Cluster Purity

Problem

Compute cluster purity: for each cluster count how many points belong to each true class, add the largest count, and divide by n.

Empty input returns 0.0.

Hint

Each cluster votes with its most frequent true class; sum those votes and normalize.

case 1

in: [[0,0,1,1],[0,0,0,1]]

expected: 0.75

case 2

in: [[0,0,0,1,1,1],[0,1,0,1,1,2]]

expected: 0.6666666666666666

case 3

in: [[7,7,7],[1,2,3]]

expected: 0.3333333333333333

case 4

in: [[],[]]

expected: 0

Starter codePython

def cluster_purity(cluster_ids, true_labels):
    # Your code here
    pass

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