Sample Size per Variant

Problem

Compute the required sample size per variant for a two-proportion test: ceil((z_alpha + z_beta)^2 * (p1*(1-p1) + p2*(1-p2)) / (p2 - p1)^2).

Return 0 when p1 == p2.

Hint

Smaller effects and lower variances both change required sample size dramatically.

case 1

in: [0.1,0.12,1.96,0.84]

expected: 3834

case 2

in: [0.5,0.55,1.96,0.84]

expected: 1561

case 3

in: [0.2,0.1,1.645,0.84]

expected: 155

case 4

in: [0.1,0.1,1.96,0.84]

expected: 0

Starter codePython

import math
def sample_size_per_variant(p1, p2, z_alpha, z_beta):
    # Your code here
    pass

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