Estimate mean rewards \(\mu\) for each treatment \(a\)

\(\mu_a = m(x) + (1-e_a(x))\tau_a(x)\)

# S3 method for causal_forest
conditional_means(object, ...)

# S3 method for causal_survival_forest
conditional_means(object, ...)

# S3 method for instrumental_forest
conditional_means(object, ...)

# S3 method for multi_arm_causal_forest
conditional_means(object, outcome = 1, ...)

conditional_means(object, ...)

Arguments

object	An appropriate causal forest type object
...	Additional arguments
outcome	Only used with multi arm causal forets. In the event the forest is trained with multiple outcomes Y, a column number/name specifying the outcome of interest. Default is 1.

Value

A matrix of estimated mean rewards

Methods (by class)

• conditional_means(causal_forest): Mean rewards \(\mu\) for control/treated

• conditional_means(causal_survival_forest): Mean rewards \(\mu\) for control/treated

• conditional_means(instrumental_forest): Mean rewards \(\mu\) for control/treated

• conditional_means(multi_arm_causal_forest): Mean rewards \(\mu\) for each treatment \(a\)

Examples

# \donttest{
# Compute conditional means for a multi-arm causal forest
n <- 500
p <- 10
X <- matrix(rnorm(n * p), n, p)
W <- as.factor(sample(c("A", "B", "C"), n, replace = TRUE))
Y <- X[, 1] + X[, 2] * (W == "B") + X[, 3] * (W == "C") + runif(n)
forest <- grf::multi_arm_causal_forest(X, Y, W)
mu.hats <- conditional_means(forest)
head(mu.hats)
#>               A          B          C
#> [1,] -0.7427991 -1.4902451 -1.0782366
#> [2,]  0.6192057  1.3040476  0.3668783
#> [3,] -1.4264362 -1.0146372 -0.7595196
#> [4,]  0.4765792  0.7675958 -0.3966188
#> [5,]  1.3470327  0.5297886  0.5861858
#> [6,]  1.7781017  1.8904472  1.4167694

# Compute conditional means for a causal forest
n <- 500
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 <- grf::causal_forest(X, Y, W)
mu.hats <- conditional_means(c.forest)
# }