Quadrification possible with Kangaroo?

Replies to This Discussion

Permalink Reply by Riccardo Majewski on March 27, 2015 at 6:57am

Interesting challenge.

Got few time. Sorry, not sufficient descriptions :P

I've used an iterative method > anemone plug-in... To load my .gh you'll need it.

Results are not "cool" as in the site of your link.

Tweaks needed.

You can maybe use this as a start.

If you have more info about how this graph should work, write it here.

You'r opening post is a bit "minimal"... (too much "How-to-do-this-pic.")

Good luck!

Attachments:

• quadrigram.gh, 37 KB

Permalink Reply by bea franko on March 30, 2015 at 8:12am

Wow!

pretty cool stuff! this helps a lot!

how to cahnge the diagram to akt like this:

thank you so much!!!

Permalink Reply by Riccardo Majewski on March 30, 2015 at 1:18pm

I see the logic you aim to.

I've tought of that... fill x > fill y > x > y > x > y > etc

But thinking about that: with an extremely wide rectangle this wouldn't work well.

x,y,x,y sequence can be done only after the starting rectangle is "shrinked" to a squared-like rectangle. And "shrinking" will be done with a x,x,x,x or y,y,y,y sequence... automated..

I'll follow this instead: "place squares on the shortest side of the remaining rectangle to fill".

Should be similar enough to what you have said.

Permalink Reply by Riccardo Majewski on March 30, 2015 at 2:13pm

Here, done.

Edited method logic as said and fixed some bugs.

Seems good.

Added some description panels.

Tell if something still unclear or if you want to change core algorythm logic.

EDIT: my first solution, instead, somehow have pros and cons: it were "selecting" side to work depending of how well the new rectangles would fit, making them more squared-ish in total, but with the last (or lasts) "rows" suffering for all the job done to other rows before...

With version 2 here, "errors" (as not being a perfect square) are "spreaded" to all rectangles, instead procrastinating most of it at the end...

Probably with pure (deep(deeper)) math, a better solution could exists...

The "algorythm" i've used here "watch" only a bit of work at time, not the whole.... thus, it's impossible this is the best solution. (Maybe :D )

Attachments:

• quadrigram_V2.gh, 33 KB

Permalink Reply by Riccardo Majewski on March 31, 2015 at 5:24am

fixed a bug (big one)

Areas are correct now.

Attachments:

• quadrigram_V3.gh, 33 KB

Permalink Reply by bea franko on March 31, 2015 at 5:59am

thank you so much!

excellent work!

Permalink Reply by Mikkel on December 14, 2015 at 8:53am

Hi Riccardo.
This is a really cool script! it is almost what i am going for, however in another context. I have tried for some time now to alter your script so I would be able to control the placement of the different sizes of rectangles according to an attractor line og point, but wit out luck. Do you have any thoughts of how this could be implemented?
Thanks.
Mikkel

Permalink Reply by Riccardo Majewski on December 14, 2015 at 9:28am

Hi Mikkel.

What you described is somehow "generic".

Can you post some hand drawings of what you need?

Like: "with this input (pic) i want this output (pic)"...

so I/we can understand you better.

I think somehow the solution/definition will be far different...

This one is too much "orthogonal" to work with curves...

Permalink Reply by Mikkel on December 14, 2015 at 9:33am

Right. I made this post earlier today.
http://www.grasshopper3d.com/forum/topics/random-rectangular-grid