Run a causalTree-style exploratory analysis

Usage

fit_causal_tree_explorer(
  data,
  outcome = "outcome",
  treatment = "treatment",
  covariates,
  sample_id = NULL,
  treatment_binary = NULL,
  treatment_cut = c("median", "mean", "zero"),
  split_rule = "CT",
  honest = TRUE,
  minsize = 20,
  xval = 5,
  prune = TRUE
)

Arguments

data: A data.frame.
outcome: Name of the outcome column.
treatment: Name of the treatment column.
covariates: Character vector of covariate column names.
sample_id: Optional sample identifier column.
treatment_binary: Optional binary treatment column. If NULL, a binary treatment is created from treatment_cut.
treatment_cut: One of "median", "mean", or "zero" when a binary treatment must be derived from a continuous treatment.
split_rule: Split rule passed to htetree::causalTree().
honest: Whether to request honest splitting.
minsize: Minimum treated and control size per leaf.
xval: Number of cross-validation folds.
prune: Whether to prune to the cross-validated optimal cp.

Value

A list containing the causal_tree, tree_table, and the analysis data used by the explorer.