) d.rast.arrow does not save a line in
>>the monitor history
>>
>>From: Hamish <hamish_nospam at yahoo.com>
>>Sent: Wed, 12 Oct 2005 19:05:02 +1300
>>
>>>Done & in CVS.
>>>
>>>d.rast.arrow now can take a raster map for arrow magnitude and there
>>>is a new skip option to only draw arrows every Nth cell.
>>>
>>>screenshot showing raster elevation level, with arrow direction and size
>>>controlled by aspect and slope maps (created with r.slope.aspect); skip
>>>factor set to only draw every third cell; and grid color set to "none":
>>>
>>>  http://bambi.otago.ac.nz/hamish/grass/d_rast_arrow_magnitude.png
>>>
>>>The question remains what to do if fed negative magnitudes. Right
>>>now I have it skipping over that arrow (as it does for NULL values),
>>>but it could take the absolute value or rotate the arrow 180 degrees.
>>
>>I'd say keep the current skip, but issue a warning; whatever the cartographer
>>needs to do he can using r.mapcalc (tell him in the warning and the docs).
>>
>>The other possibility is to add an option, but I've the impression it somehow
>>reduces flexibility...
>>
>>Daniel.
> 
> 
> _______________________________________________
> grass5 mailing list
> grass5 at grass.itc.it
> http://grass.itc.it/mailman/listinfo/grass5