Rank-Biserial Correlation

Problem

Return the rank-biserial correlation 2*(mean_rank_x - mean_rank_y)/(n1+n2), using average ranks for ties in the combined sample. Return 0.0 if either sample is empty.

Hint

It converts a rank comparison into a correlation-scale effect size.

Test cases

case 1

in: [[1,2,3],[4,5,6]]

expected: -1

case 2

in: [[4,5,6],[1,2,3]]

expected: 1

case 3

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

expected: 0

case 4

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

expected: -0.3333333333333333

case 5

in: [[5],[5]]

expected: 0

Starter codePython

def rank_biserial(x, y):
    # Your code here
    pass

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