confint.ate: Confidence intervals for causal forest ATEs

In ensley-nexant/cfeval: Causal Forest Evaluation and Visualization

Description

Computes confidence intervals for the average treatment effect estimated by a causal_forest model.

Usage

## S3 method for class 'ate'
confint(object, parm, level = 0.9, ...)

Arguments

object	An ate object
parm	Not used
level	The desired confidence level
...	Additional arguments (ignored)

Value

A one-row tibble with two columns giving lower and upper confidence bounds for the estimated ATE. The column names will be (1-level)/2% and 1 - (1-level)/2% (by default, 5.0% and 95.0%).

Examples

## Not run: 
 require(grf)

 n <- 2000; p <- 10

 X <- matrix(rnorm(n * p), n, p)
 W <- rbinom(n, 1, 0.4 + 0.2 * (X[, 1] > 0))
 Y <- pmax(X[, 1], 0) * W + X[, 2] + pmin(X[, 3], 0) + rnorm(n)

 cf <- causal_forest(X, Y, W)
 a <- ate(cf)
 confint(a)

## End(Not run)

ensley-nexant/cfeval documentation built on May 20, 2020, 12:34 a.m.