nitions prior to Karamba are to allow the genes to manipulate the form of the shell and then kangaroo to relax the form to its "equilibrium" state.
The definition, as attached, runs fine over one iteration. However, when I run the Galapagos solver, rhino slowly uses up my computers memory and then ultimately crashes (around 80 Galapagos iterations). I don't think that the surface patch, or kangaroo are the issue, as I have run other iterative definitions through them without issue.
I believe Karamba may be occupying memory each iteration that is not released when a new iteration begins. This problem is exasperated by the fact that I am running 11 load cases, 9 of which are point loads defined over each vertex of the mesh. I ran a definition with only one load case, and it reached 170 generations (with a population of 50 for each generation). However, at this point it had occupied 90% of my computer's available memory.
Do you know of a way to ensure that Karamba purges its memory after an iteration, or is this a possible memory leak bug?
Thanks again, any help you can provide is much appreciated.
Sean
…
.
I think i'm about 80% there. It may not be the most elegant procedure, but it appears to be working. I am having some slight problems.
1. I'm having trouble extracting the edge curve on one side of my mesh
a. I used a mesh plane intersection to trim my site mesh, and the resultant curve extends beyond the limit of my mesh
b. I identified the intersection point, but GH is not shattering on that intersection point. instead it happens at a point further down the polyline
2. I'm can't join my curves. If i bake them, they join into a closed curve.... GH join curves is not behaving. i've had some success joining two segments, but i can seem to get all 4 side wall curves to join.
if anybody could offer some advice, i'd be very appreciative. also, i'd like some recommendations for ways to streamline this definition. I'm sure that there are more creative ways to manage this data, and i'd love to about them. thanks - CSDG
…
use an attractor curve to adjust curves that are being lofted to create a "wavey" surface. I've attached a picture to show the end result. Right now i create all the curves and then loft them together to create a surface which could then have a rib definition applied to it, and it works, my problem is that this project is for a 80' section of wall with ribs that would be a few inches apart, so that is a ton of curves i have to change everytime i want to adjust the overall shape. So is it possible to have an attractor curve that instead of adjusting the spacing of shapes in plane with the curve, would create high points or low points perpendicular to the curve? Hopefully that makes some sense, i'm having trouble finding the words to explain it, the attached pictures should help.
Any advice on how to do this would be great. I hate being that guy but this project came up last minute and it seemed like something the Grasshopper help save a ton of time in as far as adjustments go.
Thanks,
Kyle
…
h kangaroo and have found one project that illustrates my idea exactly...
https://vimeo.com/88002087
So far my best attempt has been to use a gridded surface in which the lines of the grid are springs, with a rest length at 80% of their initial length. (call this grid A) This is to simulate as if the material has been stretched 120% of its resting dimensions.
I have been trying to anchor the springs to a secondary grid (Grid B, curves that will be deposited onto the material) at the points of intersection with grid A.
I am not sure if this is the best approach, maybe soapfilm would be better? although i require the boundary (grid B) to adapt also...
Any advice or attempts to explain how Taichi Kuma has done this in his video would be greatly appreciated.
Thanks…
A: Who created it? / Copyright?
B: Anyone have a larger resolution copy of this image (or a vector so I can do a large resize).
Long and short - I'm an artist and I'd like to make a quilt out of that image. I need a decent resolution copy of it so I can do a huge resize and get a 76" x 80" crop out of it. I've tried resizing it in Photoshop, but by the time it's large enough, the quality has dropped below what is usable for my needs.
Thanks in advance! I totally understand I'm relying on the kindness of strangers here. (Also if this is inappropriate to post - please delete. Thank you)
Michael…
e some questions.
I want to loop with a foreach loop trough a list of points do i have to make a list before or is it possible to use them coming in from a noed i set the access to list?
Also i dont understand why no plane is created. How do i need to feed the points in?
And why is c# expecting open parens in line 88 and 86?
Hope its not to much at once, probably i should try a few less steps to get the problems solved one by one, just hoped it would be easier and sometimes just a parentesis is missing or some format stuff, so maybe it is not so much i really cant say.
If anybody has the time and feels he wants to help it would be nice on the other hand i understand cause of the amount of chaotic questions.
Regards!…
" (idiomatic) and easy way of doing things.So here come some basic questions:
Is there a way to create custom components by grouping an existing sub-network together? I'm looking for a way to re-use parts of a program (something similar to subroutines), and to make the network look less cluttered. I found that it is possible to group components (ctrl-g), but this still displays them as separate blocks (too much clutter), and provides no way to re-use a sub-network in such a way that if it is modified in one place, all it's instances (all the places where it is re-used) also get modified.
Is there a component that does nothing, just passes a signal through? Suppose I need to connect block A to blocks B1, B2, B3 (all three get the same input). Then I change my mind, and I decide to connect block C to these three, not A. In this case it will be necessary to change three connections, not just one. I'm looking for an easy way to do this by a single rewiring, not three. (This came up in a practical situation).
Finally, a related question: is there a component that acts as a switch, so I can choose which signal it passes through out of a possible set of choices? For example, suppose that a set of objects can be coloured based on a number of different properties (size, positions, rotation, etc.) I'm looking for a way to switch between these very easily, without the need to do much rewiring.
Thank you in advance for any replies / useful comments, even general ones on how to easily structure a large Grasshopper program/network.…