[GRASS-dev] Re: r.prominence fails to compile under grass 7

[Benjamin Ducke]

Hi Michael

Michael Barton wrote:
> I tried out the new code in GRASS 6.5 and 6.4. It does work much better 
> with the GUI and has no trouble with integer maps. It seems to run 
> faster, though maybe that's my imagination.

I think it is ;-)
I did not put in any optimizations. However, I think you were referring
to some suggestions (by Glynn C?) for those in one of your previous
emails. Could you give me some details?

> 
> I tried it with a very small radius (3m) to see what happens and posted 
> the result.
> 
> http://www.public.asu.edu/~cmbarton/files/temp/pks_rprominence3.jpg
> http://www.public.asu.edu/~cmbarton/files/temp/pks_rprominence3_nviz.jpg
> 
> One thing I forgot to mention is that at small sizes, r.prominence 
> creates an interesting red -> black -> white color table. But with 
> larger maps (e.g., the Spearfish DEM), it does not produce a viable 
> color table and the map appears blank. Running r.colors and picking a 
> color table fixes this, but it is probably something simple to fix in 
> the code too.

I can't reproduce that. Colour scaling is independent of cell 
resolution and number (are you referring to that when you say "small size"?) 
and worked in my tests using different output map sizes.

However, the actual result value ranges may be scale-sensitive.
Are you sure you did not have a mismatch between the search radius
and the map scale?

Do you use any of the normalization options?
Could you post the data ranges for those result maps where the colour

table seemed bad?

By "blank" do you mean all white or black?

Cheers,

Ben


> 

-- 
Benjamin Ducke
Geospatial Consultant

Oxford Archaeology Digital
Janus House
Osney Mead
OX2 0ES
Oxford, U.K.

Tel: +44 (0)1865 263 800 (switchboard)
Tel: +44 (0)1865 980 758 (direct)
Fax :+44 (0)1865 793 496
benjamin.ducke at oadigital.net
http://oadigital.net





------
Files attached to this email may be in ISO 26300 format (OASIS Open Document Format). If you have difficulty opening them, please visit http://iso26300.info for more information.