Are there fractal sets in C^2? (Would be 4d, wouldn't they?)
--tommy
*************************************************************
* semper idem! * "O Frabjous Day! *
* ca...@etek.chalmers.se * Callooh! Callay!" -L.C.*
*************************************************************
>Does anybody know of a 3d fractal set, e.g. a 3d mandelbrot?
>(And I don't mean the graphic presentations using heights
>instead of colours - rather globes instead of circles, and
>an "apple" instead of the main cardioid.)
>
>Are there fractal sets in C^2? (Would be 4d, wouldn't they?)
Consider the 'vanilla' Julia and Mandelbrot sets:
Z(n+1)=Z(n)**2+C
The difference is that in one, Z changes and in the other C.
Now, what if *both* Z and C varied? This makes four dimensions
(Real Z, real C, imag Z, imag C).
Fractint has a fractal called 'julibrot' that displays a 3-d slice of
this 4-d fractal.
It looks quite neat. (Rather an eye-bender, though.)
--
|Jeff Epler Additions Welcome c(-8 ;-) >{8-) |
| :) (=( =-] (-= Celebrating the variety of faces =-> :^) {-= |-) (: |
| Lincoln, Nebraska|
I saw this mentioned in the latest "Algorithm" journal...The method used was
to vary the exponent (Z <- Z^p + C) and use p as the third dimension. I didn't
get a chance to carefully read the article, but I seem to recall it saying that
something happened near the integers...Now, I am nowhere near an expert in the
fractal area, but it occured to me that Z^p would not be really well-defined
for non-integral p. The article used exp(p*ln(Z)), but in the complex numbers
ln(Z) could be one of an infinite number of things (separated by multiples of
2\pi), and if p is not an integer, you'll get different values of exp(p*ln(Z))
depending on which you choose. But like I said, I didn't read the article that
closely, so this issue might have been brought up in the article itself...If so
I apologize...
--
-- Erick, the perfect square :-)
There are fractals in the 4d realm of complex numbers, called the quaternions.
Basically, a quaternion Q = x + i*y + j*z + k*t, where i*i = j*j = k*k = -1,
and i*j = k = -j*i, j*k = i = -k*j, and k*i = j = - i*k. 3d slices of the 4d
Mandelbrot set (let the k part of c = 0) do reveal spheres instead of circular
disks. Kind of a disappointment, really, to get these extra dimensions and
not see something new and exciting (not that I've seen all of 4d fractal
space! :-)
Kerry Mitchell
One can specify an Iterated Function System in any number of
dimensions, e.g. in 3d. An example is the famous fern of
Michael Barnsley, extented to 3d as shown on the inside of the
cover of his book `Fractals everywhere'.
Of course are fractals generated with IFS of linear type
(when using affine transformations, as mostly done), not
like the Mandelbrot-set that is non-linear.
--
Michael A. Neuhauser Dept. for Pattern Recoginition and Image Processing
m...@prip.tuwien.ac.at Technical Univerity of Vienna, Treitlstr. 3/183/2
A-1040 Vienna AUSTRIA
Does anybody know of a 3d fractal set, e.g. a 3d mandelbrot?
=============================================================
There are many ways to generate 3d fractal sets. Several other
contributors to this group have pointed out a few. Another
method which incorporates the Mandelbrot set construction is
to treat the usual Z -> Z^2 + C map of the complex plane as if
it were a map of the real plane into itself. That is, the eqns
would look like :
X -> X^2 - Y^2 + A
Y -> 2XY + B
Where Z = X + Yi and C = A + Bi .
All that's necessary to extend to 3 dimensions is to add the eqn
for the real Z component. In general, iterates of R^3 into R^3
can be written as :
X -> F(X, Y, Z)
Y -> G(X, Y, Z)
Z -> H(X, Y, Z)
Make something up !
Another kinda cool way to get 3-D fractals from iterates of maps
is to apply some simple numerical integration scheme to a 3-D flow.
For example, take the Lorenz eqns :
Xdot = -sX + sY
Ydot = -XZ + rX - Y
Zdot = XY - bZ
Where s, r, b are positive constants and dot means differentiation wrt time.
Applying the Euler forward differencing scheme of numerically integrating
these eqns with a time increment t will give a 3-D iterated map which
approximates the Lorenz eqns and gives rise to a chaotic attractor in 3-D.
-rr-
--
Verum est ... quod superius est sicut | The truth is that what is above is like
quod inferius et quod inferius est | what is below and what is below is
sicut quod superius, ad perpetrando | like what is above, to accomplish
miracula rei unius. | the miracles of the one thing.