plot.py

import numpy as np
import random as rd
import matplotlib.pyplot as plt
from tmc import TransitionMatrixCalculator as tmc
from markovDecision import MarkovDecisionSolver as mD
from validation import Validation

def make_plots():
    layout = [0, 0, 3, 0, 0, 0, 2, 0, 0, 0, 3, 0, 0, 1, 0]
    circle = False
    validation = Validation(layout, circle)
    expec, optimal_policy = mD(layout, circle).solve()

    # Plot 1: Theoretical vs Empirical Cost
    expected_costs = np.zeros(len(expec))
    for start_square in range(len(expec)):
        total_turns = 0
        for _ in range(10000):
            total_turns += validation.play_one_game(start_square)
        expected_costs[start_square] = total_turns / 10000

    squares = np.arange(len(expec))
    plt.plot(squares, expec, label="Theoretical cost")
    plt.plot(squares, expected_costs, label="Empirical cost")
    plt.xticks(np.arange(0, len(expec), step=1))
    plt.grid(True)
    plt.xlabel("Square")
    plt.ylabel("Cost")
    plt.legend()
    plt.title("Comparison between the expected cost and the actual cost")
    plt.show()

    # Plot 2: Expected number of turns for different policies
    policies = [optimal_policy, np.ones(len(expec)), np.ones(len(expec)) * 2, np.ones(len(expec)) * 3, np.random.randint(1, 4, len(expec))]
    avgn_turns = [Validation(layout, circle).empirical_results() for policy in policies]
    names = ["optimal", "safe", "normal", "risky", "random"]
    plt.bar(names, avgn_turns)
    plt.xlabel("Policy")
    plt.ylabel("Cost")
    plt.title("Expected number of turns for different policies")
    plt.show()

# Call make_plots function
if __name__ == "__main__":
    make_plots()