Measures the quality of a prediction w.r.t. some performance measure.
Arguments
- pred
(Prediction)
Prediction object.- measures
(Measure | list of Measure)
Performance measure(s) to evaluate. Default is the default measure for the task, see here getDefaultMeasure.- task
(Task)
Learning task, might be requested by performance measure, usually not needed except for clustering or survival.- model
(WrappedModel)
Model built on training data, might be requested by performance measure, usually not needed except for survival.- feats
(data.frame)
Features of predicted data, usually not needed except for clustering. If the prediction was generated from atask
, you can also pass this instead and the features are extracted from it.- simpleaggr
(logical)
If TRUE, aggregation ofResamplePrediction
objects is skipped. This is used internally for threshold tuning. Default isFALSE
.
Value
(named numeric). Performance value(s), named by measure(s).
See also
Other performance:
ConfusionMatrix
,
calculateConfusionMatrix()
,
calculateROCMeasures()
,
estimateRelativeOverfitting()
,
makeCostMeasure()
,
makeCustomResampledMeasure()
,
makeMeasure()
,
measures
,
setAggregation()
,
setMeasurePars()
Examples
training.set = seq(1, nrow(iris), by = 2)
test.set = seq(2, nrow(iris), by = 2)
task = makeClassifTask(data = iris, target = "Species")
lrn = makeLearner("classif.lda")
mod = train(lrn, task, subset = training.set)
#> Error in x[0, , drop = FALSE]: incorrect number of dimensions
pred = predict(mod, newdata = iris[test.set, ])
#> Error in predict(mod, newdata = iris[test.set, ]): object 'mod' not found
performance(pred, measures = mmce)
#> Error in performance(pred, measures = mmce): object 'pred' not found
# Compute multiple performance measures at once
ms = list("mmce" = mmce, "acc" = acc, "timetrain" = timetrain)
performance(pred, measures = ms, task, mod)
#> Error in performance(pred, measures = ms, task, mod): object 'pred' not found