algorithmic modeling for Rhino
Hi,
I am trying to create grasshopper and hoopsnake to create a dendriform structure. Page 30 here has a basic picture of what a 2d dendriform tree would look like.
Basically, I want a tree where each branch splits into 4 branches and this happens repeatedly. I made a very simple picture of it in paint below.
I am having issues with my data right now and one of the four branches will get 12 of the new branches and the other three branches will only get one new branch on top of it. (Sorry for the confusing situation). I have a poor understanding of lists and trees and I was wondering if anyone could help me find the error in my method or a better way to solve my problem.
Thank you so much in advance!
Tags:
The rhino file is the shapes i have been testing with
This a classic fractal problem using recursion: where a method calls itself (Louis call Louis who's calling Louis etc etc).
In general and due to the acyclic nature of GH this is only addressable with code (say: C#). I could provide a "modified" demo for your needs (no components around, mind) ... if no one comes with a component based solution (using Anemone or Hoopsnake).
In the mean time get this and spot the simplified binary tree thingy (where 2 branches pop-up from each node - hence the "binary").
best, Peter
Hi Peter:
I'm looking with interest at your fractal tree v3.gh, and also looking at how L- systems can be contained boundary wise within mesh or other brep shapes.
http://www.z-way.org/script-and-gizmo/houdini/lsystem
I was attempting to load a panel with the Hilbert 3D 01: parameter, into your script. which turned out NFG :(
see also the next 3 pics on attached URL. I presume your C## code is only addressing specific Tree L systems?
Mama Mia.
Er ... hmm ... this IS a program I'm afraid ... meaning that (obviously) rules are "translated" with a specific way (spot the tooMuchVodka bool: makes all the difference, he he):
In general ... doing anything the fractal way is the easiest of things (if you speak some proper language: C#, that is) ... but is utterly boring as well (zzzzzzzzzzzzzzzzzzzzzzzzz): I rather prefer to ride (and/or crash) a proper superbike.
HI Peter,
I always enjoy reading your posts when you get into a subject. Illuminating for various approaches, which always triggers another idea for me, leading to interesting things.
So, I was wondering how I could limit the generation within a boundary curve. I suppose I could translate your logic to grasshopper, which I expect would execute slowly, but perhaps there is a faster method?
Well ... there's 2 ways to skin the cat:
(a) Create recursively instructions and then execute them (this is the case with the big tree in v3 - the one that "bends" with the wind).
(b) Recursively control the generations without the wizardry (this is what the primitive and ugly binary mini tree does).
But since a fellow user asked some "similar" thingy quite recently (for AEC purposes) I've promised a far more elaborated example on (b): a super fractal tree no less (although bending is NOT an option, he he).
I think that the next week the super thingy could be ready, so we can talk having a solid example on hand. This is interesting since he wants to meet some "targets" as well: the final branches MUST pass through some random grid points (it's a paranoid column that ... hmm ... looks like a tree).
BTW: what means "translate" ? You mean that you want to do fractals without code?
Thanks for the reply Peter. Well, translate, yes I was thinking I could do the fractals without code, this is more thinking out loud than a plan. I think I could constrain the hilbert 'growth' with a region or mesh or volume, but only if I can access the logic through grasshopper/kangaroo. But say, for example, I want to use the hilbert for a decorative pattern on a surface (boring, I know, but something I can actually make for a project I am on). So, that is the goal, constraining the hilbert to a region/surface. If that needs to be done with code, then I guess it gets done with code.
Of course when I need to model something that has alot of void space, say, the brain of a politician, then we will go three dimensional, but for the moment I am staying 2d.
Any help is appreciated
Hilbert? you mean like this very simple thingy attached? (this has nothing to do with the case posted above).
BTW: I have what you need ... in fact I had but due to a major summer crash (Karma, what else) I've lost "some" stuff, he he. The dead HD is send to some top data retrieval gurus (California) ... but the question is:should I blow 2-3++K for retrieving about 666 paranoid C# definitions ?. Better buy some Motec ECU for my Panigale.
BTW: GH is acyclic by nature. Don't try to re-invent the wheel.
BTW: This shot mysteriously survived the havoc (Hilbert cube(s)) - one step before "connecting" the sides. No shot for Hilbert within regions found.
Well ... that happened at home workstations (all 3 Kaput !) where back-up policy is a bit "loose" he he.
BTW: Speaking with California Gurus on that Oops thingy was enlightening: they suggest to AVOID keeping data on SSD since they are prone to failures due to current fluctuations (you need a top quality UPS to handle this). But they do recommend SSD for O/S (speed).
BTW: I see that you are still after spaghetti maters (Hilbert al dente a la Vodka, what else?) but be warned: IF I rewrite the H thingy I'll use SOLELY sardines (as instance definitions) instead of lines > cooler than Sahara or what?
There's a good anenome example here that does this sort of thing...
http://www.grasshopper3d.com/forum/topics/stochastic-fractal-defini...
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
© 2024 Created by Scott Davidson. Powered by