[GRASS5] testing GRASS5.0.0pre1

[mberglund]

On Thu, 19 Apr 2001, Markus Neteler wrote:

> I just tried r.patch with a small map:
>
> g.region -p; date ; r.patch in=areaA,areaB out=patch; date
> projection: 99 (Transverse Mercator)
> zone:       0
> datum:      potsdam
> ellipsoid:  bessel
> north:      5772695
> south:      5766875
> west:       3559800
> east:       3569180
> nsres:      5
> ewres:      5
> rows:       1164
> cols:       1876
Ok, 1164x1876, that is small.

> Don Apr 19 17:43:43 CEST 2001
> r.patch: percent complete:  100%
> CREATING SUPPORT FILES FOR patch
> Don Apr 19 17:43:49 CEST 2001
>
> 2MB of cells take 6 seconds to complete (1Ghz AMD). The result is looking
> well. If you patch 500+MB, probably the tmp-space is too small (although
> you should get a G_malloc() error message then)?
If you mean "grass tmp", in the .tmp directory where my maps are, grass
has access to over 3 gigs of space.
Real memory doesn't seem to be a problem as it never creeps over about 20%
of 320M. It does consume all the processor it can get.

Having not run a GIG machine, I am not sure of how much faster it really
is, but being generous, lets assume 8x as fast (I am on a PII 300).
Assuming a patch of the entire set of files(which I am not doing, more
like 1/10th of the total by region) you would be patching 1gig of material
so your machine should process that in about 50mins, and mine in about 7
hours.

I let the thing run all night, I said 14 hours but more like 18/20 and got
no resulting files.

I think these numbers are probably an overshoot, but that still says
something is wrong.

> Probably you have a region problem. Try to run
If you remember about 3 weeks ago, I was having problems with r.proj and
STP becuase the nad27 file was the only one that got read. The quick fix
as per Morten Hulden was to rename grass5/etc/nad83 to grass5/etc/nad27.
To get my material to reproject I had to perform the same thing, even
under the most current cvs check out.

Could this be affecting the situation?

> g.region rast=myrastermap -p
Well, I'll try this (after I fix what I just broke).

Thanks,
Matt


---------------------------------------- 
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo at geog.uni-hannover.de with
subject 'unsubscribe grass5'