How Sequential Algorithm Portfolios can benefit Black Box Optimization - Reproducibility

# Reproducibility for Sequential Algorithm Portfolios This document provides an overview of the figures generated from various plotting notebooks. Each figure corresponds to specific analysis and visualization tasks. ## Figures Overview ### Figure 1 - **Description**: Comparison of EAF (Empirical Attainment Function) vs. performance. - **Notebook**: `plot_eaf_vs_performance.ipynb` ### Figure 2 - **Description**: Analysis of greedy portfolios using COCO framework. - **Notebook**: `plot_coco_greedy_portfolios.ipynb` ### Figures 3 and 4 - **Description**: Results from greedy portfolio analysis with 10,000 evaluations. - **Notebook**: `plot_coco_greedy_10000.ipynb` ### Figure 5 - **Description**: Analysis of greedy weight power. - **Notebook**: `plot_greedy_weight_power.ipynb` ### Figures 6, 7, and 8 - **Description**: Comparison of greedy vs. partial enumeration methods. - **Notebook**: `plot_greedy_vs_partial_enum.ipynb` ### Figure 9 - **Description**: Incremental portfolio analysis using COCO framework. - **Notebook**: `plot_coco_incremental_portfolios.ipynb` ### Figure 10 - **Description**: Comparison of greedy and random algorithms. - **Notebook**: `plot_coco_greedy_random_algo.ipynb` ### Figure 11 - **Description**: Influence of parameter `k` on continuous optimization. - **Notebook**: `plot_k_influence_on_continuous.ipynb` ### Figures 12, 13, and 14 - **Description**: Comparison of all methods. - **Notebook**: `appendix/plot_all_methods_compared.ipynb` To generate the plots, you have the following options: 1. **Generate Data**: - Use the generative code available in the `generate_data` folder (both in the main directory and the appendix). 2. **Unzip Pre-Generated Data**: - Extract the `data.zip` file for pre-generated data. **Algorithm Data**: - For data on algorithms run using the COCO framework, extract `coco.zip` into the `coco_data` folder. - For data on algorithms run locally, extract `local_algo_data.zip` into the `local_algo_data` folder.