Examples of weka.core.Matrix

weka.core.Matrix
Class for performing operations on a matrix of floating-point values.
Deprecated: Uses internally the code of the sub-package weka.core.matrix - only for backwards compatibility. @author Gabi Schmidberger (gabi@cs.waikato.ac.nz) @author Yong Wang (yongwang@cs.waikato.ac.nz) @author Eibe Frank (eibe@cs.waikato.ac.nz) @author Len Trigg (eibe@cs.waikato.ac.nz) @author Fracpete (fracpete at waikato dot ac dot nz) @version $Revision: 1.25 $ @deprecated Use weka.core.matrix.Matrix instead - only forbackwards compatibility.

  numAttributes++;
      }
    }


    // Check whether there are still attributes left
    Matrix independent = null, dependent = null;
    double[] weights = null;
    if (numAttributes > 0) {
      independent = new Matrix(m_TransformedData.numInstances(), 
             numAttributes);
      dependent = new Matrix(m_TransformedData.numInstances(), 1);
      for (int i = 0; i < m_TransformedData.numInstances(); i ++) {
  Instance inst = m_TransformedData.instance(i);
  int column = 0;
  for (int j = 0; j < m_TransformedData.numAttributes(); j++) {
    if (j == m_ClassIndex) {
      dependent.setElement(i, 0, inst.classValue());
    } else {
      if (selectedAttributes[j]) {
        double value = inst.value(j) - m_Means[j];
        
        // We only need to do this if we want to
        // scale the input
        if (!m_checksTurnedOff) {
    value /= m_StdDevs[j];
        }
        independent.setElement(i, column, value);
        column++;
      }
    }
  }
      }
      
      // Grab instance weights
      weights = new double [m_TransformedData.numInstances()];
      for (int i = 0; i < weights.length; i++) {
  weights[i] = m_TransformedData.instance(i).weight();
      }
    }


    // Compute coefficients (note that we have to treat the
    // intercept separately so that it doesn't get affected
    // by the ridge constant.)
    double[] coefficients = new double[numAttributes + 1];
    if (numAttributes > 0) {
      double[] coeffsWithoutIntercept  =
  independent.regression(dependent, weights, m_Ridge);
      System.arraycopy(coeffsWithoutIntercept, 0, coefficients, 0,
           numAttributes);
    }
    coefficients[numAttributes] = m_ClassMean;

View Full Code Here

  numAttributes++;
      }
    }


    // Check whether there are still attributes left
    Matrix independent = null, dependent = null;
    double[] weights = null;
    if (numAttributes > 0) {
      independent = new Matrix(m_TransformedData.numInstances(), 
             numAttributes);
      dependent = new Matrix(m_TransformedData.numInstances(), 1);
      for (int i = 0; i < m_TransformedData.numInstances(); i ++) {
  Instance inst = m_TransformedData.instance(i);
  int column = 0;
  for (int j = 0; j < m_TransformedData.numAttributes(); j++) {
    if (j == m_ClassIndex) {
      dependent.setElement(i, 0, inst.classValue());
    } else {
      if (selectedAttributes[j]) {
        double value = inst.value(j) - m_Means[j];
        
        // We only need to do this if we want to
        // scale the input
        if (!m_checksTurnedOff) {
    value /= m_StdDevs[j];
        }
        independent.setElement(i, column, value);
        column++;
      }
    }
  }
      }
      
      // Grab instance weights
      weights = new double [m_TransformedData.numInstances()];
      for (int i = 0; i < weights.length; i++) {
  weights[i] = m_TransformedData.instance(i).weight();
      }
    }


    // Compute coefficients (note that we have to treat the
    // intercept separately so that it doesn't get affected
    // by the ridge constant.)
    double[] coefficients = new double[numAttributes + 1];
    if (numAttributes > 0) {
      double[] coeffsWithoutIntercept  =
  independent.regression(dependent, weights, m_Ridge);
      System.arraycopy(coeffsWithoutIntercept, 0, coefficients, 0,
           numAttributes);
    }
    coefficients[numAttributes] = m_ClassMean;

View Full Code Here

    fillCovariance();


    double [] d = new double[m_numAttribs]; 
    double [][] v = new double[m_numAttribs][m_numAttribs];


    Matrix corr = new Matrix(m_correlation);
    corr.eigenvalueDecomposition(v, d);
    m_eigenvectors = (double [][])v.clone();
    m_eigenvalues = (double [])d.clone();
    
    /*for (int i = 0; i < m_numAttribs; i++) {
      for (int j = 0; j < m_numAttribs; j++) {

View Full Code Here

    fillCorrelation();


    double [] d = new double[m_numAttribs]; 
    double [][] v = new double[m_numAttribs][m_numAttribs];


    Matrix corr = new Matrix(m_correlation);
    corr.eigenvalueDecomposition(v, d);
    m_eigenvectors = (double [][])v.clone();
    m_eigenvalues = (double [])d.clone();


    // any eigenvalues less than 0 are not worth anything --- change to 0
    for (int i = 0; i < m_eigenvalues.length; i++) {

View Full Code Here

  numAttributes++;
      }
    }


    // Check whether there are still attributes left
    Matrix independent = null, dependent = null;
    double[] weights = null;
    if (numAttributes > 0) {
      independent = new Matrix(m_TransformedData.numInstances(), 
             numAttributes);
      dependent = new Matrix(m_TransformedData.numInstances(), 1);
      for (int i = 0; i < m_TransformedData.numInstances(); i ++) {
  Instance inst = m_TransformedData.instance(i);
  int column = 0;
  for (int j = 0; j < m_TransformedData.numAttributes(); j++) {
    if (j == m_ClassIndex) {
      dependent.setElement(i, 0, inst.classValue());
    } else {
      if (selectedAttributes[j]) {
        double value = inst.value(j) - m_Means[j];
        
        // We only need to do this if we want to
        // scale the input
        if (!m_checksTurnedOff) {
    value /= m_StdDevs[j];
        }
        independent.setElement(i, column, value);
        column++;
      }
    }
  }
      }
      
      // Grab instance weights
      weights = new double [m_TransformedData.numInstances()];
      for (int i = 0; i < weights.length; i++) {
  weights[i] = m_TransformedData.instance(i).weight();
      }
    }


    // Compute coefficients (note that we have to treat the
    // intercept separately so that it doesn't get affected
    // by the ridge constant.)
    double[] coefficients = new double[numAttributes + 1];
    if (numAttributes > 0) {
      double[] coeffsWithoutIntercept  =
  independent.regression(dependent, weights, m_Ridge);
      System.arraycopy(coeffsWithoutIntercept, 0, coefficients, 0,
           numAttributes);
    }
    coefficients[numAttributes] = m_ClassMean;

View Full Code Here

  numAttributes++;
      }
    }


    // Check whether there are still attributes left
    Matrix independent = null, dependent = null;
    double[] weights = null;
    if (numAttributes > 0) {
      independent = new Matrix(m_TransformedData.numInstances(), 
             numAttributes);
      dependent = new Matrix(m_TransformedData.numInstances(), 1);
      for (int i = 0; i < m_TransformedData.numInstances(); i ++) {
  Instance inst = m_TransformedData.instance(i);
  int column = 0;
  for (int j = 0; j < m_TransformedData.numAttributes(); j++) {
    if (j == m_ClassIndex) {
      dependent.setElement(i, 0, inst.classValue());
    } else {
      if (selectedAttributes[j]) {
        double value = inst.value(j) - m_Means[j];
        
        // We only need to do this if we want to
        // scale the input
        if (!m_checksTurnedOff) {
    value /= m_StdDevs[j];
        }
        independent.setElement(i, column, value);
        column++;
      }
    }
  }
      }
      
      // Grab instance weights
      weights = new double [m_TransformedData.numInstances()];
      for (int i = 0; i < weights.length; i++) {
  weights[i] = m_TransformedData.instance(i).weight();
      }
    }


    // Compute coefficients (note that we have to treat the
    // intercept separately so that it doesn't get affected
    // by the ridge constant.)
    double[] coefficients = new double[numAttributes + 1];
    if (numAttributes > 0) {
      double[] coeffsWithoutIntercept  =
  independent.regression(dependent, weights, m_Ridge);
      System.arraycopy(coeffsWithoutIntercept, 0, coefficients, 0,
           numAttributes);
    }
    coefficients[numAttributes] = m_ClassMean;

View Full Code Here

    fillCovariance();


    double [] d = new double[m_numAttribs]; 
    double [][] v = new double[m_numAttribs][m_numAttribs];


    Matrix corr = new Matrix(m_correlation);
    corr.eigenvalueDecomposition(v, d);
    m_eigenvectors = (double [][])v.clone();
    m_eigenvalues = (double [])d.clone();
    
    /*for (int i = 0; i < m_numAttribs; i++) {
      for (int j = 0; j < m_numAttribs; j++) {

View Full Code Here

TOP

Related Classes of weka.core.Matrix

weka.attributeSelection.PrincipalComponents

weka.classifiers.functions.LinearRegression

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.