import numpy as np

import seaborn as sns

import matplotlib.pyplot as plt

​

def make_confusion_matrix(cf, group_names=None, categories='auto', count=True,

                          percent=True, cbar=True, xyticks=True, xyplotlabels=True,

                          sum_stats=True, figsize=None, cmap='Blues', title=None):

    """

    Generates a labeled confusion matrix using Seaborn's heatmap.

    Args:

    - cf (array-like): 2D array representing the confusion matrix.

    - group_names (list): List of labels to use for annotating the matrix.

    - categories (list or 'auto'): List of categories for axis labels.

    - count (bool): Whether to display counts in matrix boxes.

    - percent (bool): Whether to display percentage in matrix boxes.

    - cbar (bool): Whether to display a colorbar.

    - xyticks (bool): Whether to display x and y ticks.

    - xyplotlabels (bool): Whether to display x and y labels.

    - sum_stats (bool): Whether to display summary statistics.

    - figsize (tuple): Size of the figure.

    - cmap (str): Colormap to be used.

    - title (str): Title of the heatmap.

    Returns:

    - None: Displays a heatmap.

    """

    # Ensure matrix is square

    if cf.shape[0] != cf.shape[1]:

        raise ValueError("Confusion matrix should be a square matrix.")

    # Ensure group_names match matrix size if provided

    if group_names and len(group_names) != cf.size:

        raise ValueError("Length of group_names should match the size of the confusion matrix.")

    # Annotations for each box

    blanks = ['' for _ in range(cf.size)]

    group_labels = ["{}\n".format(value) for value in group_names] if group_names else blanks

    group_counts = ["{0:0.0f}\n".format(value) for value in cf.flatten()] if count else blanks

    group_percentages = ["{0:.2%}".format(value) for value in cf.flatten() / np.sum(cf)] if percent else blanks

    box_labels = [f"{v1}{v2}{v3}".strip() for v1, v2, v3 in zip(group_labels, group_counts, group_percentages)]

    box_labels = np.asarray(box_labels).reshape(cf.shape[0], cf.shape[1])

​

    # Summary statistics

    stats_text = ""

    if sum_stats:

        accuracy = np.trace(cf) / float(np.sum(cf))

        if len(cf) == 2:  # Binary classification metrics

            precision = cf[1, 1] / sum(cf[:, 1])

            recall = cf[1, 1] / sum(cf[1, :])

            f1_score = 2 * precision * recall / (precision + recall)

            stats_text = f"\n\nAccuracy={accuracy:.3f}\nPrecision={precision:.3f}\nRecall={recall:.3f}\nF1 Score={f1_score:.3f}"

        else:

            stats_text = f"\n\nAccuracy={accuracy:.3f}"

​

    # Visualization

    if not figsize:

        figsize = plt.rcParams.get('figure.figsize')

​

    plt.figure(figsize=figsize)

    sns.heatmap(cf, annot=box_labels, fmt="", cmap=cmap, cbar=cbar, xticklabels=categories, yticklabels=categories)

    if xyplotlabels:

        plt.ylabel('True label')

        plt.xlabel('Predicted label' + stats_text)

    else:

        plt.xlabel(stats_text)

    if title:

        plt.title(title)

    plt.show()