Divisiveness Theoretical

In the text we refer to:

Experiment 1 as the results shown in Section 3.2 and Appendix A.1

Experiment 2 as the results shown in Section 4.1 and Appendix A.2

Experiment 3 as the results shown in Section 4.2 and Appendix A.3

File names

For each Experiment X $\in$ {1,2,3} and method Y $\in$ {IC, UM10, UM50} and number of agents N we include:

Experiment X Toy Example.ipynb to show how the data was collected

Data collection

The data we collected for building our figures in Experiments 2 and 3 can be found in experiment_X_method_Y_profile_N.csv

In case of Experiment 3, we split the files by voting rule R $\in$ {copeland, borda}. Due to computational time to calculate divisiveness, we run the divisiveness every S new agents (for this manuscript, S $\in$ {6, 10}) experiment_3_method_Y_profile_N_step_S_R.csv

For each figure in the paper and appendix we include a file showing how the figures were created: Experiment X Figure #.ipynb

Additional considerations

The files helpers.py and rules.py contain functions used throughout the notebooks.

The file run_manipulation.py sets the heuristic implemented in Experiment 3, along with generating the data. In addition, we include the script files cron_borda.sh and cron_copeland.sh to initialize the file run_manipulation.py.

The folder data/ also includes files generated in run_manipulation.py for Experiment 3. Those files were concatenated by using File_Creation.ipynb in deprecated/ folder.