All Packages  Class Hierarchy  This Package  Previous  Next  Index  WEKA's home

Interface weka.gui.boundaryvisualizer.DataGenerator


public interface DataGenerator
Interface to something that can generate new instances based on a set of input instances

Since:
1.0
Version:
$Revision: 1.3 $
Author:
Mark Hall

Method Index

 o buildGenerator(Instances)
Build the data generator
 o generateInstances(int[])
Generate an instance.
 o getNumGeneratingModels()
Returns the number of generating models used by this DataGenerator
 o getWeights()
Get weights
 o setSeed(int)
Set a seed for random number generation (if needed).
 o setWeightingDimensions(boolean[])
Set the dimensions to be used in computing a weight for each instance generated
 o setWeightingValues(double[])
Set the values of the dimensions (chosen via setWeightingDimensions) to be used when computing instance weights

Method Detail

 o buildGenerator
public void buildGenerator(Instances inputInstances) throws java.lang.Exception
          Build the data generator
Parameters:
inputInstances - Instances to build the generator from
Throws:
java.lang.Exception - if an error occurs
 o generateInstances
public double[][] generateInstances(int indices[]) throws java.lang.Exception
          Generate an instance. Should return a new Instance object
Returns:
an Instance value
Throws:
java.lang.Exception - if an error occurs
 o getWeights
public double[] getWeights() throws java.lang.Exception
          Get weights
 o setWeightingDimensions
public void setWeightingDimensions(boolean dimensions[])
          Set the dimensions to be used in computing a weight for each instance generated
Parameters:
dimensions - an array of booleans specifying the dimensions to be used when computing instance weights
 o setWeightingValues
public void setWeightingValues(double vals[])
          Set the values of the dimensions (chosen via setWeightingDimensions) to be used when computing instance weights
Parameters:
vals - a double[] value
 o getNumGeneratingModels
public int getNumGeneratingModels()
          Returns the number of generating models used by this DataGenerator
Returns:
an int value
 o setSeed
public void setSeed(int seed)
          Set a seed for random number generation (if needed).
Parameters:
seed - an int value

All Packages  Class Hierarchy  This Package  Previous  Next  Index  WEKA's home