Future Render Plugin: Martin Attractors

[jerkfight]

Example:

I found this to be a nice render/fractal.

Heres a quote taken from http://spanky.triumf.ca/www/FRACTINT/ma ... _type.html

These fractal types are from A. K. Dewdney's "Computer Recreations" column in "Scientific American". They are attributed to Barry Martin of Aston University in Birmingham, England.

Hopalong is an "orbit" type fractal like lorenz. The image is obtained by iterating this formula after setting z(0) = y(0) = 0:

x(n+1) = y(n) - sign(x(n))*sqrt(abs(b*x(n)-c))

y(n+1) = a - x(n)

Parameters are a, b, and c. The function "sign()" returns 1 if the argument is positive, -1 if argument is negative.

This fractal continues to develop in surprising ways after many iterations.

Another Martin fractal is simpler. The iterated formula is:

x(n+1) = y(n) - sin(x(n))

y(n+1) = a - x(n)

The parameter is "a". Try values near the number pi.

Michael Peters has based the HOP program on variations of these Martin types. You will find three of these here: chip, quadruptwo, and threeply.

Is this possible to achieve in Paint.NET

[MadJik]

This is hard to code as the pixels aren't calculated in the standard x/y loop.

The result doesn't correspond to the image...

[jerkfight]

This is hard to code as the pixels aren't calculated in the standard x/y loop.

The result doesn't correspond to the image...

Its ok, i figured how to achieve something alike

[MiguelPereira]

This is hard to code as the pixels aren't calculated in the standard x/y loop.

The result doesn't correspond to the image...

I'm sorry to contradict but I think your wrong the formulas are right there with the x's and y's...

Yes it is calculable

[Myrddin]

MiguelPereira: I'm no coder, but I'm assuming from MadJik's response that they are the wrong type of x/y's for plugin authoring, and that maybe more work is required than simply what is presented there.

That may have been phrased wrong, though I'm sure the point is still percievable.

[MiguelPereira]

Yes I know but I had today an examination that had these kind of formulas to code, and i'm now trying to transpose these to codelab...

[barkbark00]

Be sure to post your results. I'd be interested to see what you get.

[MiguelPereira]

Be sure to post your results. I'd be interested to see what you get.

me too

okay i'm having some trouble, I've written the code on Matlab to see if it worked, and i can see reults, but i'm not able to program on codelab.

[MadJik]

The effect in CodeLab needs this formulas to be reversed...(how?)

This is an excel file that shows some result:

[MiguelPereira]

Aww how come i didn't remembered that? anyways thats pretty much what i got on matlab...

The effect in CodeLab needs this formulas to be reversed...(how?)

I got there too but still reversing them will be a headache...

EDIT: I know i should look for myself on book but how do you handle vector variables on c#?

[MadJik]

Vector?

or polar!

http://en.wikipedia.org/wiki/Polar_coordinate_system

[MiguelPereira]

sorry vectors on c#, variables with dimension above 1

[MadJik]

Still not see the point, sorry, I'm programer (not math-er)...

But perhaps, this code of Martin's fractal could help you... Let me know.

int Amount1=1500;   //[0,9999]a (/10)
int Amount2=1000;   //[0,9999]b (/10)
int Amount3=1000;   //[0,9999]c (/10)

bool init = false;
int sAmount1=-1;
int sAmount2=-1;
int sAmount3=-1;

void Render(Surface dst, Surface src, Rectangle rect)
{
 double Xa1,Ya1,Xa2,Ya2;

 Rectangle selection = this.EnvironmentParameters.GetSelection(src.Bounds).GetBoundsInt();
 long cX = (long)(((selection.Right - selection.Left) / 2)+selection.Left);
 long cY = (long)(((selection.Bottom - selection.Top) / 2)+selection.Top);
 double a = (double)Amount1 / 10.0;
 double b = (double)Amount2 / 10.0;
 double c = (double)Amount3 / 10.0;

 long MaxiP = 8000;

 if ((!init) || (sAmount1 != Amount1) || (sAmount2 != Amount2) || (sAmount3 != Amount3))
 {
   init = false;
   sAmount1 = Amount1;
   sAmount2 = Amount2;
   sAmount3 = Amount3;
 }

 ColorBgra PrimColor = (ColorBgra)EnvironmentParameters.PrimaryColor;
 int BrushWidth = 1;// (int)EnvironmentParameters.BrushWidth;

 SolidBrush Brush1 = new SolidBrush(Color.FromArgb(PrimColor.A, PrimColor.R, PrimColor.G, PrimColor.);
 Pen Pen1 = new Pen(Brush1, BrushWidth);
 //Pen1.StartCap = System.Drawing.Drawing2D.LineCap.Round;
 //Pen1.EndCap = System.Drawing.Drawing2D.LineCap.Round;

 // Calculate the X,Y coords of the curve
 double [] sx1 = new double [1 + MaxiP];
 double [] sy1 = new double [1 + MaxiP];

 if (!init) 
 {
   sx1[0] = 0;
   sy1[0] = 0;
   for (int j = 1; j < MaxiP; j++)
   {
     int signx = 1;
     if (sx1[j-1]<0) signx=-1;
     sx1[j] = sy1[j-1] - signx * Math.Sqrt(Math.Abs(b*sx1[j-1]-c));
     sy1[j] = a - sx1[j-1];
   }
   init = false;
 }

 // Copy existing image (source) to destination
 for (int y = rect.Top; y < rect.Bottom; ++y)
   for (int x = rect.Left; x < rect.Right; ++x)
     dst[x, y] = src[x, y];

 Graphics g = new RenderArgs(dst).Graphics; 
 g.SmoothingMode = System.Drawing.Drawing2D.SmoothingMode.AntiAlias;
 g.Clip = new Region(rect);
 Xa2 = cX + sx1[0];
 Ya2 = cY + sy1[0];
 for (int i = 1; i < MaxiP; ++i)
 {
   Xa1 = cX + sx1[i];
   Ya1 = cY + sy1[i];

   g.DrawLine(Pen1, (float)Xa1, (float)Ya1, (float)(Xa1 + 1), (float)(Ya1 + 1));

   Xa2 = Xa1;
   Ya2 = Ya1;
 }
}

[jerkfight]

Thanks Madjic for trying to acomplish this.

I tried making it, its close but not close enough:

[salu]

doesnt a grid and wobble and rainbow work??

[jerkfight]

doesnt a grid and wobble and rainbow work??

Thats practically wat i did. grid, sine waves,new layer, gradient (plugin), then overlay mode

[MiguelPereira]

Still not see the point, sorry, I'm programer (not math-er)...

But perhaps, this code of Martin's fractal could help you... Let me know.

...

Well you're far more experienced than me at coding PdN's plugins and this one is well done... i was get there by a diferent prespective but your way is the actual working one so... Good Job

EDIT: SO can you improve this making it with a little more quality on the render process? I know I can't at least for now, I have to learn C# first, and study BoltBait's codelab tut's, so can you do it, the hard part it's already done, I can do the beta testing and so, if you need help ...

[MadJik]

I would like the reverse the formulas first (anyone?) to change and improve the code.

In this code version you could increase the number of points to calculate:

long MaxiP = 8000;

Warning: it will take time and memory wth possibly error message (out of memory)...

MaxiP = 120000 (3min)

[MiguelPereira]

Invert how?

[MadJik]

The loop is :

for each point p, calculate the x and y values corresponding to the formulas and plot a point at x,y position.

To reverse it, I need :

for each x,y couple, calculate if the point is plotted.

[MiguelPereira]

Would that work on improving the performance?

[MadJik]

I think so...

for each point p, calculate the x and y values corresponding to the formulas and plot a point at x,y position.

This could give several times (many) the same point (with the rounded values)...

[MiguelPereira]

Can you work on a third surface besides the destination and source ones?

cause you could print the calculated pixels onto the third one to keep tracking of which were already plotted

EDIT: Nah forget it it would take way to much memory

[MadJik]

Some plugins are using this method already!

It's not really so much more memory, but it won't reduce the calculation time...

[MiguelPereira]

Well it's late but I'll see if I can dig up any numerical method to enhance the calculations, keep in touch