When to be Discrete: Analyzing Algorithm Performance on Discretized Continuous Problems - Data and Code

This repository contains the code and data for reproducibility of the paper 'Modular Differential Evolution'. 

The following files are included:

- experiments: Python files which were used to create the discretized functions and collect the data (the experiment_runner files) and the code for each of the tested algorithms (two subfolders)

- Process_discr: code which takes the raw IOH data and turns it into the relevant csv-files which are used to create the figures. CSV files are included 'csv_fb_*.zip'. ecdf_creation_script is used to extract ecdf and ert data for the ECDF_plot notebook. This data is added as 'other_csvs.zip'

- Plot*: notebooks which are used to generate the figures from the papers