(Simple) Genetic Algorithm

Simple genetic algorithm with Ackley's function optimization.

Author: Dominic Hollis (GovernmentPlates) -- Uni Coursework (May 2021)

---

Running the GA

Make sure that all the files main.py, GAFunctions.py and GAConfig.py are in the same directory.

Then run main.py from your Python CLI -- after a few seconds (depending on your parameters), you should see a GAOutput.csv generated. The CSV contains the population coordinates, which can then be fed into a program like Excel to produce a graph.

IMPORTANT: Make sure that you are reading the output of the console for any messages.

Adjusting the parameters of the GA

The script will read from the configuration file called ga.config. You can open this file up in a text editor of your choice and change around the parameters on there.

If this is the first time you are running the script, it will generate a configuration file with default parameters for you. Once it does this, re-run main.py.

Example ga.config file:

populationSize:100
maximumGens:1000
minBoundsSolutionConstraint:-5
maxBoundsSolutionConstraint:5

populationSize Set's the size of the population (used in generating the population for the first time).

maximumGens Set's the maximum number of generations that the GA will go through.

minBoundsSolutionConstraint A minimum boundary value (used in generating random numbers).

maxBoundsSolutionConstraint A maximum boundary value (used in generating random numbers).

NOTE: As per the [university assignment] requirements, the minBoundsSolutionConstraint and maxBoundsSolutionConstraint must be set at -5 and 5 (constraints).

Features Implemented

• Population generation
• Genetic crossover + Mutation
• Child generation
• Ackley function optimization (used in the fitness function)
• Configuration system (innovation?)