compExp-class: Class "compExp"

Description Objects from the Class Slots Methods Author(s) References See Also Examples


This is the main class that holds the results of experimental comparisons of a set of learners over a set of predictive tasks, using some experimental methodology.

Objects from the Class

Objects can be created by calls of the form compExp(...). These objects contain information on the set of learners being compared, the set of predictive tasks being used on the comparison, the experimental settings and the overall results of all experimental comparisons.



Object of class "list" : a list of objects of the class learner.


Object of class "list" : a list of objects of the class task.


Object of class "expSettings" : an object belonging to one of the classes in this class union.


Object of class "array" : a numeric array with the overall results of the experiment. This array has 4 dimensions. The first dimension are the different repetitions/iterations of the experiment; the second dimension are the evaluation statistics being estimated; the third dimension are the different learners being compared; while the fourth dimension are the predictive tasks.



signature(x = "compExp", y = "missing"): plots the results of the experiments. It can result in an over-cluttered graph if too many learners/datasets/evaluation metrics - use the subset method (see below) to overcome this.


signature(object = "compExp"): shows the contents of an object in a proper way


signature(x = "compExp"): can be used to obtain a smaller compExp object containing only a subset of the information of the provided object. This method also accepts the arguments "its", "stats", "vars" and "dss". All are vectors of numbers or names corresponding to an indexing of each of the dimensions of the "foldResults" slot. They default to all values of each dimension. See "methods?subset" for further details.


signature(object = "compExp"): provides a summary of the experimental results.


Luis Torgo


Torgo, L. (2010) Data Mining using R: learning with case studies, CRC Press (ISBN: 9781439810187).

See Also

experimentalComparison, compAnalysis, rankSystems, bestScores, statScores, join



DMwR documentation built on May 1, 2019, 9:17 p.m.