Strange Behavior in 2D Gradient Random Point Selection

[e.tamrabadi]

Hello everyone.
I was working on a fragment to make something like “Attractor Geometrical Possibility” Component in Heteroptera. A 2D gradient semi-random attractor. And I saw these weird selection patterns in some seed numbers! I made some sliders in red groups so you can choose. Anyone has any idea where are these patterns come from? What is going on inside the random component so the effect of the three of them together, can produce a non-random result?! 2D Gradient Random Selection.gh (15.2 KB)

[e.tamrabadi]

It is attached.

[ajarindia (BVR)]

My mistake. I don’t know how I missed it.

[laurent_delrieu]

Hello if I remember well @DavidRutten said random in GH1 is not so good (uses Random from .Net). GH2 will have new random methods. But I can’t find this discussion!
I am quite sure there are some discussions on Random .Net class on others forum like Stackoverflow

Random not that random

[wim (Wim Dekeyser)]

Hi Laurent -

@laurent_delrieu

But I can’t find this discussion!

Here it is:

Random Walk Discrepancy

As well you should. It relies on the .NET random engine which is not a particularly good one. Different seeds tend to generate related sequences based on their ratios. You can clearly see this when populating a cube using seeds 1, 2, and 3 respectively for the x, y, and z components. In grasshopper 2.0 I’ve added many more random engines and picked a better default one. Specifically, Mersenne Twister, PALF, and Xoshiro256 will be available and they are considered very good ones, although not g…

-wim

[e.tamrabadi]

Thanks @laurent_delrieu and @wim for the responses. Actually I was always suspicious about the random component. One other example that I found out is if you create a random range of numbers, then apply a jitter component to the list (by the factor of 1 and by “the same seed number”), it will more or less become sorted rather than being shuffled!!! Please consider I said more or less. Cause there is some displacements in the sorted list by jitter :)) Yet this is strange that the jitter somehow neutralizes the function of random component! It seems they both use the same inner functions inside. Random, then Jitter.gh (12.5 KB)

[wim (Wim Dekeyser)]

Hi -

@e.tamrabadi

It seems they both use the same inner functions inside.

I really have no clue at all how all this works but if you want something that somewhat resembles randomness, it seems like a rather bad idea to use the same seed value for those components.

-wim