tten a class, ghDefaultDict, in order to use the defaultdict. Similarly, the dictionary itself is at variablename.d
It seems to work, however, about 50% of the time (this happens more often in a complex definition, and especially after unlocking the canvas) the component to which the data is passed returns a completely empty dictionary. For example, these two python components should print the same thing, and they do here, but fairly often, the second will print "KEYS: []" instead.
Any thoughts on why this is or how to fix it?
This file is attached below. The Python_2 component contains only the code:
print("KEYS:")print(x.d.keys())…
ge curves. The source code is available as usual on GitHub, https://github.com/mcneeleurope/ShortestWalk.
Here some examples of walks on predefined and custom grids.
With equilateral grids (1, 2, 3), the shortest walk on the network is the same both counting the edge length and the number of links. With these types of grids, there are often several solutions, one of which is selected by the ShortestWalk component. If the automatic search is used (no lengths are specified), then the A* algorithm is used and this will result in a path that departs "not much" (there are more rigorous definitions) from the straight path.
With the square grid (2), the geometry is called taxicab or Manhattan, and results in the total distance being the sum between the number of vertical steps and the number of horizontal steps.
The circular grid (4, 6) shows a case in which curve distance and "link distance" (number of edges that are walked, uses Dijkstra's algorithm) results is completely different paths. This example here selects the tangential road (4) or the "city center" (6).
Finally, Voronoi diagrams (5), Delauney triangulations (7) and random mazes or labyrinths (8) can be walked, searched and solved quickly, if a solution is possible, now even if there are multiple overlapping curves.
These examples show two-dimensional grids, but it is possible to also compute (weighted) walks on three-dimensional networks.
The compiled Grasshopper assembly (.gha) and the examples can be downloaded from Food4Rhino. Join the group if you want to get updates for new releases.
- Giulio________________
giulio@mcneel.comMcNeel Europe, Barcelona…
the baked result into a mesh in Rhino
3) Select the mesh, in Rhino and type "ReduceMesh".
4) Set the number of meshes as simple as possible (i.e. visually - click preview) without destroying the main structure. You'll see that the spheres dissapear. 500 seems to be good when handling the entire structure
5) insert the resulting, reduced, mesh into Grasshopper and subsequently into the script. The image only shows a section of the structure since the operation takes a while.
Exploding the mesh takes out the bulgy node effect we're after.
But, experiment with it.…
ern
2 - use the b+w values for a circle packing
3 - keep the circle center points
4 - copy them across (and above/below if you want a doubly periodic pattern)
5 - make the voronoi (or delaunay in your case)
6 - keep only the curves whose center lies in the original bounding box
…
nd stress of a plate that is supported at two opposite sides (rotational degrees of freedom are allowed) and gravity load is applied. By now I can only verify the displacement of the plate with a deviation of less than 3 % using ANSYS Workbench. Kirchhoff's plate theory as an analytical approach gives a similar result with 10 % deviation.
The van Mises stress and Principal stress results in Karamba are approximately 200 times higher than the results in ANSYS and the analytical results. I tried to find the mistake for several days now and would appreciate any help or similar problems with validating the shell stresses.
Here are the values of the plate:
length: 1 m, width: 0.2 m, thickness: 0.01 m
Material: Steel 'S235' (standard)
resulting gravity load: 0.157 kN
displacement in Karamba: 0.000583 m
stress in Karamba: 116 kN/cm² (=1160 MPa = 500 % utilization!)
stress in ANSYS: 0.57 kN/cm² (=5.7 MPa)
The utilization of 500 % for a steel plate under its own dead weight makes we wonder what is wrong... See the grasshopper definition and the picture attached.
Best regards and thanks for any help,
Robert…
ome struggling i managed to get the effect i wanted but i have three problems:
1) i can't really scale these, hexagons were moved in the easiest way, so i have no control over the pattern
2) i feel that i made it pretty messy with all the dispatches, rotations etc - does it make the definition run slow? how could i simplify my definition?
3) most important i have no idea how to transfer it to hexagonal grid (so i can use jpg as attractor) for a bigger pattern. i tried starting from the hexagonal grid but couldn't get it and eventually got lost, but maybe thats the right way?
Pardon my english, hope i will get some help from you, have a good day :)…
s capacitará a los participantes en el uso de Rhino y Grasshopper.
$450 (profesores y estudiantes Veritas, para quienes se incluye en el precio una licencia académica de Rhino). $500 (Público en General)
Mayor información o inscripciones: arquitectura@uveritas.ac.cr
…
edefining the axis variables, logarithmic scales, display thresholds, better marking management - or at least add contrast!
Hey Fred,
thanks for the feedback! This is a basic version, and personally I used a custom component to read and parse the history files from the canvas to be able to e.g. scroll through generations and solutions or display more solutions at once (via pathes, mostly requires modification of the initial setup) ...
but you are right. I would love to bring the solution's navigation directly into the rhino viewport but I think that would be a major hack .. unless you can give me a hint how to do that. the displaying and user-preference-handling are besides a re-entrant history, some more algorithms and parallelization the next things to tackle, but display is definitely one of the easiest, so ... soon! work will begin in january i guess, since the project then starts i hope - but it will start for sure.
best
r
…