Majority Vote Accuracy

Problem

For each item, pick the most frequent answer in its sample list (ties broken by the lexicographically smallest label) and compare it with the gold answer. Return the fraction of items where the majority answer is correct. Return 0.0 for an empty sample list, and count empty per-item lists as incorrect.

Test cases

case 1

in: [[["a","a","b"],["b","b","a"]],["a","b"]]

expected: 1

case 2

in: [[[],["a"]],["a","a"]]

expected: 0.5

case 3

in: [[["x","y"]],["z"]]

expected: 0

case 4

in: [[],[]]

expected: 0

Starter codePython

def majority_vote_accuracy(samples, gold):
    # Your code here
    pass

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