question about kangaroo

Hello everyone, does anyone know if is possible group objects such as spheres or boxes with a kangaroo (by springs) but maintaining a predetermined hierarchy of connections? I have seen an example of Daniel Piker Self_Aligning_Geometry that I am attaching here.can this be a good starting point?if yes someone of good will which understand better than me .could give me an explanation in broad terms about this definition?

Thanks in advance

Attachments:

Self_Aligning_Geometry2.gh, 58 KB

Replies to This Discussion

Permalink Reply by Daniel Piker on March 11, 2012 at 11:54am

Could you clarify what you mean by maintaining a hierarchy of connections ?

Perhaps a sketch of what you are aiming for ?

Permalink Reply by Massimiliano Trave on March 11, 2012 at 12:28pm

I m sorry for my bad English. for hierarchy, I intend to establish adjacency constraints, creating connections so as to drive the attraction between them . I thought this could be done with the springs and their force of attraction, I'm sorry if I'm saying something stupid but I saw the video of marc SYP about space planning and I was interested to know if I could get something like this in kangaroo or forget it because too complex.

Marc Syp video

http://vimeo.com/15563685

Permalink Reply by Massimiliano Trave on March 11, 2012 at 3:39pm

Maybe I'm way off?:)

Permalink Reply by Massimiliano Trave on March 12, 2012 at 4:54am

I tried setting the problem. I think I understand (but I'm not sure) how to set the forces of attraction(like in the image), but I have two problems 1) how to group balls of different radii, without overlapping 2) create anchor points by which i can further manipulate the solution as seen in the video. I am attaching the files. Thanks in advance

Massimiliano

Attachments:

BallCluster,space planning.gh, 21 KB
ball cluster space planning.3dm, 57 KB

Permalink Reply by Massimiliano Trave on March 12, 2012 at 2:00pm

Permalink Reply by Massimiliano Trave on March 18, 2012 at 6:03am

nobody is interested?

Permalink Reply by Mr. Gudjon Thor Erlendsson on August 13, 2012 at 5:56am

I have had a look at this myself and I think the solution is based on the tree structure that you create prior to a set of multiple Spring components. Each node in fact needs to deal with multiple relations that each affect the attraction.

Firstly the clustering force between nodes needs to be the sum of the radius of sphere A and sphere B (applied to all of them.) *As an initial starting point, this needs then to be adjusted as noted below

This initial start will only closely cluster all the balls together, without any hierarchy of relations. So the next thing to do is to adjust primary clustering force so that it is near the radius sum.

Secondly you can then add another "spring" force that adds any internal primary relations within the graph.

You can also add another component that adds repellent force to the graph

Lastly you add a "spring" component that is directly linked to the fixed points in Rhino.

I presume that you want to work with either SQL or Excel input, which you can easily do. A big challenge is to organize which nodes have a stronger attraction or repelling to other nodes. Ideally the user should be able to organise this either in Rhino itself, or in the SQL/Excel Document. This is where the data stream manipulation comes in.

Good luck, I would love to see the outcome.

Permalink Reply by Mr. Gudjon Thor Erlendsson on August 13, 2012 at 8:16am

I had a stab at this and enclose the initial solution. Hope this helps. As you can see, this is about organizing the list for the radius sums. If you use this same principle, it should start to work for you. Let me know how it goes.

Attachments: