loss_adapt_mlr3proba function

Adapt mlr3proba measures for use with survex

This function allows for usage of standardized measures from the mlr3proba package with survex.

loss_adapt_mlr3proba(measure, reverse = FALSE, ...)

Arguments

• measure: * a MeasureSurv object from the mlr3proba package, the object to adapt
• reverse: * boolean, FALSE by default, whether the metric should be reversed in order to be treated as loss (for permutational variable importance we need functions with lower values indicating better performance). If TRUE, the new metric value will be (1 - metric_value)
• ...: * other parameters, currently ignored

Returns

a function with standardized parameters (y_true, risk, surv, times) that can be used to calculate loss

Examples

if(FALSE){
  measure <- msr("surv.calib_beta")
  mlr_measure <- loss_adapt_mlr3proba(measure)
}