genyal 0.1.11

A framework for genetic algorithms in Python

Genyal: The cheerful genetic algorithm framework

This work is licensed under a Creative Commons Attribution 4.0 International License

Genyal is a genetic algorithm framework aimed to be as simple to use as it can be. Evolution is natural, so using evolutionary algorithms should feel natural as well.

One of the main benefits of evolutionary programming is making it simple to solve complicated problems, and that's something that most implementations of this kind of algorithms seems to forget. You'll find either, very simple implementations of genetic algorithms from scratch, or full fledged libraries that are, by no means, something that someone with little experience in the field of evolutionary can easily use; nothing in between. Genyal is that missing "in between", by providing a simple interface so users new to the topic can learn and use it without need of being experts and flexible enough so that people can adapt it to mor sophisticated scenarios.

Usage

Just like using it, installing Genyal should be a piece of cake (great, now I want cake). To do so, you can get it directly from PyPi or, with pip:

pip install genyal

Examples

Guessing a word

Let's check a very simple example. A program to guess a given word.

First, we need a way to create the individuals of the population, in this case, each individual will represent a word, and each gene of the individual will be a letter (lowercase). For this we can implement a generator function like:

import random
import string

def random_char():
    return random.choice(string.ascii_lowercase)

And then we create a factory with that function as its generator:

from genyal.genotype import GeneFactory

gene_factory = GeneFactory[str]()
gene_factory.generator = random_char

Next, we're gonna need a way to check how close the individual is to our target word, let's try to make it guess owo.

from genyal.engine import GenyalEngine

# This is how the engine will select the fittest individuals
def fitness_fun(word: list[str]) -> float:
    return sum([word[i] == "owo"[i] for i in range(0, 3)])

# This is the condition to stop the evolution
def target(genyal_engine: GenyalEngine) -> bool:
    return "".join(genyal_engine.fittest.genes) == "owo"

Now, we'll start the engine. The engine will take care of maintaining and evolving the population.

engine = GenyalEngine(fitness_function=fitness_fun, terminating_function=target)
# We create an initial population of 16 words (Individuals) with words of 3 characters (genes)
# using our previously defined gene factory.
engine.create_population(16, 3, gene_factory)
engine.evolve()
print(f"Found solution in {engine.generation} generations")
print(
    f"Fittest individual: \"{''.join(engine.fittest.genes)}\" with fitness: "
    f"{engine.fittest.fitness}")

If you run the code the script will print something like:

Found solution in 543 generations
Fittest individual: "owo" with fitness: 3

More complex examples will be added in future iterations.