plot_service.py

import matplotlib.pyplot as plt
import seaborn


def plot_intra_scores(scores, name):
    plt.figure(figsize=(10, 6))
    seaborn.histplot(scores, bins=30, kde=True, color='skyblue', alpha=0.7)
    plt.title(f'Distribution of {name}')
    plt.xlabel(f'{name}_scores')
    plt.ylabel('Frequency')
    plt.savefig(f"{name} Scores.png")
    plt.show()


def compare_similarity_distributions(tfidf_scores, ann_scores_normalized):
    plt.figure(figsize=(10, 6))
    seaborn.histplot(tfidf_scores, label="Cosine Similarity (TF-IDF)", kde=True, color="blue", alpha=0.5)
    seaborn.histplot(ann_scores_normalized, label="ANN Similarity", kde=True, color="red", alpha=0.5)

    plt.title("Comparison of Cosine Similarity (TF-IDF) and ANN Scores")
    plt.xlabel("Similarity Score")
    plt.ylabel("Frequency")
    plt.legend()
    plt.show()


def plot_distribution(all_recommendations):
    tfidf_scores = [rec["avg_tfidf_score"] for rec in all_recommendations]
    sbert_scores = [rec["avg_sbert_score"] for rec in all_recommendations]
    ann_scores = [rec["avg_ann_score_normalized"] for rec in all_recommendations]

    score_data = {
        "TF-IDF": tfidf_scores,
        "SBERT": sbert_scores,
        "ANN": ann_scores,
    }

    plot_data = [(key, value) for key, values in score_data.items() for value in values]

    # Boxplot
    plt.figure(figsize=(12, 6))
    seaborn.histplot(tfidf_scores, label='TF-IDF CS', kde=True, bins=20, color='blue', alpha=0.5)
    seaborn.histplot(sbert_scores, label='SBERT', kde=True, bins=20, color='orange', alpha=0.5)
    seaborn.histplot(ann_scores, label='TF-IDF ANN', kde=True, bins=20, color='red', alpha=0.5)

    plt.title('Distribution of Average Scores for Each Recommender')
    plt.xlabel('Average Score')
    plt.ylabel('Frequency')
    plt.legend()
    plt.savefig("distribution_of_scores_boxplot.png")
    plt.show()