By placing the same data on a different location.

1. void setup() {
2.   size(224, 56, P2D);
3.   loadPixels();
4. }
5. void draw() {
6.   background(0);
7.   for (int y=0,x=0;y<56;y++,x=x+224) {     
8.     pixels[x + y]=color(255);
9.     pixels[(x + y)+56]=color(255);
10.     pixels[(x + y)+112]=color(255);
11.     pixels[(x + y)+168]=color(255);
12.   } 
13.   updatePixels();
14. }

You asked the same question at two different places, that's bad practice in forums.
I asked in the previous thread, not knowing you had an answer here.
I would normally delete this thread, but that would delete the given answer, so I won't do it, but avoid this.
Or link the new question in the old topic, so no duplicate work is done.

Hi PhiLho , sorry, i asked this  in another thread so i decided to create a new thread only for this .

The problem is that with your approach and with el_dippos approach i get irregular patterns of points, while my example is a regular simetric pattern of points.
I dont know which values do i need to have the same patterns.

any idea?

1.    loadPixels();
2.    for (int i = 0; i < 32; i++) {
3.         for (int j = 0; j < 56; j++) {
4.           pixels[(j * 30) * 532 + i * 30] = c;
5.         }
6.    }
7.    updatePixels();