AIPW Estimate

Problem

Return the augmented IPW estimate of the treated mean, (1/n)*sum(mu1_i + t_i*(y_i - mu1_i)/e_i), where mu1_i is the outcome model prediction. Return 0.0 when lengths differ, a treated unit has e_i <= 0, or the input is empty.

Hint

The model predictions are corrected by the weighted residuals.

Test cases

case 1

in: [[1,0,1,0],[10,5,8,4],[0.5,0.5,0.4,0.6],[9,6,7,5]]

expected: 7.875

case 2

in: [[1],[7],[0.7],[6]]

expected: 7.428571428571429

case 3

in: [[0,0],[1,2],[0.2,0.3],[1.5,2.5]]

expected: 2

case 4

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

expected: 0

case 5

in: [[1],[5],[0],[1]]

expected: 0

Starter codePython

def aipw_estimate(treatment, outcome, propensity, mu1):
    # Your code here
    pass

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