mlr (version 2.7)

plotCritDifferences: Plot critical differences for a selected measure.

Description

Plots a critical-differences diagram for all classifiers and a selected measure. If a baseline is selected for the Bonferroni-Dunn test, the critical difference interval will be positioned arround the baseline. If not, the best performing algorithm will be chosen as baseline. The positioning of some descriptive elements can be moved by modifying the generated data.

Usage

plotCritDifferences(obj, baseline = NULL)

Value

ggplot2 plot object.

References

Janez Demsar, Statistical Comparisons of Classifiers over Multiple Data Sets, JMLR, 2006

See Also

Other benchmark: BenchmarkResult, benchmark, convertBMRToRankMatrix, friedmanPostHocTestBMR, friedmanTestBMR, generateCritDifferencesData, getBMRAggrPerformances, getBMRFeatSelResults, getBMRFilteredFeatures, getBMRLearnerIds, getBMRLearners, getBMRMeasureIds, getBMRMeasures, getBMRPerformances, getBMRPredictions, getBMRTaskIds, getBMRTuneResults, plotBMRBoxplots, plotBMRRanksAsBarChart, plotBMRSummary

Other plot: plotBMRBoxplots, plotBMRRanksAsBarChart, plotBMRSummary, plotCalibration, plotFilterValuesGGVIS, plotFilterValues, plotLearningCurveGGVIS, plotLearningCurve, plotPartialPredictionGGVIS, plotPartialPrediction, plotROCCurves, plotROCRCurvesGGVIS, plotROCRCurves, plotThreshVsPerfGGVIS, plotThreshVsPerf

Examples

lrns = list(makeLearner("classif.nnet"), makeLearner("classif.rpart"))
tasks = list(iris.task, sonar.task)
rdesc = makeResampleDesc("CV", iters = 2L)
meas = list(acc, mmce)
res = benchmark(lrns, tasks, rdesc, meas)
r = generateCritDifferencesData(res, mmce, p.value = 0.3, test = "bd")
plotCritDifferences(r)