Feature Importance Rank Stability

Problem

Measure feature importance rank stability with Spearman's correlation. Convert both importance vectors to average ranks so ties share the mean rank, then return the Pearson correlation of the two rank vectors using population statistics. Return 0.0 for empty input, unequal lengths, or a constant vector.

Hint

Rank correlation tracks whether two explanations agree on ordering, even if scores are not comparable.

Test cases

case 1

in: [[0.1,0.5,0.3],[0.2,0.9,0.4]]

expected: 1

case 2

in: [[1,2,3],[3,2,1]]

expected: -1

case 3

in: [[1,1,2],[3,1,2]]

expected: 0

case 4

in: [[5],[5]]

expected: 0

case 5

in: [[],[]]

expected: 0

Starter codePython

def importance_rank_stability(importance_a, importance_b):
    # Your code here
    pass

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