[GRASS-user] announcing v.strahler public alpha

[Florian Kindl]

Hello lists!

I am writing a module that calculates the Strahler Order for all  
lines of a given dendritic network.

The program works on input data that meets some requirements:
The input vector map must be free of cycles. More than one tree in  
the input data is allowed. No given flow direction is needed. To find  
the outlet of each tree, a DEM must be given.

Strahler Order is not attached as a new layer yet, it is merely  
written to an ASCII file.

Everyone interested in this functionality can get the sources along  
with sparse documentation from
http://geo4.uibk.ac.at/tilde/kindl/grass/v.strahler.tgz

Copy the file to the "vector" directory of your grass source tree and  
unpack it with
`tar xzvf v.strahler.tgz`
`cd v.strahler` and type `make` to compile.
It should compile cleanly against current grass 6.1-cvs


If you like, you can help me developing the program:

Users:
Apply the module on a dataset of yours. Does it hang? Does it produce  
incorrect output? How can the behaviour of the program be improved?  
How should it format its output, what other options would you like to  
see? Can input data be treated (v.clean ..?) to make v.strahler work  
better? If it works, is it acceptable, performance-wise?

You can use g.gisenv set="DEBUG=4" to see more (maybe too much)  
information about the process.

Developers:
Are there any fundamental flaws in the design of the algorithm? Any  
severe quirks in the code?
Is there any solution on the treatment of cyclic graphs? How would  
you realize the output?
How can the code be optimized? - This is my first try in C  
programming, I'm sure there's a lot that can be improved.


I am looking forward to your opinions and experience, may this alpha  
one day become a beta!
However, I will be out-of-office for some weeks from Tuesday 8th - so  
this public alpha phase will last until mid-september at least.

\Flo.




Florian Kindl
Dept. of Geography
University of Innsbruck, Austria