Hi Vic,

I haven't seen anything like this yet, but I love the idea of  
developing specific geo-processing packages built on Shapely. I wonder  
if we could reuse the WSGI model to make interoperable pipelines?

I'm not qualified to comment on your strategy. IIRC, Martin Davis (of  
Refractions) does this kind of stuff with JTS, and might have open  
source code with useful algorithms.

Cheers,
Sean

On Aug 16, 2009, at 4:29 AM, Vic Kelson wrote:

> Hi All,
>
> I think shapely is astonishingly cool, and I'm using it for  
> preprocessing groundwater model input. I had somehow overlooked the  
> shapely.ops.polygonize function, and I seriously need to be able to  
> manufacture polygons from collections of shapes. What I want to do  
> is have the user digitize line-strings that, when interescted, map  
> out polygons. However, I don't want them to have to exactly digitize  
> the nodes where they line-strings intersect. When I tried  
> polygonize(), it only works if my line-string end points connect. I  
> need something that's more like a "clean-and-build" operation. Also,  
> in addition to the polygons (they map out domains of varying aquifer  
> properties), I need the modified line-strings that bound the  
> polygons, as I impose analytic functions along the line segments,  
> and a way to identify the "left" and "right" polygons for each  
> modified line-string. I'll throw away the dangling ones...
>
> Anyhow. has anyone implemented this in a more generic way? If it  
> exists, that would be great. If not, should I just use  
> LineString.intersect() to find intersections, split strings, and  
> then send the resulting strings to polygonize? Then I can scan the  
> modified line-strings, looking for "left" and "right" by moving a  
> bit to the side of a segment and asking which polygon the point is  
> in. If a line-string has the same polygon on both sides, they're  
> dangles and I throw them away.
>
> Has anyone done this already? If not, will my strategy work? If I  
> need to implement it, I will, and I'll gladly release it under LGPL.
>
> THANKS!
> --vic
>
> -- 
> Vic Kelson
> [email protected]
> _______________________________________________
> Community mailing list
> [email protected]
> http://lists.gispython.org/mailman/listinfo/community

--
Sean Gillies
Programmer
Institute for the Study of the Ancient World
New York University



--
Sean

_______________________________________________
Community mailing list
[email protected]
http://lists.gispython.org/mailman/listinfo/community

Reply via email to