galgopy

A lightweight Python package providing a basic but flexible implementation of the Genetic Algorithm (GA) for optimization tasks. It can be easily integrated into other projects or extended with custom selection, crossover, and mutation strategies.

Installation

galgopy is available on PyPI:

pip install galgopy

Usage

Sphere Function Optimization example

import numpy as np

from galgopy.crossover import IntermediateRecombination
from galgopy.ga import GA
from galgopy.mutation import UniformMutation
from galgopy.selection import TournamentSelection

ga = GA(
    population_size=100,
    chromosome_size=2,
    gene_bounds=(-5.12, 5.12),
    fitness_func=lambda x: -np.sum(np.square(x)),
    mutation=UniformMutation(individual_rate=0.1, gene_bounds=(-5.12, 5.12)),
    crossover=IntermediateRecombination(),
    selection=TournamentSelection(),
)

# The course of 50 generations.
for _ in range(50):
    next(ga)

# Result
best_chromosome, _ = ga.best_chromosome()
print(
    f"Result:\t{best_chromosome}",
    "Global minimum:\t[0. 0.]",
    sep="\n",
)

More examples

More examples can be found in the examples directory.

• Booth Function Optimization
• One Max Problem
• String generation
• Travelling salesman problem (TSP)

License

This project is licensed under the MIT License.