Package weka.classifiers.trees

Class RandomForest

  • All Implemented Interfaces:
    java.io.Serializable, java.lang.Cloneable, Classifier, AdditionalMeasureProducer, Aggregateable<Bagging>, BatchPredictor, CapabilitiesHandler, CapabilitiesIgnorer, CommandlineRunnable, OptionHandler, PartitionGenerator, Randomizable, RevisionHandler, TechnicalInformationHandler, WeightedInstancesHandler
    public class RandomForest
extends Bagging
    Class for constructing a forest of random trees.

    For more information see:

    Leo Breiman (2001). Random Forests. Machine Learning. 45(1):5-32.

    BibTeX: 
     @article{Breiman2001,
    author = {Leo Breiman},
    journal = {Machine Learning},
    number = {1},
    pages = {5-32},
    title = {Random Forests},
    volume = {45},
    year = {2001}
 }


    Valid options are: 

     -P
  Size of each bag, as a percentage of the
  training set size. (default 100)
 
     -O
  Calculate the out of bag error.
 
     -store-out-of-bag-predictions
  Whether to store out of bag predictions in internal evaluation object.
 
     -output-out-of-bag-complexity-statistics
  Whether to output complexity-based statistics when out-of-bag evaluation is performed.
 
     -print
  Print the individual classifiers in the output
 
     -attribute-importance
  Compute and output attribute importance (mean impurity decrease method)
 
     -I <num>
  Number of iterations (i.e., the number of trees in the random forest).
  (current value 100)
 
     -num-slots <num>
  Number of execution slots.
  (default 1 - i.e. no parallelism)
  (use 0 to auto-detect number of cores)
 
     -K <number of attributes>
  Number of attributes to randomly investigate. (default 0)
  (<1 = int(log_2(#predictors)+1)).
 
     -M <minimum number of instances>
  Set minimum number of instances per leaf.
  (default 1)
 
     -V <minimum variance for split>
  Set minimum numeric class variance proportion
  of train variance for split (default 1e-3).
 
     -S <num>
  Seed for random number generator.
  (default 1)
 
     -depth <num>
  The maximum depth of the tree, 0 for unlimited.
  (default 0)
 
     -N <num>
  Number of folds for backfitting (default 0, no backfitting).
 
     -U
  Allow unclassified instances.
 
     -B
  Break ties randomly when several attributes look equally good.
 
     -output-debug-info
  If set, classifier is run in debug mode and
  may output additional info to the console
 
     -do-not-check-capabilities
  If set, classifier capabilities are not checked before classifier is built
  (use with caution).
 
     -num-decimal-places
  The number of decimal places for the output of numbers in the model (default 2).
 
     -batch-size
  The desired batch size for batch prediction  (default 100).
    Version:
    $Revision: 15312 $
    Author:
    Richard Kirkby (rkirkby@cs.waikato.ac.nz)
    See Also:
    Serialized Form

    • Constructor Detail

      • RandomForest

        public RandomForest()
        Constructor that sets base classifier for bagging to RandomTre and default number of iterations to 100.

    • Method Detail

      • globalInfo

        public java.lang.String globalInfo()
        Returns a string describing classifier
        Overrides:
        globalInfo in class Bagging
        Returns:
        a description suitable for displaying in the explorer/experimenter gui

      • numIterationsTipText

        public java.lang.String numIterationsTipText()
        Returns the tip text for the number of iterations. Overridden here to be more informative.
        Overrides:
        numIterationsTipText in class IteratedSingleClassifierEnhancer
        Returns:
        tip text for this property suitable for displaying in the explorer/experimenter gui

      • setRepresentCopiesUsingWeights

        @ProgrammaticProperty
public void setRepresentCopiesUsingWeights​(boolean representUsingWeights)
        This method only accepts true as its argument
        Overrides:
        setRepresentCopiesUsingWeights in class Bagging
        Parameters:
        representUsingWeights - must be set to true.

      • numFeaturesTipText

        public java.lang.String numFeaturesTipText()
        Returns the tip text for this property
        Returns:
        tip text for this property suitable for displaying in the explorer/experimenter gui

      • getNumFeatures

        public int getNumFeatures()
        Get the number of features used in random selection.
        Returns:
        Value of numFeatures.

      • setNumFeatures

        public void setNumFeatures​(int newNumFeatures)
        Set the number of features to use in random selection.
        Parameters:
        newNumFeatures - Value to assign to numFeatures.

      • computeAttributeImportanceTipText

        public java.lang.String computeAttributeImportanceTipText()
        Returns the tip text for this property
        Returns:
        tip text for this property suitable for displaying in the explorer/experimenter gui

      • setComputeAttributeImportance

        public void setComputeAttributeImportance​(boolean computeAttributeImportance)
        Set whether to compute and output attribute importance scores
        Parameters:
        computeAttributeImportance - true to compute attribute importance scores

      • getComputeAttributeImportance

        public boolean getComputeAttributeImportance()
        Get whether to compute and output attribute importance scores
        Returns:
        true if computing attribute importance scores

      • maxDepthTipText

        public java.lang.String maxDepthTipText()
        Returns the tip text for this property
        Returns:
        tip text for this property suitable for displaying in the explorer/experimenter gui

      • getMaxDepth

        public int getMaxDepth()
        Get the maximum depth of trh tree, 0 for unlimited.
        Returns:
        the maximum depth.

      • setMaxDepth

        public void setMaxDepth​(int value)
        Set the maximum depth of the tree, 0 for unlimited.
        Parameters:
        value - the maximum depth.

      • breakTiesRandomlyTipText

        public java.lang.String breakTiesRandomlyTipText()
        Returns the tip text for this property
        Returns:
        tip text for this property suitable for displaying in the explorer/experimenter gui

      • getBreakTiesRandomly

        public boolean getBreakTiesRandomly()
        Get whether to break ties randomly.
        Returns:
        true if ties are to be broken randomly.

      • setBreakTiesRandomly

        public void setBreakTiesRandomly​(boolean newBreakTiesRandomly)
        Set whether to break ties randomly.
        Parameters:
        newBreakTiesRandomly - true if ties are to be broken randomly

      • setDebug

        public void setDebug​(boolean debug)
        Set debugging mode.
        Overrides:
        setDebug in class AbstractClassifier
        Parameters:
        debug - true if debug output should be printed

      • toString

        public java.lang.String toString()
        Returns description of the bagged classifier.
        Overrides:
        toString in class Bagging
        Returns:
        description of the bagged classifier as a string

      • computeAverageImpurityDecreasePerAttribute

        public double[] computeAverageImpurityDecreasePerAttribute​(double[] nodeCounts)
                                                    throws WekaException
        Computes the average impurity decrease per attribute over the trees
        Parameters:
        nodeCounts - an optional array that, if non-null, will hold the count of the number of nodes at which each attribute was used for splitting
        Returns:
        the average impurity decrease per attribute over the trees
        Throws:
        WekaException

      • listOptions

        public java.util.Enumeration<Option> listOptions()
        Returns an enumeration describing the available options.
        Specified by:
        listOptions in interface OptionHandler
        Overrides:
        listOptions in class Bagging
        Returns:
        an enumeration of all the available options

      • getOptions

        public java.lang.String[] getOptions()
        Gets the current settings of the forest.
        Specified by:
        getOptions in interface OptionHandler
        Overrides:
        getOptions in class Bagging
        Returns:
        an array of strings suitable for passing to setOptions()

      • setOptions

        public void setOptions​(java.lang.String[] options)
                throws java.lang.Exception
        Parses a given list of options.

        Valid options are: 

         -P
  Size of each bag, as a percentage of the
  training set size. (default 100)
 
         -O
  Calculate the out of bag error.
 
         -store-out-of-bag-predictions
  Whether to store out of bag predictions in internal evaluation object.
 
         -output-out-of-bag-complexity-statistics
  Whether to output complexity-based statistics when out-of-bag evaluation is performed.
 
         -print
  Print the individual classifiers in the output
 
         -attribute-importance
  Compute and output attribute importance (mean impurity decrease method)
 
         -I <num>
  Number of iterations (i.e., the number of trees in the random forest).
  (current value 100)
 
         -num-slots <num>
  Number of execution slots.
  (default 1 - i.e. no parallelism)
  (use 0 to auto-detect number of cores)
 
         -K <number of attributes>
  Number of attributes to randomly investigate. (default 0)
  (<1 = int(log_2(#predictors)+1)).
 
         -M <minimum number of instances>
  Set minimum number of instances per leaf.
  (default 1)
 
         -V <minimum variance for split>
  Set minimum numeric class variance proportion
  of train variance for split (default 1e-3).
 
         -S <num>
  Seed for random number generator.
  (default 1)
 
         -depth <num>
  The maximum depth of the tree, 0 for unlimited.
  (default 0)
 
         -N <num>
  Number of folds for backfitting (default 0, no backfitting).
 
         -U
  Allow unclassified instances.
 
         -B
  Break ties randomly when several attributes look equally good.
 
         -output-debug-info
  If set, classifier is run in debug mode and
  may output additional info to the console
 
         -do-not-check-capabilities
  If set, classifier capabilities are not checked before classifier is built
  (use with caution).
 
         -num-decimal-places
  The number of decimal places for the output of numbers in the model (default 2).
 
         -batch-size
  The desired batch size for batch prediction  (default 100).
        Specified by:
        setOptions in interface OptionHandler
        Overrides:
        setOptions in class Bagging
        Parameters:
        options - the list of options as an array of strings
        Throws:
        java.lang.Exception - if an option is not supported

      • main

        public static void main​(java.lang.String[] argv)
        Main method for this class.
        Parameters:
        argv - the options