transport_ittate {transportr} | R Documentation |
Transported Intent-to-Treat Average Treatment Effect
Description
Implements a TMLE for the transported intent-to-treat average treatment effect. Nuisance parameters are estimated using the Super Learner algorithm.
Usage
transport_ittate(
data,
instrument,
trt,
outcome,
covar,
pop,
obs = NULL,
id = NULL,
weights = NULL,
learners_instrument = "glm",
learners_trt = "glm",
learners_pop = "glm",
learners_outcome = "glm",
folds = 1,
control = transport_control()
)
Arguments
data |
[ |
instrument |
[ |
trt |
[ |
outcome |
[ |
covar |
[ |
pop |
[ |
obs |
[ |
id |
[ |
weights |
[ |
learners_instrument |
[ |
learners_trt |
[ |
learners_pop |
[ |
learners_outcome |
[ |
folds |
[ |
control |
[ |
Value
An object of class transported_ittate
containing the parameter estimate.
Examples
gendata <- function(n, A = NULL, S = NULL) {
if (is.null(S)) S <- rbinom(n, 1, 0.5)
W1 <- rbinom(n, 1, 0.4 + (0.2 * S))
W2 <- rnorm(n, 0.1 * S, 1)
W3 <- rnorm(n, 1 + (0.2 * S), 1)
if (is.null(A)) A <- rbinom(n, 1, 0.5)
Z <- rbinom(n, 1, plogis(-log(1.6) + log(4)*A - log(1.1)*W2 - log(1.3)*W3))
Yi <- rbinom(n, 1, plogis(log(1.6) + log(1.9)*Z - log(1.3)*W3 - log(1.2)*W1 + log(1.2)*W1*Z))
Y <- ifelse(S == 1, Yi, NA_real_)
data.frame(W1 = W1, W2 = W2, W3 = W3,
S = S,
A = A,
Z = Z,
Y = Y,
Yi = Yi)
}
set.seed(123)
n <- 1000
tmp <- gendata(n)
if (requireNamespace("ranger", quietly = TRUE)) {
transport_ittate(data = tmp,
trt = "Z",
instrument = "A",
outcome = "Y",
covar = c("W1", "W2", "W3"),
pop = "S",
folds = 1)
}