infer_policy: Conduct inference for a personalized treatment policy.

In bonander/CEAforests: Tree-based heterogenous cost-effectiveness and policy learning with individual-level data (beta)

Description

Conduct inference for a personalized treatment policy, either using a manually specified policy or a learned policy.

Usage

infer_policy(
  forest,
  treat.policy,
  WTP = NULL,
  ci.level = 0.95,
  robust.se = FALSE,
  boot.ci = FALSE,
  R = 999
)

Arguments

forest	A trained CEA forest.
treat.policy	A logical vector or cea policy tree defining the subset covered by the policy.
WTP	Willingness to pay for a one unit increase in the outcome. If NULL, the WTP supplied to the CEA forest is used.
ci.level	Desired significance level (for confidence intervals).
robust.se	Whether or not robust (sandwich) standard errors are desired. Defaults to FALSE. Ignored when boot.ci=TRUE.
boot.ci	Whether or not bootstrapped confidence intervals are desired. Defaults to FALSE.
R	The number of bootstrap replications. Defaults to 999. Ignored when boot.ci=FALSE.

Value

Returns a matrix containing estimates for the average welfare gain per population member under various treatment policies (treat everyone vs. treat no one; treat suggested subset vs. treat no one; treat suggested subset vs. treat everyone). Also outputs the share of the popuation covered by the policy.

Examples

## Not run: 
To be added...

## End(Not run)

bonander/CEAforests documentation built on April 1, 2021, 10:57 a.m.