control_ptl {polle}

Control arguments for Policy Tree Learning

Description

control_ptl sets the default control arguments for doubly robust policy tree learning, type = "ptl". The arguments are passed directly to policytree::policy_tree() (or policytree::hybrid_policy_tree()) if not specified otherwise.

Usage

control_ptl(
  policy_vars = NULL,
  hybrid = FALSE,
  depth = 2,
  search.depth = 2,
  split.step = 1,
  min.node.size = 1
)

Arguments

policy_vars

Character vector/string or list of character vectors/strings. Variable names used to construct the V-restricted policy tree. The names must be a subset of the history names, see get_history_names(). Not passed to policy_tree().

hybrid

If TRUE, policytree::hybrid_policy_tree() is used to fit a policy tree. Not passed to policy_tree().

depth

Integer or integer vector. The depth of the fitted policy tree for each stage.

search.depth

(only used if hybrid = TRUE) Integer or integer vector. Depth to look ahead when splitting at each stage.

split.step

Integer or integer vector. The number of possible splits to consider when performing policy tree search at each stage.

min.node.size

Integer or integer vector. The smallest terminal node size permitted at each stage.

Value

list of (default) control arguments.