Grasshopper

algorithmic modeling for Rhino

Request is for a dynamic multislider object, which creates X number of sliders based on an integer.  Ideally this object would be connectable to Galapagos and understand each slider as a separate genome.  

This would be incredibly useful for creating dynamic systems that have a different number of controllable elements, depending on the geometry, use case, etc.  As it stands, it is necessary to create a large number of dummy sliders that represent the maximum number of elements in the system.

Would also be nice to have the same type of object for toggles, etc.  Even nicer would be able to define a "group" of objects that would duplicate ad infinitum.  Attached is an image of a section of a definition that uses the old way (note I'm using panels to identify which slider banks are "active").

Views: 3313

Attachments:

Replies to This Discussion

Isn't that what gene pool is? Or do I misunderstand?

Gene Pool is on the right track but is missing a dynamic input.  I'd like to be able to provide an integer like "15" to the component to determine how many sliders there are.  I would also like to be able to determine their extents dynamically (not the same extents for all sliders), instead of through the UI...

Ah, got ya. Yea I think that could also benifiet the image sampler domain.

From what I understand, genomes and individuals are described by their genome vector. With equal length genomes, you can easily define genome proximity. Justi like distance of two points in 3D space. This is essential for combining individuals for the next generation.

Even in biology a successful breeding recombines two half sets of chromosomes to a new full set. Two unequal sets would end up in chaos. Imagine you have a 2D point and one in 3D but you cannot tell whether those 2D values are XY XZ or YZ, how would you measure their distance and how do you calculate the average of both? Of course you can have the 2D coordinates in slots where 1 is X, 2 is Y and 3 is Z but that is just a 3D point with one coordinate set to 0. This is what you do now with a lot of sliders in the gene pool, shortening the list at will.

Since you will feed the genome into Galapagos and probably will never manually adjust the values with the sliders, you can kind of mimic variable lengths. All you need is a function that will deterministically generate a variable number of n values from constant number of input parameters. Don't try to make the function too random or complex as this will break the porximity logic of Galapagos.

RSS

About

Translate

Search

© 2024   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service