pairsDifferencesSelection {ClassifyR} | R Documentation |
Ranks pre-specified pairs of features by the largest difference of the sum of measurement differences over all samples within a class and chooses the pairs of features which have the best resubstitution performance.
## S4 method for signature 'matrix' pairsDifferencesSelection(measurements, classes, featurePairs = NULL, ...) ## S4 method for signature 'DataFrame' pairsDifferencesSelection(measurements, classes, featurePairs = NULL, datasetName, trainParams, predictParams, resubstituteParams, selectionName = "Pairs Differences", verbose = 3) ## S4 method for signature 'MultiAssayExperiment' pairsDifferencesSelection(measurements, target = names(measurements)[1], featurePairs = NULL, ...)
measurements |
Either a |
classes |
Either a vector of class labels of class |
featurePairs |
An S4 object of type |
target |
If |
... |
Variables not used by the |
datasetName |
A name for the data set used. Stored in the result. |
trainParams |
A container of class |
predictParams |
A container of class |
resubstituteParams |
An object of class |
selectionName |
A name to identify this selection method by. Stored in the result. |
verbose |
Default: 3. A number between 0 and 3 for the amount of progress messages to give. This function only prints progress messages if the value is 3. |
Instead of considering whether one feature in a pair of features is consistently lower or higher than the other in the pair, this method takes the sum of differences across all samples within a class, to prevent ties in the ranking of pairs of features.
An object of class SelectResult
or a list of such objects, if the classifier which was
used for determining the specified performance metric made a number of prediction varieties.
Dario Strbenac
Simple decision rules for classifying human cancers from gene expression profiles, Aik C Tan, Daniel Q Naiman, Lei Xu, Raimond L. Winslow and Donald Geman, 2005, Bioinformatics, Volume 21 Issue 20, https://academic.oup.com/bioinformatics/article/21/20/3896/203010.
kTSPclassifier
for a classifier which makes use of the pairs of selected features in classification.
featurePairs <- Pairs(c('A', 'A'), c('B', 'C')) # Difference in differences for features A and C between classes. measurements <- matrix(c(9.9, 10.5, 10.1, 10.9, 11.0, 6.6, 7.7, 7.0, 8.1, 6.5, 8.5, 10.5, 12.5, 10.5, 9.5, 8.5, 10.5, 12.5, 10.5, 9.5, 6.6, 7.7, 7.0, 8.1, 6.5, 11.2, 11.0, 11.1, 11.4, 12.0, 8.1, 10.6, 7.4, 7.1, 10.4, 6.1, 7.3, 2.7, 11.0, 9.1, round(rnorm(60, 8, 1), 1)), ncol = 10, byrow = TRUE) classes <- factor(rep(c("Good", "Poor"), each = 5)) rownames(measurements) <- LETTERS[1:10] colnames(measurements) <- names(classes) <- paste("Patient", 1:10) # The features are pairs and there are only two in this example. resubstituteParams <- ResubstituteParams(nFeatures = 1:2, performanceType = "balanced error", better = "lower") predictParams <- PredictParams(NULL) pairsDifferencesSelection(measurements, classes, featurePairs = featurePairs, datasetName = "Example", trainParams = TrainParams(kTSPclassifier), predictParams = predictParams, resubstituteParams = resubstituteParams)