Measure Class

This is the abstract base class for measures like MeasureClassif and MeasureRegr. Predefined measures are stored in mlr_measures.


R6::R6Class object.


m = Measure$new(id, task_type, range, minimize, predict_type = "response",
     task_properties = character(0L), packages = character(0L))
  • id :: character(1) Identifier for the measure.

  • task_type :: character(1) Type of the task the measure can operator on. E.g., \"classif\" or \"regr\".

  • range :: numeric(2) Feasible range for this measure as c(lower_bound, upper_bound).

  • minimize :: logical(1) Set to TRUE if good predictions correspond to small values.

  • predict_type :: character(1) Required predict type of the Learner.

  • task_properties :: character() Required task properties, see Task.

  • packages :: character() Set of required packages. Note that these packages will be loaded via requireNamespace(), and are not attached.


  • id :: character(1) Stores the identifier of the measure.

  • minimize :: logical(1) Is TRUE if the best value is reached via minimization and FALSE by maximization.

  • packages :: character() Stores the names of required packages.

  • range :: numeric(2) Stores the feasible range of the measure.

  • task_type :: character(1) Stores the required type of the Task.

  • task_properties :: character() Stores required properties of the Task.


  • aggregate(rr) ResampleResult -> numeric(1) Aggregates multiple performance scores into a single score using the aggregate function of the measure. Operates on a ResampleResult as returned by resample.

  • calculate(e) Experiment -> numeric(1) Takes an Experiment, extracts the predictions (as well as other possibly needed objects), and calculates a score.

See Also

Other Measure: MeasureClassif, MeasureRegr, mlr_measures

  • Measure
Documentation reproduced from package mlr3, version 0.1.0-9000, License: MIT + file LICENSE

Community examples

Looks like there are no examples yet.