generate_figure_4.py

import numpy as np
from scipy.stats import kendalltau
import matplotlib.pyplot as plt

# Adding the ablation for with and without priority

def bootstrap_kendall_tau(reference, prediction, n=1000, confidence_level=0.90):
    tau_values = []
    n_samples = len(reference)
    for _ in range(n):
        indices = np.random.randint(0, n_samples, n_samples)
        sample_ref = np.array(reference)[indices]
        sample_pred = np.array(prediction)[indices]
        tau, _ = kendalltau(sample_ref, sample_pred)
        tau_values.append(tau)

    sorted_tau_values = np.sort(tau_values)
    cdf = np.arange(1, len(sorted_tau_values) + 1) / len(sorted_tau_values)
    return sorted_tau_values, cdf

if __name__ == "__main__":

    # Define vectors
    GT = [1, 2, 3, 4, 2, 1, 3, 4, 2, 3, 4, 1, 4, 1, 3, 2, 3, 2, 1, 4, 4, 1, 3, 2, 1, 3, 2, 4, 2, 3, 1, 4, 1, 3, 2, 4, 4,
          1, 3, 2, 3, 1, 4, 2, 1, 3, 2, 4, 3, 4, 2, 1]
    autodoserank_with_priority = [1, 2, 3, 4, 2, 3, 1, 4, 1, 4, 2, 3, 4, 3, 1, 2, 3, 4, 2, 1, 3, 4, 1, 2, 1, 4, 2, 3, 4,
                                  2, 3, 1, 2, 1, 3, 4, 1, 3, 2, 4, 2, 1, 4, 3, 3, 4, 1, 2, 1, 4, 2, 3]
    autodoserank_without_priority = [1, 2, 3, 4, 1, 2, 3, 4, 1, 4, 2, 3, 3, 2, 1, 4, 3, 4, 2, 1, 4, 3, 1, 2, 1, 3, 2, 4,
                                     4, 3, 2, 1, 1, 2, 3, 4, 1, 4, 2, 3, 4, 1, 3, 2, 4, 3, 1, 2, 1, 3, 2, 4]
    # Perform bootstrapping analysis
    sorted_with_priority, cdf_with_priority = bootstrap_kendall_tau(GT, autodoserank_with_priority)
    sorted_without_priority, cdf_without_priority = bootstrap_kendall_tau(GT, autodoserank_without_priority)

    plt.figure(figsize=(10, 6))
    plt.plot(sorted_without_priority, cdf_without_priority, label='GT vs. AutoDoseRank', color='orange', linestyle='-')
    plt.plot(sorted_with_priority, cdf_with_priority, label='GT vs. AutoDoseRank-no-Priority', color='orange', linestyle=':')

    plt.xlabel('Kendall’s Tau')
    plt.ylabel('Cumulative Distribution')
    plt.title('CDF of Kendall’s Tau Values for Bootstrapping Comparisons')
    plt.legend()
    plt.grid(True)
    plt.tight_layout()
    plt.show()