Pretty print a confusion matrix with seaborn

confusion_matrix_pretty_print.py

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

def print_confusion_matrix(confusion_matrix, class_names, figsize = (10,7), fontsize=14):
    """Prints a confusion matrix, as returned by sklearn.metrics.confusion_matrix, as a heatmap.
    
    Note that due to returning the created figure object, when this funciton is called in a
    notebook the figure willl be printed twice. To prevent this, either append ; to your
    function call, or modify the function by commenting out the return expression.
    
    Arguments
    ---------
    confusion_matrix: numpy.ndarray
        The numpy.ndarray object returned from a call to sklearn.metrics.confusion_matrix. 
        Similarly constructed ndarrays can also be used.
    class_names: list
        An ordered list of class names, in the order they index the given confusion matrix.
    figsize: tuple
        A 2-long tuple, the first value determining the horizontal size of the ouputted figure,
        the second determining the vertical size. Defaults to (10,7).
    fontsize: int
        Font size for axes labels. Defaults to 14.
        
    Returns
    -------
    matplotlib.figure.Figure
        The resulting confusion matrix figure
    """
    df_cm = pd.DataFrame(
        confusion_matrix, index=class_names, columns=class_names, 
    )
    fig = plt.figure(figsize=figsize)
    try:
        heatmap = sns.heatmap(df_cm, annot=True, fmt="d")
    except ValueError:
        raise ValueError("Confusion matrix values must be integers.")
    heatmap.yaxis.set_ticklabels(heatmap.yaxis.get_ticklabels(), rotation=0, ha='right', fontsize=fontsize)
    heatmap.xaxis.set_ticklabels(heatmap.xaxis.get_ticklabels(), rotation=45, ha='right', fontsize=fontsize)
    plt.ylabel('True label')
    plt.xlabel('Predicted label')
    # Note that due to returning the created figure object, when this funciton is called in a notebook
    # the figure willl be printed twice. To prevent this, either append ; to your function call, or
    # modify the function by commenting out this return expression.
    return fig

Thank you!

Hey, thanks for this! I modified it to work for normalized cm as well, you can just add this:

    if normalize:
        confusion_matrix = confusion_matrix.astype('float') / confusion_matrix.sum(axis=1)[:, np.newaxis]
        print("Normalized confusion matrix")
    else:
        print('Confusion matrix, without normalization')

and them add the format option:

    fmt = '.2f' if normalize else 'd'

and add fmt=fmt in your try-catch :)

Thanks for sharing

I can't seem to figure out why my heat map is printing twice? here is my code:

y_pred_knn_raw = knn_model.predict(X_test) print(metrics.confusion_matrix(y_true = y_test, y_pred = y_pred_knn)) #print basic confusion matrix print('Accuracy = ', metrics.accuracy_score(y_true = y_test, y_pred = y_pred_knn)) #printing accurary print_confusion_matrix(metrics.confusion_matrix(y_true = y_test, y_pred = y_pred_knn), labels) #print cm heatmap

Hoping someone could help me see what i'm doing wrong?

I can't seem to figure out why my heat map is printing twice? here is my code:

y_pred_knn_raw = knn_model.predict(X_test) print(metrics.confusion_matrix(y_true = y_test, y_pred = y_pred_knn)) #print basic confusion matrix print('Accuracy = ', metrics.accuracy_score(y_true = y_test, y_pred = y_pred_knn)) #printing accurary print_confusion_matrix(metrics.confusion_matrix(y_true = y_test, y_pred = y_pred_knn), labels) #print cm heatmap

Hoping someone could help me see what i'm doing wrong?

Printing Twice for me too!

Add a ';' after you call the function. Should only print once.

Thank you, very helpful

Prints twice for no apparent reason.

Prints twice for no apparent reason.

Just remove return fig

Thank you

Awesome

awesome, thanks.
BTW "class_names" sort alphabetically, when there are a lot of attributes it will be better to do this as @scott Boston said
https://stackoverflow.com/questions/54875846/how-to-print-labels-and-column-names-for-confusion-matrix

and the argument normalize{‘true’, ‘pred’, ‘all’}, default=None can deal with the normalization when we generate the confusion matrix @BrunoGomesCoelho
eg:
confusion_matrix_array = confusion_matrix(ture_label, predict_label, normalize='all' )

THANK YOU DEAR