Abnormal Mutations: Evolution Strategies Don't Require Gaussianity

# Requirements

In order to run the algorithms and create all the figures for the paper `Abnormal Mutations`, we need the following Python packages installed. The Python version that was used to generate the data is Python 3.10.12.

    - iohinspector==0.0.1

    - numpy==1.26.4

    - matplotlib==3.8.3

    - modcma==1.0.9

    - pandas==2.2.2

    - polars==1.17.1

    - seaborn==0.13.2

    - scipy==1.12.0

# Scripts

All scripts and notebooks used to generate the data can be found in the repository: https://github.com/IOHprofiler/ModularCMAES. To get the version that was used here, please use the v1.0.9 tag, or download the corresponding release. Specific scrips are also included here for convinience.

# Collecting data

All the data can be collected by runnning the `run.sh` script, found in the directory scripts/distributions. This calles the script `run.py`, which executes the algorithms, and stores performance data in the `data` folder. This folder is also uploaded as a zipped directory on zenodo.

# Creating figures

The notebook `plots.ipynb` can be used to generate all the non-ecdf plots in the paper. The notebook `ecdf_plots.ipynb` can be used to generate all the performance plots used in the paper. The second notebook requires the installation of the `iohinspector` library, currently in pre-release. It can be downloaded and installed via github: https://github.com/IOHprofiler/IOHinspector. The version used here can be accessed via the commit: 8e6bb06. All figures from the paper, included a few additional ones, are included on zenodo.