Helper function for estimation of propensity score using sl3.

Helper function for estimation of propensity score using sl3.

Usage

estimate_pi(
  W,
  A,
  binomial_learner = Lrnr_gam$new(family = "binomial"),
  weights = NULL,
  treatment_level = max(A),
  cross_fit_and_cv = TRUE,
  return_learner = FALSE,
  folds = 10,
  ...
)

Arguments

W

Covariate matrix or data frame used to estimate the propensity score.

A

Observed treatment assignments.

binomial_learner

Base sl3 learner used for the binomial treatment regression.

weights

Optional observation weights.

treatment_level

Treatment level defining the positive class for the returned propensity estimates.

cross_fit_and_cv

Whether to wrap the nuisance learner in cross-fitting and learner selection.

return_learner

Whether to include the trained learner in the returned list.

folds

Cross-validation fold specification passed to the constructed sl3_Task.

...

Additional arguments forwarded to sl3_Task.