Create setting for creating ensemble from model settings

Create setting for creating ensemble from model settings

setEnsembleFromResults(resultList, filterSettings, combinerSettings)

Arguments

resultList

A list of runPlp results to combine.

filterSettings

Setting specifying rules to use to filter (remove) any model specified in the list of model designs that performs insufficiently (these models get ignored from the ensemble)

combinerSettings

Settings specifying how to combine the remaining models into an ensemble

Examples

if (FALSE) {

plpResult1 <- PatientLevelPrediction::loadPlpResult("./result1")
plpResult2 <- PatientLevelPrediction::loadPlpResult("./result2")
plpResult3 <- PatientLevelPrediction::loadPlpResult("./result3")

ensembleSettings <- setEnsembleFromResults(resultList = list(plpResult1, plpResult2, plpResult3),
                                           filterSettings = list(minValue = 0.5, maxValue = 1),
                                           combinerSettings = createFusionCombiner(type = "uniform",
                                                                                   evaluation = "CV",
                                                                                   scaleFunction = "normalize"))
}