Class SimPLS


  • public class SimPLS
    extends java.lang.Object

    Title: SimPLS

    Description: SimPLS: Simple partial least squares algorithm according: de Jong, S. SIMPLS: An alternative approach to partial least squares regression Chemometrics and Intelligent Laboratory Systems (1993) 18 pp 251-263.

    Copyright: Actelion Ltd., Inc. All Rights Reserved This software is the proprietary information of Actelion Pharmaceuticals, Ltd. Use is subject to license terms.

    Version:
    1.0 21.11.2003 MvK: Start implementation
    Author:
    Modest von Korff
    • Constructor Detail

      • SimPLS

        public SimPLS()
    • Method Detail

      • invLinReg_Yhat

        public static Matrix invLinReg_Yhat​(Matrix B,
                                            Matrix Xtrain,
                                            Matrix Xtest,
                                            Matrix Ytrain)
        Parameters:
        B -
        Xtrain - uncentered matrix. Is used to center Xtest.
        Xtest - uncentered matrix. Will be centered in method.
        Ytrain - uncentered matrix. Is used to de-center Yhat.
        Returns:
      • invLinReg_Yhat

        public static Matrix invLinReg_Yhat​(Matrix BFromCentered,
                                            Matrix Xtest)
        Parameters:
        BFromCentered - must be calculated from centered X and Y data.
        Xtest -
        Returns:
      • getExplainedVarS

        public Matrix getExplainedVarS()
      • getExplainedVarX

        public Matrix getExplainedVarX()
      • getExplainedVarY

        public Matrix getExplainedVarY()
      • simPlsSave

        public void simPlsSave​(Matrix Xtrain,
                               Matrix Ytrain,
                               int facmax)
        No explained variance is calculated. B = null; EX = null; EY = null; ES = null;
        Parameters:
        Xtrain -
        Ytrain -
        facmax -
      • simPlsSaveExplainedVariance

        public void simPlsSaveExplainedVariance​(Matrix Xtrain,
                                                Matrix Ytrain,
                                                int facmax)
        The explained variance is calculated. This is a time consuming process, the calculation of the correlation coefficient is takes longer than the PLS calculation itself.
        Parameters:
        Xtrain -
        Ytrain -
        facmax -
      • getT

        public Matrix getT​(Matrix XPreprocessed)
        Parameters:
        XPreprocessed - has to be preprovcessed with XTrain. I.e. cebntered by the mean values of XTrain.
        Returns:
      • main

        public static void main​(java.lang.String[] args)
      • toString

        public java.lang.String toString​(int iDigits)