ORION (version 1.0.3)

subcascades: Subcascades Evaluation

Description

Subcascades returns all cascades found within the data or evaluates a set of specific cascades.

Usage

subcascades(
  predictionMap = NULL,
  sets = NULL,
  thresh = 0,
  size = NA,
  numSol = 1000
)

Arguments

predictionMap

A PredictionMap object as it is returned by predictionMap-function. It is made up of a list of two matrices(pred and meta). Both matrices provide information on individual samples column-wise. The rownames of the pred-matrix (e.g. [0vs1]) show the classes of the binary base classifier. The elements are the prediction result of a specific training. The rows that correspond to base classifiers that would separate the same class consists of -1. Those rows are not used within the analysis. The meta information connects the values in the pred-matrix to a specific fold, run and contains the original label.

sets

Contains the set used for filtering. It is either a list of numeric vectors, a numeric vector, or a vector of characters representing a cascade of the following format '1>2>4'. Empty vectors are not allowed.

thresh

A numeric value between 0 and 1. The minimal sensitivity threshold used to filter the returned cascades. Only cascades that pass this threshold are returned. If 0 is used the returned cascades are filtered for >0 and otherwise >= thresh. For low thresholds the calculation lasts longer.

size

A numeric value that defines the size of the cascades that should be returned. The smallest size is 2 and the largest the maximal number of classes of the current dataset. If size is NA (the default setting), all cascades from 2 to the maximal number of classes are evaluated.

numSol

The maximum number of cascades that should pass the first sensitivity bound and are further evaluated.

Value

A Subcascades object comprising the evaluated cascades and their performances. The Subcascades object is made up of a list of matrices. Each matrix comprises the evaluation results of cascades of a specific length and is sorted row-wise according to the achieved minimal classwise sensitivities of the cascades (decreasing). The rownames show the class order by a character string of type '1>2>3' and the entries the sensitivity for each position of the cascade.

Details

This function can either be used to evaluate the performance of a specific cascade, a set of cascades or to filter out the set of cascades of a specific size and passing a given threshold. If the sets-variable is given no size can be set.

See Also

print.Subcascades, plot.Subcascades, summary.Subcascades

Examples

# NOT RUN {
library(TunePareto)
data(esl)
data = esl$data
labels = esl$labels
foldList = generateCVRuns(labels  = labels,
                          ntimes      = 2,
                          nfold       = 2,
                          leaveOneOut = FALSE,
                          stratified  = TRUE)
predMap = predictionMap(data, labels, foldList = foldList, 
                       classifier = tunePareto.svm(),  kernel='linear')

# use default parameter settings 
# -> returns cascades of all lengths that show a minimal classwise sensitivity >0.
subc = subcascades(predMap)
# change the threshold 
# -> returns cascades of all lengths that show a minimal classwise sensitivity >=0.6.
subc = subcascades(predMap, thresh=0.6)
# search only for cascades of length 2 and 4 
# -> returns cascades of length 2 and 4 that show a minimal classwise sensitivity >=0.6.
subc = subcascades(predMap, thresh=0.6, size=c(2,4))
# evaluates the performance of the cascade '0>1>2>3>4'.
subc = subcascades(predMap, sets = c('0>1>2>3>4'))
# }