Looking for a gmic + python scripter to help me out here.

Software G'MIC
89

I wanted users to have the option to use non-square seed image. Hence, why you’re seeing hyp_img, and the initial n being based off solving b for a=sqrt(2b^2). Non-square images is not quite the correct way of using this script, but eh I don’t really care about whether users will use a square image or not, and I do know how to create the option of forcing a non-square image to be a square seed.

I did that, it still doesn’t seem to work.

In addition, that can be simplified to:

NN=(N-1)^2
increment=1/NN
incrementValue=-NN
zeroV=4*NN

Here’s the new code which is a lot easier to read.

rep_hamiltonian:

repeat $! l[$>]
 if s>1 to_gray. fi
 ge. {avg(im,iM)}

 hyp_img={sqrt(sqr(w)+sqr(h))}
 n={int(sqrt(($hyp_img^2)/2))}
 nn={($n-1)^2}
 npts={sqr($n)}
 npts_n={$npts-$n}
 npts_1={$npts-1}
 
 increment={1/$nn}
 incrementValue={-$nn}
 zeroV={4*$nn}
 
 $npts,$npts
 
 $npts,5,1,1,"begin(
   const inc_v="$incrementValue";
   const zv="$zeroV";
   cv=[crop(#-2)];
   initial_pos_x=[0,0,0,1,"$n"];
   initial_pos_y=["$n",1,0,0,0];
   v(a)=inc_v*cv[a];
  );
  y==2?(
   i(#-1,x,x)=zv;
  ):(
   pos_x=x+initial_pos_x[y];
   pos_y=x+initial_pos_y[y];
   i(#-1,pos_x,pos_y)=v(x);
  );"
 k..
 eigen.
 k.
endl done

Also, this doesn’t work either:

 increment={(1/($n-1))^2}
 incrementValue={-1/$increment}
 zeroV={4/$increment}

EDIT: unroll instead of crop does not work either. I’m stumped.

1 Like