Package weka.core

Class TestInstances

  • All Implemented Interfaces:
    Serializable, Cloneable, OptionHandler, RevisionHandler
    public class TestInstances
extends Object
implements Cloneable, Serializable, OptionHandler, RevisionHandler
    Generates artificial datasets for testing. In case of Multi-Instance data the settings for the number of attributes applies to the data inside the bag. Originally based on code from the CheckClassifier.

    Valid options are: 

     -relation <name>
  The name of the data set.
 
     -seed <num>
  The seed value.
 
     -num-instances <num>
  The number of instances in the datasets (default 20).
 
     -class-type <num>
  The class type, see constants in weka.core.Attribute
  (default 1=nominal).
 
     -class-values <num>
  The number of classes to generate (for nominal classes only)
  (default 2).
 
     -class-index <num>
  The class index, with -1=last, (default -1).
 
     -no-class
  Doesn't include a class attribute in the output.
 
     -nominal <num>
  The number of nominal attributes (default 1).
 
     -nominal-values <num>
  The number of values for nominal attributes (default 2).
 
     -numeric <num>
  The number of numeric attributes (default 0).
 
     -string <num>
  The number of string attributes (default 0).
 
     -words <comma-separated-list>
  The words to use in string attributes.
 
     -word-separators <chars>
  The word separators to use in string attributes.
 
     -date <num>
  The number of date attributes (default 0).
 
     -relational <num>
  The number of relational attributes (default 0).
 
     -relational-nominal <num>
  The number of nominal attributes in a rel. attribute (default 1).
 
     -relational-nominal-values <num>
  The number of values for nominal attributes in a rel. attribute (default 2).
 
     -relational-numeric <num>
  The number of numeric attributes in a rel. attribute (default 0).
 
     -relational-string <num>
  The number of string attributes in a rel. attribute (default 0).
 
     -relational-date <num>
  The number of date attributes in a rel. attribute (default 0).
 
     -num-instances-relational <num>
  The number of instances in relational/bag attributes (default 10).
 
     -multi-instance
  Generates multi-instance data.
 
     -W <classname>
  The Capabilities handler to base the dataset on.
  The other parameters can be used to override the ones
  determined from the handler. Additional parameters for
  handler can be passed on after the '--'.
    Version:
    $Revision$
    Author:
    FracPete (fracpete at waikato dot ac dot nz)
    See Also:
    CheckClassifier, Serialized Form

    • Constructor Detail

      • TestInstances

        public TestInstances()
        the default constructor

    • Method Detail

      • clone

        public Object clone()
        creates a clone of the current object
        Returns:
        a clone of the current object

      • assign

        public void assign​(TestInstances t)
        updates itself with all the settings from the given TestInstances object
        Parameters:
        t - the object to get the settings from

      • listOptions

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

      • setOptions

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

        Valid options are: 

         -relation <name>
  The name of the data set.
 
         -seed <num>
  The seed value.
 
         -num-instances <num>
  The number of instances in the datasets (default 20).
 
         -class-type <num>
  The class type, see constants in weka.core.Attribute
  (default 1=nominal).
 
         -class-values <num>
  The number of classes to generate (for nominal classes only)
  (default 2).
 
         -class-index <num>
  The class index, with -1=last, (default -1).
 
         -no-class
  Doesn't include a class attribute in the output.
 
         -nominal <num>
  The number of nominal attributes (default 1).
 
         -nominal-values <num>
  The number of values for nominal attributes (default 2).
 
         -numeric <num>
  The number of numeric attributes (default 0).
 
         -string <num>
  The number of string attributes (default 0).
 
         -words <comma-separated-list>
  The words to use in string attributes.
 
         -word-separators <chars>
  The word separators to use in string attributes.
 
         -date <num>
  The number of date attributes (default 0).
 
         -relational <num>
  The number of relational attributes (default 0).
 
         -relational-nominal <num>
  The number of nominal attributes in a rel. attribute (default 1).
 
         -relational-nominal-values <num>
  The number of values for nominal attributes in a rel. attribute (default 2).
 
         -relational-numeric <num>
  The number of numeric attributes in a rel. attribute (default 0).
 
         -relational-string <num>
  The number of string attributes in a rel. attribute (default 0).
 
         -relational-date <num>
  The number of date attributes in a rel. attribute (default 0).
 
         -num-instances-relational <num>
  The number of instances in relational/bag attributes (default 10).
 
         -multi-instance
  Generates multi-instance data.
 
         -W <classname>
  The Capabilities handler to base the dataset on.
  The other parameters can be used to override the ones
  determined from the handler. Additional parameters for
  handler can be passed on after the '--'.
        Specified by:
        setOptions in interface OptionHandler
        Parameters:
        options - the list of options as an array of strings
        Throws:
        Exception - if an option is not supported

      • getOptions

        public String[] getOptions()
        Gets the current settings of this object.
        Specified by:
        getOptions in interface OptionHandler
        Returns:
        an array of strings suitable for passing to setOptions

      • setRelation

        public void setRelation​(String value)
        sets the name of the relation
        Parameters:
        value - the name of the relation

      • getRelation

        public String getRelation()
        returns the current name of the relation
        Returns:
        the name of the relation

      • setSeed

        public void setSeed​(int value)
        sets the seed value for the random number generator
        Parameters:
        value - the seed

      • getSeed

        public int getSeed()
        returns the current seed value
        Returns:
        the seed value

      • setNumInstances

        public void setNumInstances​(int value)
        sets the number of instances to produce
        Parameters:
        value - the number of instances

      • getNumInstances

        public int getNumInstances()
        returns the current number of instances to produce
        Returns:
        the number of instances

      • setClassType

        public void setClassType​(int value)
        sets the class attribute type
        Parameters:
        value - the class attribute type

      • getClassType

        public int getClassType()
        returns the current class type
        Returns:
        the class attribute type

      • setNumClasses

        public void setNumClasses​(int value)
        sets the number of classes
        Parameters:
        value - the number of classes

      • getNumClasses

        public int getNumClasses()
        returns the current number of classes
        Returns:
        the number of classes

      • setClassIndex

        public void setClassIndex​(int value)
        sets the class index (0-based)
        Parameters:
        value - the class index
        See Also:
        CLASS_IS_LAST, NO_CLASS

      • getClassIndex

        public int getClassIndex()
        returns the current class index (0-based), -1 is last attribute
        Returns:
        the class index
        See Also:
        CLASS_IS_LAST, NO_CLASS

      • setNoClass

        public void setNoClass​(boolean value)
        whether to have no class, e.g., for clusterers; otherwise the class attribute index is set to last
        Parameters:
        value - whether to have no class
        See Also:
        CLASS_IS_LAST, NO_CLASS

      • getNoClass

        public boolean getNoClass()
        whether no class attribute is generated
        Returns:
        true if no class attribute is generated

      • setNumNominal

        public void setNumNominal​(int value)
        sets the number of nominal attributes
        Parameters:
        value - the number of nominal attributes

      • getNumNominal

        public int getNumNominal()
        returns the current number of nominal attributes
        Returns:
        the number of nominal attributes

      • setNumNominalValues

        public void setNumNominalValues​(int value)
        sets the number of values for nominal attributes
        Parameters:
        value - the number of values

      • getNumNominalValues

        public int getNumNominalValues()
        returns the current number of values for nominal attributes
        Returns:
        the number of values

      • setNumNumeric

        public void setNumNumeric​(int value)
        sets the number of numeric attributes
        Parameters:
        value - the number of numeric attributes

      • getNumNumeric

        public int getNumNumeric()
        returns the current number of numeric attributes
        Returns:
        the number of numeric attributes

      • setNumString

        public void setNumString​(int value)
        sets the number of string attributes
        Parameters:
        value - the number of string attributes

      • getNumString

        public int getNumString()
        returns the current number of string attributes
        Returns:
        the number of string attributes

      • setWords

        public void setWords​(String value)
        Sets the comma-separated list of words to use for generating strings. The list must contain at least 2 words, otherwise an exception will be thrown.
        Parameters:
        value - the list of words
        Throws:
        IllegalArgumentException - if not at least 2 words are provided

      • getWords

        public String getWords()
        returns the words used for assembling strings in a comma-separated list.
        Returns:
        the words as comma-separated list

      • setWordSeparators

        public void setWordSeparators​(String value)
        sets the word separators (chars) to use for assembling strings.
        Parameters:
        value - the characters to use as separators

      • getWordSeparators

        public String getWordSeparators()
        returns the word separators (chars) to use for assembling strings.
        Returns:
        the current separators

      • setNumDate

        public void setNumDate​(int value)
        sets the number of date attributes
        Parameters:
        value - the number of date attributes

      • getNumDate

        public int getNumDate()
        returns the current number of date attributes
        Returns:
        the number of date attributes

      • setNumRelational

        public void setNumRelational​(int value)
        sets the number of relational attributes
        Parameters:
        value - the number of relational attributes

      • getNumRelational

        public int getNumRelational()
        returns the current number of relational attributes
        Returns:
        the number of relational attributes

      • setNumRelationalNominal

        public void setNumRelationalNominal​(int value)
        sets the number of nominal attributes in a relational attribute
        Parameters:
        value - the number of nominal attributes

      • getNumRelationalNominal

        public int getNumRelationalNominal()
        returns the current number of nominal attributes in a relational attribute
        Returns:
        the number of nominal attributes

      • setNumRelationalNominalValues

        public void setNumRelationalNominalValues​(int value)
        sets the number of values for nominal attributes in a relational attribute
        Parameters:
        value - the number of values

      • getNumRelationalNominalValues

        public int getNumRelationalNominalValues()
        returns the current number of values for nominal attributes in a relational attribute
        Returns:
        the number of values

      • setNumRelationalNumeric

        public void setNumRelationalNumeric​(int value)
        sets the number of numeric attributes in a relational attribute
        Parameters:
        value - the number of numeric attributes

      • getNumRelationalNumeric

        public int getNumRelationalNumeric()
        returns the current number of numeric attributes in a relational attribute
        Returns:
        the number of numeric attributes

      • setNumRelationalString

        public void setNumRelationalString​(int value)
        sets the number of string attributes in a relational attribute
        Parameters:
        value - the number of string attributes

      • getNumRelationalString

        public int getNumRelationalString()
        returns the current number of string attributes in a relational attribute
        Returns:
        the number of string attributes

      • setNumRelationalDate

        public void setNumRelationalDate​(int value)
        sets the number of date attributes in a relational attribute
        Parameters:
        value - the number of date attributes

      • getNumRelationalDate

        public int getNumRelationalDate()
        returns the current number of date attributes in a relational attribute
        Returns:
        the number of date attributes

      • setNumInstancesRelational

        public void setNumInstancesRelational​(int value)
        sets the number of instances in relational/bag attributes to produce
        Parameters:
        value - the number of instances

      • getNumInstancesRelational

        public int getNumInstancesRelational()
        returns the current number of instances in relational/bag attributes to produce
        Returns:
        the number of instances

      • setMultiInstance

        public void setMultiInstance​(boolean value)
        sets whether multi-instance data should be generated (with a fixed data structure)
        Parameters:
        value - whether multi-instance data is generated

      • getMultiInstance

        public boolean getMultiInstance()
        Gets whether multi-instance data (with a fixed structure) is generated
        Returns:
        true if multi-instance data is generated

      • setRelationalFormat

        public void setRelationalFormat​(int index,
                                Instances value)
        sets the structure for the bags for the relational attribute
        Parameters:
        index - the index of the relational attribute
        value - the new structure

      • getRelationalFormat

        public Instances getRelationalFormat​(int index)
        returns the format for the specified relational attribute, can be null
        Parameters:
        index - the index of the relational attribute
        Returns:
        the current structure

      • setRelationalClassFormat

        public void setRelationalClassFormat​(Instances value)
        sets the structure for the relational class attribute
        Parameters:
        value - the structure for the relational attribute

      • getRelationalClassFormat

        public Instances getRelationalClassFormat()
        returns the current strcuture of the relational class attribute, can be null
        Returns:
        the relational structure of the class attribute

      • getNumAttributes

        public int getNumAttributes()
        returns the overall number of attributes (incl. class, if that is also generated)
        Returns:
        the overall number of attributes

      • getData

        public Instances getData()
        returns the current dataset, can be null
        Returns:
        the current dataset

      • setHandler

        public void setHandler​(CapabilitiesHandler value)
        sets the Capabilities handler to generate the data for
        Parameters:
        value - the handler to generate the data for

      • getHandler

        public CapabilitiesHandler getHandler()
        returns the current set CapabilitiesHandler to generate the dataset for, can be null
        Returns:
        the handler to generate the data for

      • generate

        public Instances generate()
                   throws Exception
        Generates a new dataset
        Returns:
        the generated data
        Throws:
        Exception - if something goes wrong

      • generate

        public Instances generate​(String namePrefix)
                   throws Exception
        generates a new dataset.
        Parameters:
        namePrefix - the prefix to add to the name of an attribute
        Returns:
        the generated data
        Throws:
        Exception - if something goes wrong

      • forCapabilities

        public static TestInstances forCapabilities​(Capabilities c)
        returns a TestInstances instance setup already for the the given capabilities.
        Parameters:
        c - the capabilities to base the TestInstances on
        Returns:
        the configured TestInstances object

      • toString

        public String toString()
        returns a string representation of the object
        Overrides:
        toString in class Object
        Returns:
        a string representation of the object

      • main

        public static void main​(String[] args)
                 throws Exception
        for running the class from commandline, prints the generated data to stdout
        Parameters:
        args - the commandline parameters
        Throws:
        Exception - if something goes wrong