EA presentation

Code repository for the presentation on evolutionary algorithms "What can we learn from evolution?" for Capgemini August 25th 2023.

The algorithm showcased here is a variation on the simple genetic algorithm (SGA), whit the following parameters:

Parameter	Value
Representation	Bit-strings
Recombination	1-Point crossover
Mutation	Bit flip
Parent selection	Tournament selection*
Survival selection	Generational

*Traditionally, roulette wheel has been used by the SGA. Tournament selection was chosen here instead due to its simplicity and speed.

Installation

To install the project dependencies, run the following at the terminal:

pip install -r requirements.txt

Run the optimizer

The main script is located (conveniently) in main.py.

To start it, run the following at the terminal:

python main.py

Unit tests

The unit tests are located in the tests/-folder.

To execute them, run the following at the terminal:

python -m pytest tests/

Results

The results of the optimization become available after each run in the results/-folder.

Each run contains the following:

📂results
┣ 📁1
┣ 📁2
┣ ...
┣ 📁3
┣ ┣ 📈ga.png        # generational plot of the GA optimization progress 
┣ ┣ 🧾metrics.csv   # population metrics for each generation (generation, baseline, best, average, entropy)
┣ ┣ 🐍parameters.py # parameters used to obtain the respective results
┣ ┣ 📄solution.txt  # best solution found at termination