Convert propensity score to sample weight for causal inference

computer_sample_weight.py

def compute_sample_weight(treatment_label, propensity_score, enforce_positivity=True, max_sample_weight=1e3):
    """
    Demystifying Double Robustness
    https://projecteuclid.org/download/pdfview_1/euclid.ss/1207580167
    https://arxiv.org/pdf/1706.10029.pdf
    weights = [ti/g(Xi) + (1−ti)/(1−g(Xi))]
    :param treatment_label: known treatment labels
    :param propensity_score: estimated propensity scores
    :param enforce_positivity: self explanatory
    :param max_sample_weight: this is to prevent inf in subsquent calculation. N
    :return:
    """
    sample_weight = (treatment_label / propensity_score) + ((1 - treatment_label) / (1 - propensity_score))
    if enforce_positivity:
        positively_control = np.isclose(propensity_score, 0)
        positively_treatment = np.isclose(propensity_score, 1)
        positivity_violations = positively_control | positively_treatment
        sample_weight[positivity_violations] = 0  # This means the model ignores this value
    return np.minimum(sample_weight, max_sample_weight)