Grasshopper

algorithmic modeling for Rhino

Issue populating surface with random {u,v} and with "populate geometry"

Hi everybody!

Does anybody know what algorithm is behid "populate geometry"?

I am testing populating a semi-spheric surface and with that component I don´t get more density of point at the top, which does happen evaluating the surface with random u´s and v´s.

Please find attached a screenshot of what´s going on.

Many thanks.

Cheers!

Views: 1948

Replies to This Discussion

not sure at all but... As U&V subdivision is more dense in the top of the sphere, you are more likely to find a point there. Maybe the populate geometry component doesn't use u&v... anyone to complete?

The populate components use a multiple random insert approach. Basically, a number of random points is generated and the one that is furthest from any existing point is added to the collection. Then repeat until the desired number of points was added.

Picking random points on a shape (be it a mesh, curve, surface or brep) is typically quite fast, the slow bit is finding the best point. To this end I maintain a spatial octree of all points which allows me to quickly find nearest points. However even with this optimization Populate is still quite slow.

--

David Rutten

david@mcneel.com

Tirol, Austria

Thank you, David. Very explanatory, as always.

Cheers!

RSS

About

Translate

Search

Videos

  • Add Videos
  • View All

© 2024   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service