IPW Estimator — est_via_ipw • CausalModel

Inverse Probability Weighting estimator for ATE.

Usage

est_via_ipw(obj, ...)

# S3 method for class 'observational'
est_via_ipw(
  obj,
  propensity_model = NULL,
  propensity = NULL,
  normalize = TRUE,
  ...
)

Arguments

obj: An observational or clustered object.
...: Additional arguments passed to methods.
propensity_model: A model wrapper with fit and predict_proba. Default uses logistic regression.
propensity: Optional numeric vector of pre-estimated propensity scores.
normalize: Logical. Normalize weights (Hajek estimator). Default TRUE.

Value

A cm_result object.