Effective Sample Size for Importance Weights

Problem

Compute the effective sample size of a weight vector: (sum(w))^2 / sum(w^2).

Return 0.0 when the sum of squares is 0 or the input is empty.

Hint

Unequal weights reduce the effective sample size below the raw count.

case 1

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

expected: 4

case 2

in: [[1,2,3]]

expected: 2.5714285714285716

case 3

in: [[0,0]]

expected: 0

case 4

in: [[]]

expected: 0

Starter codePython

def effective_sample_size(weights):
    # Your code here
    pass

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