[GRASSLIST:8477] Re: r.le and multiple sessions

[Paolo Cavallini]

Hi Roan.
I would not suggest you to spend much time on r.le. We did it, but the code is 
old, unmaintained and difficult to improve. Efforts have been made by  
Jachym Cepicky <jachym.cepicky at centrum.cz> to include fragstats into GRASS, 
and I believe this could be a better approach. We still need some work to 
have it running well, but I believe it is worth the while. Also Markus 
Neteler seems to be interested on this.
I can send you details and source code, if you're interested.
All the best.
pc

At 06:30, martedì 04 ottobre 2005 you presumably wrote:
> Hi All,
>
> I want to use the GRASS 6.0 application that I have developed on a
> computing cluster, which is now possible with the GRASS 6.0 multi-session
> capability (whereas GRASS 5< did not have this capability).
>
> The application involves a large number of simulations of a spatial
> process from which metrics are calculated using the r.le program.
> However, when I run r.le.patch (for example) under multi-sessions you get
> a simultaneous overwriting of the r.le.out files, thus "stuffing up" the
> data. Under GRASS 5 this was solved by the r.le.rename function. However,
> it does not seem that r.le.rename exists for GRASS 6.
>
> Using the "out" option on r.le.patch does not provide the output I am
> interested in - I am looking for whole map summary statistics (although I
> could construct one from the other).
>
> Please advise.
>
> Regards
> Rohan Sadler
> PhD Candidate
> The University of Western Australia
> School of Mathematics and Statistics and School of Plant Biology

-- 
Paolo Cavallini
cavallini at faunalia.it   cavallini at jabber.org www.faunalia.it
Piazza Garibaldi 5 - 56025 Pontedera (PI), Italy   Tel: (+39)348-3801953