one_boot_scrnp {nlpred}

Internal function used to perform one bootstrap sample. The function trys to fit learner on a bootstrap sample. If for some reason (e.g., the bootstrap sample contains no observations with Y = 1) the learner fails, then the function returns NA. These NAs are ignored later when computing the bootstrap corrected estimate.

Description

Internal function used to perform one bootstrap sample. The function trys to fit learner on a bootstrap sample. If for some reason (e.g., the bootstrap sample contains no observations with Y = 1) the learner fails, then the function returns NA. These NAs are ignored later when computing the bootstrap corrected estimate.

Usage

one_boot_scrnp(Y, X, n, correct632, learner, sens)

Arguments

Y

A numeric binary outcome

X

A data.frame of variables for prediction.

n

Number of observations

correct632

A boolean indicating whether to use the .632 correction.

learner

A wrapper that implements the desired method for building a prediction algorithm. See ?glm_wrapper or read the package vignette for more information on formatting learners.

sens

The sensitivity constraint to use.

Value

If learner executes successfully, a numeric estimate of AUC on this bootstrap sample. Otherwise the function returns NA.