Metadata-Version: 2.1
Name: symchaos
Version: 0.9.0
Summary: Symmetric Chaos
Home-page: https://gitlab.com/RolfSander/symchaos
Author: Rolf Sander
Author-email: mail@rolf-sander.net
License: UNKNOWN
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3.6
Classifier: License :: OSI Approved :: GNU General Public License (GPL)
Classifier: Operating System :: OS Independent
Classifier: Natural Language :: English
Classifier: Topic :: Multimedia :: Graphics
Requires-Python: >=3.6
Description-Content-Type: text/markdown
Requires-Dist: configparser
Requires-Dist: datetime
Requires-Dist: matplotlib
Requires-Dist: numpy

# Symmetric Chaos

The script symchaos.py generates chaotic images with translational
(wallpaper) or rotational symmetry, as pioneered by Field and Golubitsky
[1].

![](http://www.rolf-sander.net/software/symchaos/quilt.png)
![](http://www.rolf-sander.net/software/symchaos/rot9.png)

## Installation:

First, install the package with pip:

`pip3 install symchaos`

Next, find the location where the installed files are:

`pip3 show -f symchaos`

Copy the whole `symchaos` directory to a place where you have read and
write access. Now you can change into this directory and start the
script `symchaos.py`.

## Dependencies:

Symchaos needs the python modules `argparse`, `ast`, `configparser`,
`datetime`, `math`, `matplotlib.pyplot`, `numpy`, `os`, `pylab`,
`random`, `re`, `shutil`, and `sys`.

The pure python version is fine for testing but very slow. To speed up
the calculations, it is strongly recommended to use the Fortran90 code
`symchaos.f90` and use the fortran to python interface generator f2py
from numpy (https://numpy.org/doc/stable/f2py).

## Usage:

Execute `symchaos.py` and provide a config file from the `ini/`
directory, for example:

  `symchaos.py ini/example.ini`

For a quick test without installing a Fortran90 compiler, type:

  `symchaos.py ini/nof90.ini`

If `f2py` is available, the Fortran90 code can be compiled with:

`f2py -c symchaos.f90 -m symchaos_f90`

Now, all config files from the `ini/` directory can be used, e.g.,
`clifford.ini`, `fern.ini`, `mondrian.ini`, `quilt.ini`, `rot9.ini` etc.

The images are saved inside the `output` directory.

## Configuration:

Generation of the images is controlled via a config (`*.ini`) file. The
keys are:

- `method`: The method determines which formula is used by
  `symchaos.f90`. Available methods are:
  - `clifford`: Clifford attractor
  - `fern`: Barnsley fern
  - `hopalong`: Hopalong attractor by Barry Martin
  - `mondrian`: images reminding me of Piet Mondrian
  - `quilt*`: several quilts (wallpapers)
  - `rot_sym`: images with rotational C_n symmetry
  - `triangle`: images looking a bit like triangles
- `a`, `b`, `c`, `d`, `e`, `n`: These parameters are part of the
   formulas of the methods.
- `vary_steps`: If `vary_steps` > 1, then several images are created,
   using different parameters `a`, `b`, `c`, `d`, `e`.
- `vary_type`: If `vary_steps` > 1, then `vary_type` determines how the
   parameters are varied:
  - `cyclic`: starting and ending at the initial parameters
  - `neighborhood`: random values in the neighborhood of the initial parameters
  - `random`: random values
  - `random_walk`: a random walk, starting at the initial parameters
- `vary_value`: The magnitude of the parameter variation
- `palette`: The matplotlib color palette [3] used to create an image.
   The default is the rainbow palette, setting `palette = *` uses all palettes.
- `resolution`: A tuple defining the x,y resolution of the image, e.g.,
  `resolution = (100,100)`.
- `MAKETILES`: If set to `True`, a tiled image is also created. This is
   only useful for wallpapers with translational symmetry.
- `USE_FORTRAN`: If set to `True`, use the fast Fortran90 code
  `symchaos.f90`.

Further explanations can be found in the comments of the `example.ini`
config file.

## References:

[1] Symmetric Chaos,
    Mike Field, and Martin Golubitsky,
    Computers in Physics 4, 470 (1990),
    https://doi.org/10.1063/1.4822939

[2] Christmas in the House of Chaos,
    Scientific American, Dec 1992,
    JSTOR 24939340

[3] List of matplotlib palettes:
    https://matplotlib.org/stable/gallery/color/colormap_reference.html






