NaturalComputingProject

The code and pdf of our project for Natural Computing At Radboud University

Manual

1. Paste the script found in sketch_script.js in the p5.js studio or run it locally.
2. Adjust the top parameters as desired. To reproduce the experiments from the paper you need to vary the parameters:
   • FoodCoolDown
   • N_food
   • The name of the file that you want to save as in WriteResults saveJSON(weights, "FoodCoolDown10/Run" + G.toString() + '/results.json');. Note this is a code which runs on the web and it has no knowledge of local directories etc. The final output will be written to a file.
3. The result is downloaded as .json file. We ran the experiments 10 times so run it as well at least some times and then place them all in a folder.
4. Open the notebook called Data_analysis.ipynb. Only change these lines:

results = read_and_average('experiments/NameOfMyFolder',names,30,15)
visualize(results, 'results')

NOTE:

1. The code here also has an implementation for Tournament K selection which we ended up not using.