cvtn_cvtmle: Compute CVTML estimates of cross-validated AUC

In benkeser/cvtmleAUC: Cross-validated TML estimates of cross-validated area under the receiver operating characteristic curve

Description

TO DO: Add

Usage

cvtn_cvtmle(Y, X, K = 20, sens = 0.95, learner = "glm_wrapper",
  nested_cv = TRUE, nested_K = K - 1, parallel = FALSE,
  maxIter = 10, icTol = 1/length(Y), quantile_type = 8,
  prediction_list = NULL, ...)

Arguments

Y	The outcome
X	The predictors
K	The number of folds
learner	The learner wrapper
parallel	Compute the predictors in parallel?
maxIter	Maximum number of iterations for cvtmle
icTol	Iterate until maxIter is reach or mean of cross-validated efficient influence function is less than icTol
...	other arguments, not currently used
seed	A random seed to set

Value

A list TO DO: Seems like this needs to return the quantiles in order to be useful in a practical sense.

Examples

n <- 200
p <- 10
X <- matrix(rnorm(n*p), nrow = n, ncol = p)
Y <- rbinom(n, 1, plogis(X[,1] + X[,10]))
fit <- cvtn_cvtmle(Y = Y, X = X, K = 5, learner = "glm_wrapper")

benkeser/cvtmleAUC documentation built on May 16, 2019, 2:30 a.m.