2017-03-2807:28 AM - last edited on 2023-05-1111:11 AM by Noemi Balogh
I am just starting to play with the connection plugin and see that V20 has many more parameters to play with. I am currently doing a randomized facade that is working well. However I am wanting to randomize the colour of some fins between two different surface materials. Any thoughts as to which grasshopper component will allow me to plug different surface or composite inputs into?
As an aside the my Archicad points dont seem to be translating into grasshopper points like they should. They seem to shift in strange directions. What could be affecting this?
in this case the problem is related to the way grasshopper works with data trees. I know it because while completing the excercise for yafimski i experimented the same issue.
When a gh component has to deal with two (or more) inputs and each one of those input has different lenght (element count) or different Data Tree topology (a data structure used by grasshopper, similar two a multidimensional array) some logic has been established by David Rutten.
In this case, two data trees are passed two Wall component, a data tree of curves and a data tree of wall settings. One has to be conscious of the structure (the paths where the data is allocated in the data tree) of the input data in order to things happen as one is expecting.
Two inputs data will be used internally on the logic behind the component, as long as they belong to the same Path of the data tree. Generally.
this is much better explained in the links I previously posted, and is a fundamental topic to understand if we want to work with grasshopper.
Edit: Barry was right, it was "also" a problem related to Surface Override...
I double-checked this again and you're right, the surface override was on.
The issue is that the override is on by default.
- Do you know how to change that in the GH or alternatively in the composite wall settings in AC? I couldn't see a setting for that.
Ieceta, thanks for all the help so far, I appreciate it a lot.
I've corrected the reparametrization, my screenshot wasn't the updated one.
Now the walls are randomized by profile, the only issue is the surface override as I mentioned above.
- Did you have that as well and you copied the walls to the side and removed the override, or were the colors directly visible in AC?
to set wall default parameters: in archicad, with wall command activated (and making sure you havent any element, wall included, selected) change "surface override" parameter or any other parameter to set them as default.
As a continuation of this topic, I was wondering 2 things:
1 - I'm not exactly sure how the bezier graph mapper I'm using is actually determining the weight ratio for colors. I can get it to be 'all one color' for various colors but I'm not sure exactly how it's divided into sectors. Perhaps you know a good resource for this?
2 - I thought to use an MD Slider instead of a graph mapper, but the MD has no input, and it acts a bit weird regarding colors. Do you think it's wiser to use it?
3 - If I have a linear row of walls, such as the one you created, is there any component in GH that can have as input the colors of the profiles? I'd like to do
a "gradient-attractorPoint" combination to further expand the capabilities of this script but I don't think GH has something like that yet...
1- the input is a random set of numbers (abscissa), and the output values (ordinate) is ruled by graph curve. This way you control how likely a random input will generate an output between 0 and 3. its quite simple logic... For sure there will be a better way to obtain weighted random numbers...
2-MDsilder is a 2d version of slider. I dont understand how you would like to use it.
3-a sketch with your intentions would be helpfull...
I modified your image in photoshop to illustrate the color gradient. I would like to achieve this by using an attractor point at one end of the array for example, and set the color spectrum according to native composite/complex profiles of AC walls.
Is it possible?
Ordinate interval should match the number of material.
Domain interval match the abscissas interval, as indicated in the picture.
If you need great control on material distribution, may be you should think on other approaches, Graph mapper is just a intuitive tool, but you will no have a great control in the sense of "fine granularity", so to say..
bwt, this is not more about "material randomization"