Top-K Accuracy

Problem

Compute top-k accuracy: the fraction of samples whose true label appears among the k classes with the highest predicted probability.

Ties in probability are ordered by the smallest class index. If k is at least the number of classes, every sample counts as correct. Empty input returns 0.0.

Hint

Sort classes by (-probability, index) and check the first k entries.

Test cases

case 1

in: [[0,1,2],[[0.7,0.2,0.1],[0.1,0.6,0.3],[0.2,0.3,0.5]],1]

expected: 1

case 2

in: [[2,0],[[0.5,0.4,0.1],[0.6,0.3,0.1]],2]

expected: 0.5

case 3

in: [[2,0],[[0.5,0.4,0.1],[0.6,0.3,0.1]],3]

expected: 1

case 4

in: [[1],[[0.5,0.5,0]],1]

expected: 0

Starter codePython

def top_k_accuracy(y_true, y_pred_probs, k):
    # Your code here
    pass

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