bootstrappingMetric: Create an matrix of bootstrapped predictions.

In Sage-Bionetworks/challengescoring: Reducing Overfitting in Challenges With the BootLadderBoot Approach

Description

Create an matrix of bootstrapped predictions.

Usage

bootstrappingMetric(
  goldStandardMatrix,
  predictionsMatrix,
  scoreFun = scoreFun,
  bootstrapN = bootstrapN,
  seed = seed,
  doParallel = F,
  ...
)

Arguments

goldStandardMatrix	A single column matrix with the gold standard predictions.
predictionsMatrix	Columns of prediction sets, in the same order as the goldStandardMatrix.
bootstrapN	Number of total bootstraps to perform.
seed	Set a seed for bootstrap sampling.
doParallel	Bootstrap in parallel. Only works on UNIX based OS. Default FALSE.

Value

An MxN matrix of bootstrapped predictions where M is the number of bootstraps performed and N is the number of prediction sets.

Sage-Bionetworks/challengescoring documentation built on Sept. 28, 2020, 8:27 a.m.