MATLAB File Help: prtDataSetClass
prtDataSetClass
 prtDataSetClass < prtDataSetStandard & prtDataInterfaceCategoricalTargets
  dataSet = prtDataSetClass generates a prtDataSetClass object
 
      dataSet = prtDataSetClass(X,Y) generates a prtDataSetClass from
      the nObservations x nFeatures double matrix X and the
      nObservations x 1 label vector Y.
 
      dataSet = prtDataSetClass(PROPERTY1, VALUE1, ...) constructs a
      prtDataSetClass object dataSet with properties as specified by
      PROPERTY/VALUE pairs.
 
      A prtDataSetClass object inherits all properties from the
      prtDataSetStandard class. In addition, it has the following properties:
 
      nClasses             - The number of classes
      uniqueClasses        - An array of the integer class labels
      isUnary              - True if the number of classes = 1
      isBinary             - True if the number of classes = 2
      isMary               - True if the number of classes > 2
      isZeroOne            - True if the unique classes are 0 and 1
      nObservationsByClass - The number of observations per class.
 
      A prtDataSetClass inherits all methods from the prtDataSetStandard
      class. In addition, it has the following methods:
 
      getObservationsByClass     - Return the observations per class
      getObservationsByClassInd  - Return the observations by class and index
      getTargetsAsBinaryMatrix   - Return a binary matrix of targets.
      explore                    - Explore the prtDataSetClass object
      plot                       - Plot the data set
      plotFeatureDensity         - Plot the probability density estimate
                                   of a single feature, labeled by class.
      plotStar                   - Create a star plot to visualize higher
                                   dimensional data
      plotAsTimeSeries           - Plot the prtDataSetClass object as a
                                   time series
      plotPairs                  - Plot a grid of plots containing each
                                   set of two-features with density
                                   estimates on the diagonals
      plotDensity                - Plot the density of each feature
                                   independently as a volume
 
      Note: As of Jan 5, 2013 prtDataSetClass enables the use of NAN
      values in the target vector to specify unlabled data.  This data
      and class value (NAN) are treated specially.  e.g., NAN labels do
      not count towards the nClasses and will not appear in
      uniqueClasses.  It is not possible to set the name of
      the class corresponding to NAN.  See
      prtDataInterfaceCategoricalTargets for more information.
Class Details
Superclasses prtDataSetStandard, prtDataInterfaceCategoricalTargets
Sealed false
Construct on load false
Constructor Summary
prtDataSetClass Constructor for class prtDataSetClass 
Property Summary
classNames  
data  
description A string with a verbose description of the data set 
featureInfo Additional data (structure) per feature 
featureInfoInternal  
featureNameModificationFunctionInternal  
featureNameModificationMaskInternal  
hasUnlabeled True if any of the labels are nan 
internalData  
internalSizeConsitencyCheck  
internalTargets  
isBinary True if the number of classes = 2 
isLabeled Whether or not the data has target labels 
isMary True if the number of classes > 2 
isUnary True if the number of classes = 1 
isZeroOne True if the uniqueClasses are 0 and 
nClasses  
nFeatures The number of features 
nObservations The number of observations 
nObservationsByClass histogram of samples x class 
nTargetDimensions The number of target dimensions 
name A string naming the data set 
observationInfo Struct of observation information 
observationInfoInternal  
observationNamesInternal The observations names 
targetNamesInternal The target names. 
targets  
uniqueClasses  
userData Additional data per data set 
Method Summary
  bootstrap dsBoot = bootstrap(dataSet,nSamples) 
  bootstrapByClass  
  catClasses self = catClasses(self,ds1,ds2,...) 
  catFeatures dsOut = catFeatures(dataSet1,dataSet2) 
protected   catObservationInfo  
  catObservations dsOut = catObservations(dataSet1,dataSet2) 
  catTargets dsOut = catTargets(dataSet1,dataSet2) 
protected Static   checkConsistency  
  classNamesToClassInd classInds = classNamesToClassInd(dataSet,classNames) 
  explore Explore the prtDataSetObject 
  exploreSimple Explore the prtDataSetObject using basic 
  getBinaryTargetsAsZeroOne Return the target vector from a 
  getClassNames cn = getClassNames(self,indices) 
  getClassNamesByClassInd  
  getData data = getData(dataSet) 
  getDataByClass Return the Data by class 
  getDataByClassInd  
  getDataUnlabeled  
  getFeatureInfo  
  getFeatureNames names = getFeatureNames(dataSet) 
  getFeatures Return the features of a prtDataSetStandard 
  getKFoldKeys keys = getKFoldKeys(dataSet,K) 
  getNumFeatures  
  getNumObservations nObs = getNumObservations(dataSet) 
  getNumTargetDimensions nTargets = getNumTargetDimensions(dataSet) 
  getObservationInfo obsInfo = getObservationInfo(dataSet) 
  getObservationNames Return DataSet's Observation Names 
  getObservations d = getObservations(dataSet) 
  getObservationsByClass  
  getObservationsByClassInd  
  getObservationsUnlabeled  
  getTargetNames Return the target names from a dataset 
  getTargets targets = getTargets(dataSet) 
  getTargetsAsBinaryMatrix binaryMatTargets Return the targets as a binary matrix 
  getTargetsClassInd getTargetsClassIndex Return the targets by class index 
  getUniqueClasses  
  getX Shortcut for GetObservations 
  getXY Shortcut for getObservationsAndTargets 
  getY Shortcut for getTargets 
Static   loadobj dataSet = loadobj(obj) 
  numClasses  
  permuteFeatures dsOut = permuteFeatures(dataSet,permuteFeatureInds) 
  plot Plot the prtDataSetClass object 
  plotAsTimeSeries Plot the data set as time series data 
  plotBeeSwarm Plot the density of each feature independently as a scatter 
  plotDensity Plot the density of each feature independently as a volume. 
  plotFeatureDensity Plot the densities of a single feature 
  plotPairs () - Plot each pair of features in feature space 
  plotStar Create a star plot 
  plotStarIndividual Create a star plot 
  removeClasses remove observations corresponding to 
  removeClassesByInd removeClasses remove observations corresponding to 
  removeFeatures dsOut = removeFeatures(dataSet,removeFeatureInds) 
  removeLabeled  
  removeObservations dsOut = removeObservations(dataSet,indices) 
  removeTargets dsOut = removeTargets(dataSet,indices) 
  removeUnlabeled  
  retainClasses retain observations corresponding to specified 
  retainClassesByInd retain observations corresponding to 
protected   retainFeatureNames  
  retainFeatures dsOut = retainFeatures(dataSet,retainFeatureInds) 
  retainLabeled  
protected   retainObservationInfo  
  retainObservations dsOut = removeObservations(dataSet,indices) 
  retainTargets dsOut = retainTargets(dataSet,indices) 
  retainUnlabeled  
  select observations to retain by specifying a function 
  setClassNames ds = ds.setClassNames({'fasdf','asdf','asdfdsfdsf'}); 
  setData dsOut = setData(dataSet,dataIn) 
  setFeatureInfo  
  setFeatureNames dataSet = setFeatureNames(dataSet,names) 
  setFeatures dsOut = setFeatures(dataSet,data,indices) 
  setObservationInfo dataSet = setObservationInfo(dataSet,obsInfoStruct) 
  setObservationNames Set the observation names of a data set 
  setObservations dataSet = setObservations(dataSet,data) 
  setObservationsAndTargets dataSet = setObservationsAndTargets(dataSet,data,targets) 
  setTargetNames Set the data set target names 
  setTargets dataSet = setTargets(ds,targets) 
  setX Shortcut for setObservations 
  setXY Shortcut for setObservationsAndTargets 
  setY Shortcut for setTargets 
  summarize Summarize the prtDataSetStandard object 
protected   update Updated chached target info 
protected   updateObservationsCache default behaviour is to do nothing