Skip to contents

Applies a transparent reporting rule to the policy-brief table returned by `margot_policy_workflow()`. The default rule graphs a policy tree only when the selected policy value and the uplift among treated units both have lower confidence limits above the specified thresholds.

Usage

margot_select_grf_policy_trees(
  policy_brief,
  model_names = NULL,
  outcome_labels = NULL,
  policy_value_lower_threshold = 0,
  treated_uplift_lower_threshold = 0,
  require_policy_value = TRUE,
  require_treated_uplift = TRUE
)

Arguments

policy_brief

Data frame with `Outcome`, `Policy Value (95 `Uplift in Treated (95 `policy_workflow$policy_brief_df`.

model_names

Optional character vector of model names. If supplied with `outcome_labels`, the returned table includes model names aligned by outcome label.

outcome_labels

Optional character vector of outcome labels aligned with `model_names`.

policy_value_lower_threshold

Numeric threshold for the policy-value lower confidence limit. Default is 0.

treated_uplift_lower_threshold

Numeric threshold for the treated-uplift lower confidence limit. Default is 0.

require_policy_value

Logical. Require policy-value lower confidence limit to exceed `policy_value_lower_threshold`. Default TRUE.

require_treated_uplift

Logical. Require treated-uplift lower confidence limit to exceed `treated_uplift_lower_threshold`. Default TRUE.

Value

A data frame containing parsed estimates, lower confidence limits, and `graph_policy_tree`.