algorithmic modeling for Rhino
All-
Anyone have a good thought on categorizing / grouping wall panels based on a panel dimensions -
I have approached it from two ways - matching all panels to a series of know sizes and looking at an individual panel and finding its match. Both give results but I think they are over complicating the issue.
I think what I should be doing is categorizing? or creating types based on dimensional similarity and outputting the number of variations...? This may be similar to the Morphosis Phare Tower panelization / typing...
Any thoughts would be appreciated.
Anthony
Tags:
thanks for the link- what if you didn't know the number of clusters just that it should be minimized based on a tolerance - and it would be searching keys among multiple dimensional criteria...
I think you can plug Galapagos in the equation, or just a serie of clusters numbers and keep the best one.
About the multi-dimensional criteria, well I dont know, are you talking about (length, width) ?
I think it should be doable to apply k-means on the length, then reapply it on width and get a useful (even if not correct) classification.
I've been a fan of the sort list component. Heres a definition that sorts random placed objects (left side) areas and sorts them along lines neatly by area (right side) using the sort list than delete duplicate and moving with a cluster I made found here. http://www.grasshopper3d.com/forum/topics/move-object-based-on-refe.... Hope it helps, and you can sort anything this way, by color, volume, distance, dimensions. Just change the area list to the desired sorting list.
Welcome to
Grasshopper
Added by Parametric House 0 Comments 0 Likes
Added by Parametric House 0 Comments 0 Likes
Added by Parametric House 0 Comments 0 Likes
Added by Parametric House 0 Comments 0 Likes
Added by Parametric House 0 Comments 0 Likes
© 2024 Created by Scott Davidson. Powered by