Project 4: Community

Citation and License

This project belongs to Department of Computer Science, Columbia University. It may be used for educational purposes under Creative Commons with proper attribution and citation for the author TAs Raavi Gupta (First Author), Divyang Mittal and the Instructor, Prof. Kenneth Ross.

Summary

Course: COMS 4444 Programming and Problem Solving (Fall 2024)
Problem Description: https://www.cs.columbia.edu/~kar/4444f24/node21.html
Course Website: https://www.cs.columbia.edu/~kar/4444f24/
University: Columbia University
Instructor: Prof. Kenneth Ross
Project Language: Python

TA Designer for this project

Raavi Gupta

Teaching Assistants for Course

1. Divyang Mittal
2. Raavi Gupta

Installation

To install tkinter on macOS, run the following command:

brew install [email protected]

For Windows, tkinter can be installed using pip:

pip install tk

Usage

To view all options, use python community.py -h.

python community.py  [--seed] [--num_members] [--num_turns] [--num_abilities] [--seed_ability] [--seed_task_difficulty] [--group_task_distribution] [--group_abilities_distribution] [--log] [--gui] [--gi]

Description of Flags:

• seed: The seed for the entire game.
• num_members: Number of members in the community.
• num_turns: Total number of turns.
• num_abilities: Number of abilities present (between 3-8).
• seed_ability: Separate seed for ability generation (optional).
• seed_task_difficulty: Separate seed for task difficulty generation (optional).
• group_task_distribution: Group number whose task difficulty distribution function will be used.
• task_distribution_difficulty: Difficulty level of tasks. One of easy, medium or hard.
• group_abilities_distribution: Group number whose ability distribution function will be used.
• abilities_distribution_difficulty: Difficulty level for the abilities. One of easy, medium or hard.
• log: If this flag is present, print statements won't appear on the console and will be saved to the log-results/ folder.
• gui: Activates the simulator with a graphical user interface (GUI).
• gi: Specifies the number of players from each group in the community. [i] ranges from 1-10, with $\sum_{i=1}^{10} g_{i} = $num_members. If the sum is less than num_members, the remaining players will be from team 0 by default. The sum cannot exceed the specified num_members.

Example Usage

python .\community.py --num_members 40 --num_turns 100 --num_abilities 8 --g1 20 --gui --log

Submission

To submit the code for each deliverable, open a pull request to merge your proposed preferences.py and distributions.py files for player preferences and task/abilities distributions. These two Python files, named preferences.py and distributions.py, should be located in the teams/team_{group-number} folder (e.g., teams/team_0, where the group number corresponds to the allotted group number for this project). The Python files should contain the following functions:

preferences.py

1. phaseIpreferences(player, community, global_random): This function defines your strategy for choosing preferences in phase I of the game. It should return a list of lists, where each element is a [task_index, partner_id] pair. This represents your bid to form partnerships with players on tasks.

2. phaseIIpreferences(player, community, global_random): This function defines your strategy for bidding on tasks in phase II. It should return a list of task indices from community.tasks that the player chooses to bid on during phase II in a turn.

distributions_{easy/medium/hard}.py

1. ability_distribution(num_abilities, seed, player_id, global_random): This function defines the ability distribution for a player, given the number of abilities and the player ID. For a given seed, the function should output the same list of abilities across runs to ensure reproducibility. Refer to the sample in the team_0 folder for guidance.

2. task_difficulty_distribution(num_abilities, seed, task_generation_id, global_random): This function defines the task difficulty distribution, given the number of abilities and task generation ID. For a given seed, the function should output the same list of task difficulties across runs to ensure reproducibility. Refer to the sample in the team_0 folder for guidance.

A sample submission format for team 0 (a non-existent team) has been provided in the teams folder for reference.