Re: Iterating the error function on the complex plane?

On Dec 2, 3:42 pm, glenn <glenn...@xxxxxxxxx> wrote:
O/H mike3 ??????:

On Dec 2, 2:53 pm, glenn <glenn...@xxxxxxxxx> wrote:
<snip>
Sorry to say but all these financial calculations qualifies you as more
than good for research in z(n+1)<--erf(z(n)).

"Sorry"?

That's good if I could do it. But I need that algorithm,
so I can compute it on the complex plane.

For Christ sake, what can be so difficult to write a program:

1. Put an initial value z<-z0.
2. Run for 1 to say 10000 the routine z(n+1)<--erf(z(n)).
3. See if z(n) converge.
4. If yes color that pixel z0 with yellow.
5. If no color it black
6. Go to the next z0, according to x=x+1/10000.
7. Finishing with the x's, set y=y+1/1000.

I did it 20 years ago (with a different recursion) in an IBM PS/2 8Mhz
and got the Mandelbrot fractal after 15 minutes.

Nothing about this is so difficult: the rub is calculating erf()
itself. The series does not converge quickly for reasonably large z.
And if the map behaves chaotically, who knows how large that z
might get?
.

Relevant Pages