[GRASS-dev] does r.composite respect a mask?

[Michael Barton]

It was really odd.

The original RGB channels were around 300 Mb each for the entire maps, with a resolution of 0.5m. Only about 1/4 was within the MASK and computational region. It took over 4 hours to run r. composite and produced a resulting map of over 5 Gb, but only for the area in the MASK and region. 

When I reset the resolution to 1m, r.composite ran in a minute or so and produced a resulting map of 77Mb. 

I’m OK with the 1m resolution I guess but the whole thing seemed strange.

Michael
______________________________
C. Michael Barton 
Director, Center for Social Dynamics & Complexity
Professor of Anthropology, School of Human Evolution & Social Change
Head, Graduate Faculty in Complex Adaptive Systems Science
Arizona State University
Tempe, AZ  85287-2402
USA

voice: 	480-965-6262 (SHESC), 480-965-8130/727-9746 (CSDC)
fax:          480-965-7671(SHESC), 480-727-0709 (CSDC)
www: 	http://csdc.asu.edu, http://shesc.asu.edu
		http://www.public.asu.edu/~cmbarton

On Sep 9, 2014, at 8:34 AM, Glynn Clements <glynn at gclements.plus.com> wrote:

> 
> Michael Barton wrote:
> 
>> It is taking a VERY long time to run r.composite. It is a big map in
>> its entirety. But only a part of the map falls within the MASK.
>> 
>> FWIW, I have the computational region set to match the MASK too.
> 
> r.composite doesn't do anything special with regard to the region or
> mask.
> 
> A "sparse" map (with large null areas) should have some advantage over
> a dense map, as the colour lookup optimises the case where a cell's
> value matches the previous cell's value, and also checks for null
> before other values.
> 
> -- 
> Glynn Clements <glynn at gclements.plus.com>