ate.randomForest: Average treatment effect estimation for randomized...

Description Usage Arguments Value Author(s) References

Description

Use random forest to adjust covariates and estimate average treatment effects and confidence intervals

Usage

1
ate.randomForest(X, Y, W, nodesize = 20, conf.level = 0.9)

Arguments

X

Data matrix with covariates, one observation per row

Y

Outcome vector for assigned treatment for each observation

W

Treatment vector for each observation

nodesize

Node size for random forest. Default 20

conf.level

Confidence level for intervals. Default 0.9

Value

Returns list containing the following

tau

Average treatment effect estimate

var

Variance estimate

conf.int

Confidence interval for true tau

conf.level

Confidence level

Author(s)

Stefan Wager, Wenfei Du, Jonathan Taylor, Rob Tibshirani

References

S Wager, W Du, J Taylor, and R Tibshirani. "High-dimensional regression adjustments in randomized experiments". PNAS, November 8, 2016 vol. 113 no. 45 12673-12678


swager/crossEstimation documentation built on May 30, 2019, 9:33 p.m.