Interaction H-Statistic

Problem

Compute the Friedman H-statistic for a feature pair: sqrt(sum((pdp_jk - pdp_j - pdp_k)^2) / sum(pdp_jk^2)).

Return 0.0 when the denominator is 0.

Hint

The H-statistic measures how much the joint partial dependence deviates from additivity.

case 1

in: [[[2,3],[3,4]],[1,2],[1,2]]

expected: 0

case 2

in: [[[2,3],[4,5]],[1,2],[1,2]]

expected: 0.19245008972987526

case 3

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

expected: 0

Starter codePython

def interaction_h_statistic(pdp_jk, pdp_j, pdp_k):
    # Your code here
    pass

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