Comparing genetic optimisation algorithms: scratch developed algorithm against published algorithm

This repository contains all the scripts, codes and datasets for performing two optimisation algorithms (GA's), storing corresponding results and comparing them.

Requisites

Python is required for reproducing the analysis. Further documentation here.

Reproducibility

In order to reproduce the analysis, it is necessary to follow the subsecuent steps:

1. Run both algorithms
2. Compare results

Run both algorithms

As the aim of this work is to compare two optimisation algorithms (scratch developed algorithm & DEAP library algorithm), the user should run the two algorithms. The order in which they are performed is unimportant.

Scratch algorithm

Algorithm contained inside scratch folder within code. scratch contains two files:

• scratch.py: stores the complete implementation of the GA.
• scratch_analysis.py: performs the GA by calling objects from scratch.py.

For running and saving the findings of the scratch algorithm just excecute scratch_analysis.py, the two procedures mentioned will be automatically done (furniture_products.csv source dataset imported from data folder).

DEAP algorithm

Algorithm contained inside distributed folder within code. distributed contains one file:

• distributed.py: stores and performs the complete implementation of the DEAP GA.

For running and saving the findings of the DEAP algorithm just excecute distributed.py, the two procedures mentioned will be automatically done (furniture_products.csv source dataset imported from data).

Compare results

Results from both algorithms are stored inside results folder, divided correspondingly into two subfolders: scratch and distributed. For comparing the GA's open the comparison folder inside code, there will be three files:

• box_plot.py: creates a box plot comparing the objective values of the two GA's. Exports box_plot.pdf to ìmages folder.
• line_plot.py: creates a line plot comparing the score ~ space relation between both algorithms. Exports line_plot.pdf to ìmages.
• cleaning.py: contains a specific function in order to clean the results dataframes. box_plot.py and line_plot.py use this function.

Report

Report available as report.pdf inside report folder. To replicate it please download and unzip report.zip, then compile it using LaTex.