在Python中是否已经实现了一些东西，可以为多类混淆矩阵计算TP，TN，FP和FN？

Question

Sklearn.metrics has great functions for obtaining classification metrics, although something that I think is missing is a function to return the TP, FN, FP and FN counts given the predicted and actual label sequences. Or even from the confusion matrix.

I know it's possible to obtain the confusion matrix using sklearn , but I need the actual TP, FN, FP and FN counts (for multilabel classification - more than 2 labels), and to obtain those counts for each of the classes.

So say, I have the confusion matrix below with 3 classes. Is there some package available to get the counts for each class from this? I was unable to find anything.

Answer 1

Scikit-learn can calculate and plot a multiclass confusion matrix, see this example from the documentation ( Demo on a Jupiter notebook ):

import itertools
import numpy as np
import matplotlib.pyplot as plt

from sklearn import svm, datasets
from sklearn.model_selection import train_test_split
from sklearn.metrics import confusion_matrix

# import some data to play with
iris = datasets.load_iris()
X = iris.data
y = iris.target
class_names = iris.target_names

# Split the data into a training set and a test set
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)

# Run classifier, using a model that is too regularized (C too low) to see
# the impact on the results
classifier = svm.SVC(kernel='linear', C=0.01)
y_pred = classifier.fit(X_train, y_train).predict(X_test)


def plot_confusion_matrix(cm, classes,
                          normalize=False,
                          title='Confusion matrix',
                          cmap=plt.cm.Blues):
    """
    This function prints and plots the confusion matrix.
    Normalization can be applied by setting `normalize=True`.
    """
    plt.imshow(cm, interpolation='nearest', cmap=cmap)
    plt.title(title)
    plt.colorbar()
    tick_marks = np.arange(len(classes))
    plt.xticks(tick_marks, classes, rotation=45)
    plt.yticks(tick_marks, classes)

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

    print(cm)

    thresh = cm.max() / 2.
    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):
        plt.text(j, i, cm[i, j],
                 horizontalalignment="center",
                 color="white" if cm[i, j] > thresh else "black")

    plt.tight_layout()
    plt.ylabel('True label')
    plt.xlabel('Predicted label')

# Compute confusion matrix
cnf_matrix = confusion_matrix(y_test, y_pred)
np.set_printoptions(precision=2)

# Plot non-normalized confusion matrix
plt.figure()
plot_confusion_matrix(cnf_matrix, classes=class_names,
                      title='Confusion matrix, without normalization')

# Plot normalized confusion matrix
plt.figure()
plot_confusion_matrix(cnf_matrix, classes=class_names, normalize=True,
                      title='Normalized confusion matrix')

plt.show()

Result (txt):

Confusion matrix, without normalization
[[13  0  0]
 [ 0 10  6]
 [ 0  0  9]]

Normalized confusion matrix
[[ 1.    0.    0.  ]
 [ 0.    0.62  0.38]
 [ 0.    0.    1.  ]]

Plot results:

---

See this code working on the link bellow:
DEMO ON A JUPYTER NOTEBOOK

Answer 2

I ended up implementing it myself, since I didn't find anything. Here is the code, case someone else looks for this in the future:

def counts_from_confusion(confusion):
    """
    Obtain TP, FN FP, and TN for each class in the confusion matrix
    """

    counts_list = []

    # Iterate through classes and store the counts
    for i in range(confusion.shape[0]):
        tp = confusion[i, i]

        fn_mask = np.zeros(confusion.shape)
        fn_mask[i, :] = 1
        fn_mask[i, i] = 0
        fn = np.sum(np.multiply(confusion, fn_mask))

        fp_mask = np.zeros(confusion.shape)
        fp_mask[:, i] = 1
        fp_mask[i, i] = 0
        fp = np.sum(np.multiply(confusion, fp_mask))

        tn_mask = 1 - (fn_mask + fp_mask)
        tn_mask[i, i] = 0
        tn = np.sum(np.multiply(confusion, tn_mask))

        counts_list.append({'Class': i,
                            'TP': tp,
                            'FN': fn,
                            'FP': fp,
                            'TN': tn})

    return counts_list