Skip to content

Latest commit

 

History

History
86 lines (54 loc) · 2.57 KB

README.md

File metadata and controls

86 lines (54 loc) · 2.57 KB
Raw

mandelpy

A Mandelbrot and Buddhabrot viewer with GPU acceleration using NVIDIA's CUDA toolkit.

Showcase

What makes this different?

There are tonnes of mandelbrot programs, but this is special in 3 ways:

  • It is the only one that provides a transform setting to everything. This means that every point z on the screen goes through a that transform which can completely change the shape of the output into something never seen before.

  • There is also zero randomness for all buddhabrots which is rares with these projects. I like to think that it makes computation faster.

  • It is GPU accelerated and is sometimes faster than previous accelerated Java and C++ applications even though the whole project is written in pure python.

Installation

You can clone this repository however you feel.

One way would be using the command line:

git clone https://github.com/Maelstrom6/mandelpy

Requirements

  • Python 64bit

  • NumPy

    Standard I assume.

  • Pillow

    The main image manipulation library.

  • imageio

    To create GIFs if wanted

  • Numba

    Requires the Python 64bit interpreter. Chances are, if you managed to install TensorFlow for GPU, you would simply be able to install it with no problems and no extra steps. Otherwise, according to their documentation, one needs the CUDA Toolkit 8.0 or later. Please run the numba_cuda_test.py to check if everything is working properly which is located in user_utilities.

  • PyQt5

    Required for the user interface. PyQt5-tools is not required.

Usage

Please see program_examples and the showcase for examples on how to use the core package.

Please run main.py in the gui folder in order to use the GUI.

Goals

  • Add new color schemes.

  • Add tooltip text to all settings on the GUI.

  • Set up 'undo' and 'redo' functionality in the GUI.

  • Decide how the GUI should look on startup and after pressing 'New'.

  • Find a way to evaluate string input as functions without the eval method.

  • See if I can form it into a PyPi package. How-to.

Contributing

Contributions are highly appreciated especially bug fixes, new color schemes and presets.