Pyaccell

A GPU Accelerated Cellular Automata Library. Allows you to create totalistic Cellular Automatas, by specifiying rules of the CA.

Installation

Recommended Python version: 3.10 and Requires: OpenGL Core profile supported GPU

pip install pyaccell

Building the project from source

git clone --recurse-submodules https://github.com/ahmedhus22/pyaccell.git
pip install .

To Build wheel:

pip wheel .

Pyaccell Documentation

Import the package as:

import pyaccell

or

import pyaccell as CA

you can now use >>>help(pyaccell) to get a brief overview of the package.

Automata Class

This class defines methods and attributes required to create and use cellular automatas.

Constructors:

CA.Automata(rule, states)
CA.Automata(rule, states, sim_width, sim_height)

where rule is a 1D list that defines Automatas transitions for all states and neighbours. states is the number of states in Automata. width and height defines size of automata.

To create a rule define a transition function:

def transition(state: int, neighbour: tuple):
    ...

state is current state and neighbour represents, neighbour count of each state. for example: neighbour[1] = count of neighbours having state 1.

Now use the transistion function to creat rule array:

rule = CA.create_rule(transition, states)

Alternatively, you can define a rule array yourself:

The Rows represent index, use get_index(neighbour: list, states: int) function. For 2 state CA index is just the count of neighbours having state 1.

Columns represent states, thus [state] [index] gives next state.

Class Attributes:

• sim_width: int (default = 800)
• sim_height: int (default = 600)

Sets the width and height of the framebuffer to handle simulation, it defines the number of cells.

• input: list

It has to be of size sim_width * sim_height, otherwise the Automat runs with random inputs. It contains the initial state of automata as a list of ints. By default it runs with random input.

• output: UIntVector

It is read only, it gives the output of cellular automata, if it is run for set number of iterations. To read output in python list use vec_to_list(output) function.

Class Methods:

A run() method is provied, It has 2 overloads.

• run(): Automata runs indefinitely with given input and rules in a window.
• run(iterations: int): Automata runs for set iterations and then halts. The Output is stored is output attribute of the class.

Useful Helper Functions:

• get_index(neighbour: list, states: int) -> index: int
• vec_to_list(output) -> list

Note : python 3.10 is required for this function.

• create_rule(transition, states) -> list

References

• Rule Representation of multistate Cellular Automata using combinatorial indices: Benjamin Mastripolito's Article