PSI Computation

Problem

Return the population stability index between two aligned lists of bin proportions: sum((actual - expected) * ln(actual / expected)). Proportions at or below 1e-6 are smoothed to 1e-6, and bins where both are non-positive are skipped. Return 0.0 when the inputs are empty or their lengths differ.

Test cases

case 1

in: [[0.5,0.5],[0.6,0.4]]

expected: 0.04054651081081642

case 2

in: [[0.25,0.25,0.5],[0.2,0.3,0.5]]

expected: 0.02027325540540821

case 3

in: [[],[]]

expected: 0

case 4

in: [[0.5],[0.5]]

expected: 0

Starter codePython

import math
def psi(expected, actual):
    # Your code here
    pass

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