bootstrap: bootstrap

In Peder2911/evallib: Simple binary classifier evaluation

Description

Apply fun to the data n times. The data is randomly sampled each time. This yields both more robust results and approximate confidence intervals for the results. Due to the random sampling, the results of this function are stochastic. Set the seed with set.seed(n) for reproducible results.

Usage

bootstrap(pred, actual, fun, draws = 100, parallel = FALSE, ...)

Arguments

pred	A vector of predictions
actual	A vector of actual values
fun	The function to apply
draws	Number of times to randomly sample the data for bootstrapping. Larger numbers yield better results, but also increase runtime.
parallel	Bool. Run in parallel?
...	Arguments passed to fun

Value

A list containing the results of each iteration

Examples

pred <- sample(seq(0,1,0.001), size = 150, replace = TRUE)
actual <- sample(c(0,1), size = 150, replace = TRUE)
bootstrap(pred,actual,roc)

Peder2911/evallib documentation built on Dec. 18, 2019, 2:41 a.m.