Build GRF-Backed CATE Learners

Usage

make_grf_cate_learners(
  method = c("ep", "r"),
  hte3_task = NULL,
  treatment_level = 1,
  control_level = 0,
  tune = c("light", "none", "all"),
  grf_params = list(),
  sieve_num_basis = NULL,
  sieve_basis_grid = NULL,
  sieve_interaction_order = 3,
  screen_basis_with_lasso = FALSE,
  ep_targeting_style = "r",
  ep_r_targeting_basis = "v_plus_propensity"
)

Arguments

method: Character vector of GRF-supported CATE methods.
hte3_task: Optional task used when expanding EP basis-size portfolios.
treatment_level: Treated level for the target contrast.
control_level: Control level for the target contrast.
tune: Tuning mode for the final GRF learner(s).
grf_params: Optional named list of GRF arguments passed to the learners.
sieve_num_basis: Optional EP basis size for single-fit EP learners.
sieve_basis_grid: Optional EP basis grid.
sieve_interaction_order: EP interaction order.
screen_basis_with_lasso: Whether EP basis screening is enabled.
ep_targeting_style: EP targeting variant used when method includes "ep".
ep_r_targeting_basis: First-stage basis construction used for EP-R.

Value

A list of learner objects.