Simple Genetic Algorithm

This repo contains my first genetic algorithm in Elixir using Nx.

Rules are simple :

Each chromosome is composed of 26 genes (letters from A to Z)
Chromosomes are randomly generated for the first generation
The goal is to obtain a perfect chromosome (the 26 letters of the alphabet in order)

Requirements

- Elixir v1.12.0 (for the `Mix.install` function)
- Tested with OTP 24

Usage

> git clone https://github.com/ImNotAVirus/simple_genetic_algorithm_ex
> cd simple_genetic_algorithm_ex
> elixir genetic_algorithm.exs
Generation 1: the best chromosome is 'QNODDBHODXABXEEOTRFNSJMXNL' (fitness score: 3)
Generation 2: the best chromosome is 'QNODDBHODXABXEEOTRFNSJMXNL' (fitness score: 3)
Generation 3: the best chromosome is 'QNUCEBHYDXAUXREPTRONNQMXNL' (fitness score: 4)
Generation 4: the best chromosome is 'QNUCEBHYDXAUXREPTRONNQMXNL' (fitness score: 4)
Generation 5: the best chromosome is 'QNUCEBHYDXAUXREPTRONNQMXNL' (fitness score: 4)
Generation 6: the best chromosome is 'QPTIHWKHWXATTEENORSBSVXXNL' (fitness score: 5)
[ ... ]
Generation 3371: the best chromosome is 'ABCDEFZHIJKLMNOPQRSTUVWXYZ' (fitness score: 25)
Generation 3372: the best chromosome is 'ABCDEFZHIJKLMNOPQRSTUVWXYZ' (fitness score: 25)
Generation 3373: the best chromosome is 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' (fitness score: 26)
Generation 3374: the best chromosome is 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' (fitness score: 26)
Generation 3375: the best chromosome is 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' (fitness score: 26)

Notes

Currently all variables are hardcoded but I plan to create a basic library for Genetic Algoithms when I'll have more knowledge about Nx.

Any contribution (PR, issues, ....) to improve my code is welcome!