Closest point as a mean to find open space

I have been using grasshopper for 3 days now and I am trying to create a script that will find the largest and (hopefully) the best spaces to intervene in a very large city.

The approach I am using is finding the closest point from a building (internal element) to the street (external element). The subdivision and overlapping of these distances gives a graphic representation of which spaces are more susceptible to an intervention.

However I just hit a wall since I still don't grasp fully the concept and notions of trees and data in grasshopper.

the next step is to create a 2d polysurface (and someday 3D) with the points that conform the closest point lines. I need to sum 450 end point with 450 beginning points to generate 900 points. But using the math sum operator gives me the sum of the XYZ coordinate of the points.

After I solve this issue I will get to finding a way to replace the closest points that go inside a building with the second closest point. But for now, finding how to sum points and creating the surface is more than enough.

Attachments:

closest-point-for-plan-analysis.3dm, 129 KB
closest-point-for-plan-analysis.ghx, 112 KB
closest-point.jpg, 147 KB

Replies to This Discussion

Permalink Reply by Daniel Piker on January 27, 2013 at 3:12pm

Looks good!
Not sure how well the technique I showed for the medial axis will cope with very large data sets, because it relies on computing the voronoi of a large number of points (depending on how densely you sample the input curves), and I believe there are much faster algorithms out there for computing the medial axis...

Permalink Reply by maxmelone on January 27, 2013 at 3:51pm

If my last post looked good, then this must look awesome! ;)
In fact, its much better. I found it in this discussion [1] posted by Mateusz Zwierzycki. He didn't post the definition, but showed a picture, so I interpretated it and added the [flatten].

greetings,