Emergent Goals during Kangaroo solution

Hi Daniel, all

I have a question about emergent goals (goals created according to the particle state), whether it is possible or not: Given a large group of particles, I don´t want all the forces (goals) among them to be created at the initial state for two main reasons: due to optimization (execution time) and because that goals may depend on the state of every iteration not (only) the first step.

Furthermore, I am trying to develop any kind of restrictions (constraints) based on planes in order to get some of that particles not fixed (not an XYZanchor that is attached to the initial position) but let them move through certain relative axis when a condition is reached (that constraints should be created only at a certain iteration, not from the initial state)

Is that possible?

Thank you in advance

Attachments:

emergentgoals.gif, 2.7 MB
emergentgoalsplanes.gif, 2.2 MB

Replies to This Discussion

Permalink Reply by paul hackett on September 4, 2016 at 10:56am

Anyone can help?

Thanks in advance

Permalink Reply by Bathsheba Grossman on September 6, 2016 at 2:38am

It's my understanding that this is possible, but you cannot do it within a single Kangaroo solver. I've seen two ways to go about it.

You can wrap a Kangaroo in an Anemone (or Hoopsnake) loop, setting it up and iterating a limited number of times within the loop. An example of this might be here: http://www.grasshopper3d.com/group/kangaroo/forum/topics/random-arb... I haven't tested it. I was put off because it uses the old Kangaroo and I'm trying to learn this using mainly the new one, and also the technique looks basically a bit clunky.

Or you can write a custom component that wraps the solver, like in pointcreation.gh or growing_line.gh in http://www.grasshopper3d.com/group/kangaroo/forum/topics/new-script.... This will only work if you can fake up some knowledge of C# (it's not likely you'll ever need it for anything else), or else figure out how these examples would look in Python. I couldn't find a Python component and don't have enough general knowledge to translate, so I adapted the C# and saw it work.

Luck with the project. I don't actually know anything, I'm just Googling the same as you.

RSS