grf (version 0.9.5)

causal_forest: Causal forest

Description

Trains a causal forest that can be used to estimate conditional average treatment effects tau(X). When the treatment assignmnet W is binary and unconfounded, we have tau(X) = E[Y(1) - Y(0) | X = x], where Y(0) and Y(1) are potential outcomes corresponding to the two possible treatment states. When W is continuous, we effectively estimate an average partical effect Cov[Y, W | X = x] / Var[W | X = x], and interpret it as a treatment effect given unconfoundedness.

Usage

causal_forest(X, Y, W, sample.fraction = 0.5, mtry = NULL,
  num.trees = 2000, num.threads = NULL, min.node.size = NULL,
  honesty = TRUE, ci.group.size = 2, precompute.nuisance = TRUE,
  alpha = 0.05, lambda = 0, downweight.penalty = FALSE, seed = NULL)

Arguments

X

The covariates used in the causal regression.

Y

The outcome.

W

The treatment assignment (may be binary or real).

sample.fraction

Fraction of the data used to build each tree. Note: If honesty is used, these subsamples will further be cut in half.

mtry

Number of variables tried for each split.

num.trees

Number of trees grown in the forest. Note: Getting accurate confidence intervals generally requires more trees than getting accurate predictions.

num.threads

Number of threads used in training. If set to NULL, the software automatically selects an appropriate amount.

min.node.size

A target for the minimum number of observations in each tree leaf. Note that nodes with size smaller than min.node.size can occur, as in the original randomForest package.

honesty

Whether or not honest splitting (i.e., sub-sample splitting) should be used.

ci.group.size

The forest will grow ci.group.size trees on each subsample. In order to provide confidence intervals, ci.group.size must be at least 2.

precompute.nuisance

Should we first run regression forests to estimate y(x) = E[Y|X=x] and w(x) = E[W|X=x], and then run a causal forest on the residuals? This approach is recommended, computational resources permitting.

alpha

Maximum imbalance of a split.

lambda

A tuning parameter to control the amount of split regularization (experimental).

downweight.penalty

Whether or not the regularization penalty should be downweighted (experimental).

seed

The seed of the c++ random number generator.

Value

A trained causal forest object.

Examples

# NOT RUN {
# Train a causal forest.
n = 50; p = 10
X = matrix(rnorm(n*p), n, p)
W = rbinom(n, 1, 0.5)
Y = pmax(X[,1], 0) * W + X[,2] + pmin(X[,3], 0) + rnorm(n)
c.forest = causal_forest(X, Y, W)

# Predict using the forest.
X.test = matrix(0, 101, p)
X.test[,1] = seq(-2, 2, length.out = 101)
c.pred = predict(c.forest, X.test)

# Predict on out-of-bag training samples.
c.pred = predict(c.forest)

# Predict with confidence intervals; growing more trees is now recommended.
c.forest = causal_forest(X, Y, W, num.trees = 4000)
c.pred = predict(c.forest, X.test, estimate.variance = TRUE)
# }
# NOT RUN {
# }