Weighted F1 Score

Problem

Compute the support-weighted F1 score over the classes present in y_true.

For each class compute F1 = 2*p*r/(p+r) using 0.0 when a denominator is zero, weight it by support_c / n, and sum the contributions. Empty input returns 0.0.

Hint

Weight each class F1 by its share of the true labels rather than averaging equally.

Test cases

case 1

in: [[0,0,0,1],[0,0,1,1]]

expected: 0.7666666666666667

case 2

in: [[1,1,0,0],[1,1,0,0]]

expected: 1

case 3

in: [[0,1,2,0,1,2],[0,1,2,0,1,2]]

expected: 1

case 4

in: [[],[]]

expected: 0

Starter codePython

def weighted_f1(y_true, y_pred):
    # Your code here
    pass

4 test cases · run your code against them in the editor