home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
PC World Komputer 1995 November
/
PCWK1195.iso
/
inne
/
dos
/
fraktale
/
fracxtr6.exe
/
OTHER
/
F(G(Z))).TXT
< prev
next >
Wrap
Text File
|
1993-06-24
|
4KB
|
92 lines
This is the Readme for files:
PHYSICS.FRM METHODS.FRM HALLWEEN.FRM PHOENIX2.FRM
F(G(Z))).FRM FRAME.FRM BESSELS.FRM
also author of tworain3.map, rainbow6.map, rainbow7.map.
----------------------------------------------------------------------------
Here are some of my best ??? fractal formulas. Some are whimsical
dreamings of my own. Others (most actually) are based on equations
that I have found from reading various books on chaos, dynamical
systems, and complex numbers. The formulas in the methods.frm file
are the result of my attempts to write Newton's method for something
more complex than z ^ n - 1. But, unfortunately that requires some
tricky programming and mathematics. So I settled for trying to use
the result as a normal equation. This produced some ungainly looking
objects. Then, thinking that setting the OUTSIDE setting to REAL was
like testing only the real protion of z. Boy!!! was I ever wrong.
But BOY!!! did it ever produce some interesting images. Some even get
more beautiful if you zoom out rather than zoom in. They begin to look
like biomorphs.
--------------------------------
LRCMIKE@LRC.OLDSCOLLEGE.AB.CA
Michael G. Wareman
P.O. Box 1856
Olds, Alberta, Canada
T0M 1P0
--------------------------------
-----------------------------------------------------------
Info on F(G(Z))).FRM (by michael g. wareman)
-----------------------------------------------------------
To me complex numbers, fractals, and the interesting images they spawn
are both fascinating and enjoyable. I would like to share with my fellow
programmers an interesting discovery I have made. My fractal is like
other fractals, that is, it is based on the basic mathematical operations
of complex numbers. The idea for this fractal came from a chapter in
Clifford A. Pickover's book titled: Computers, patterns, chaos, and beauty
on composite functions. What I did was ask: What would happen if I would
put the Mandelbrot fractal equation through a second equation. Below are
the two equations that I used to create FGZ (which is what I have come to
know this fractal as).
Z = Z * Z + C
Z = (3 * Z * Z) / (Z + 3) + C
If you have access to the Fractint program mentioned in Algorithm
issue 3.3, or available when you buy Timothy Wegner's, and Mark
Peterson's book (see further reading). You can use the following algorithm:
{NOTE: this formula and its variations are found in the F(G(Z)).frm on this
disk.}
FGZ {
z = c = pixel:
z = z * z + c;
z = (3 * z * z) / (z + 3) + c,
|z| <= 4
}
The first noteworthy thing about the image is that it appears to be two
separate sets. Upon closer examination it is actually one image. Another
interesting thing is that it has several floating images in front of it,
and shooting out from its branches. Here are some suggestions for some
interesting experiments to try. I have created some very interesting
Julia images based on the coordinates obtained from magnified portions
of the original fractal image. Since there were two "C values" (one in
each of the two equations) I decided to try replacing only one of the
two "C values. This resulted in distorted fractal images that tended to
be too chaotic. Who knows maybe there is a sensible fractal image and
I have not found it yet. If both "C values" are replaced by the same
coordinate set then a proper Julia appears exhibiting the behavior of
the magnified main image. Interesting Julias can be created by using
only the real or imaginary components of previously calculated Julias.
So far only the real component of the complex number generates reasonable
Julias. It remains to be seen if there are any imaginary components
that result in reasonable Julia images, or is there some other
relationship dependant on the location of the coordinate pair that
determines which value will result in good Julia images.
SOME C VALUES TO TRY:
For Fractint users : replace the c with:
(-0.6882, -0.1729)
(-0.6904, -0.2958)
(-0.4919, 0.4572)
(0.3379, 0)
Another thing to try is replace the constants (ie the 3's) with other
values. My initial experiments show that constant values less than 3
cause the arms to merge. In fact the value 2 you can still see little
fractal lakes with more little fractal lakes. See the fgz2-arm for what
I mean.