DS5100 Final Project

Metadata

• Author: Hang Yu, zdd3ga
• Course: DS5100 Programming for Data Science
• Date: May 2025
• License: MIT

Synopsis

The DS5100 Final Project provides three modular Python classes for modeling, playing, and analyzing weighted dice and related games:

1. Die Create a custom die with any set of faces and weights.

import numpy as np
from die import Die

faces = np.array(['A', 'B', 'C'])
d = Die(faces)
d.change_weight('A', 2.0)       # make 'A' twice as likely
rolls = d.roll(5)               # e.g. ['A','C','B','A','A']
print(d.show())                 # View current faces & weights in a data frame

2. Game Simulate repeated rolls of one or more dice.

from game import Game

dice = [Die(faces) for i in range(3)]
game = Game(dice)
game.play(100)                          # roll 3 dice 100 times
wide_df   = game.show_results()         # wide format
narrow_df = game.show_results('narrow') # Narrow(stacked) format

3. Analyzer Analyze statistical patterns from a completed game.

from analyzer import Analyzer

analyzer = Analyzer(game)
print(analyzer.jackpot())               # count of all?same rolls
print(analyzer.face_counts_per_roll())  # DataFrame: roll face counts
print(analyzer.combo_count())           # combos & their counts
print(analyzer.permutation_count())     # permutations & their counts

API

Die

• init(self, faces: np.ndarray)
• change_weight(self, face: Any, weight: float) -> None
• roll(self, n: int = 1) -> List[Any]
• show(self) -> pd.DataFrame

Game

• init(self, dice: List[Die])
• play(self, n: int) -> None
• show_results(self, form: str = 'wide') -> pd.DataFrame

Analyzer

• init(self, game: Game)
• jackpot(self) -> int
• face_counts_per_roll(self) -> pd.DataFrame
• combo_count(self) -> pd.DataFrame
• permutation_count(self) -> pd.DataFrame

Installation

# for development
pip install -e .