Data generation scripts and libraries

This directory contains scripts and libraries for the generation of data.

Players used in this work

$ python players.py  # Lists the number of players for which data was generated
175

Also contains a function generate_players which returns a list of axelrod player instances.

Abbreviations

The abbreviations.py file contains a dictionary with some abbreviations for player names.

Theoretic results

The file theoretic.py contains a number of functions used for the calculation of analytic results for Moran processes.

Generate the cache.

$ python generate_cache.py

This generates two data files in ../data:

• outcomes.csv
• outcomes_noise.csv

These files are of the format:

player1_name, player2_name, player1_score, player2_score, count

where count is the number of times that particular score pair occurs.

Also contains a function read_csv which reads in the file to give nested dictionaries of match outcomes.

Run the Moran processes

The file moran.py is used to generate data files for the Moran process.

$ python moran.py 4 2 ../data/outcomes.csv ../data/sims_n_over_2/ sims_4.csv

This will run the Moran process for all pairs of players in a population of size 4 and 2 players of the first type. A cached outcome of match results if read from ../data/outcomes.csv and the output is ..data/sims_4.csv.

Preprocessing of the raw data

The file clean_raw_moran.py is used to clean all the data generated from moran.py. Creates one data file ..data/sims_summary.csv of the form:

Noise, P1, P2, N, repetitions, P1_fixation, P2_fixation

The file preproces.py is used to write the fixation probabilities and relative fitness for each strategy pair to ..data/main.csv:

player, opponent, N, noise, p_1, p_{N/2}, p_{N-1}

Where:

• p_1: is relative fitness of 1 player with N - 1 opponents
• p_{N/2}: is relative fitness of N/2 players with N/2 opponents
• p_{N-1}: is relative fitness of N-1 players with 1 opponent.

This is automatically re written when running clean_raw_moran.py.

main.csv is the main file used for all the analysis.