Wrapper for the highly-adaptive lasso (HAL) algorithm
implemented in the hal9001 package
Usage
SL.hal9001(Y, X, newX, family, obsWeights, ...)
Arguments
- Y
Outcome variable
- X
Matrix of covariates
- newX
Matrix of covariates for prediction
- family
Family of the outcome variable
- obsWeights
Observation weights
- ...
Additional arguments to pass to the HAL function
Value
A list containing the prediction and the fitted model
Examples
if (FALSE) { # \dontrun{
n <- 200
X <- data.frame(X1 = rnorm(n))
Y <- X$X1 + rnorm(n)
fit <- SL.hal9001(Y, X, newX = X, family = gaussian(),
obsWeights = rep(1, n))
head(fit$pred)
} # }