WolfeEvolutionVerifierMetrics.java

  1. package org.drip.function.rdtor1descent;

  3. /*
  4.  * -*- mode: java; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*-
  5.  */

  7. /*!
  8.  * Copyright (C) 2020 Lakshmi Krishnamurthy
  9.  * Copyright (C) 2019 Lakshmi Krishnamurthy
  10.  * Copyright (C) 2018 Lakshmi Krishnamurthy
  11.  * Copyright (C) 2017 Lakshmi Krishnamurthy
  12.  * Copyright (C) 2016 Lakshmi Krishnamurthy
  13.  * 
  14.  *  This file is part of DROP, an open-source library targeting analytics/risk, transaction cost analytics,
  15.  *      asset liability management analytics, capital, exposure, and margin analytics, valuation adjustment
  16.  *      analytics, and portfolio construction analytics within and across fixed income, credit, commodity,
  17.  *      equity, FX, and structured products. It also includes auxiliary libraries for algorithm support,
  18.  *      numerical analysis, numerical optimization, spline builder, model validation, statistical learning,
  19.  *      and computational support.
  20.  *  
  21.  *      https://lakshmidrip.github.io/DROP/
  22.  *  
  23.  *  DROP is composed of three modules:
  24.  *  
  25.  *  - DROP Product Core - https://lakshmidrip.github.io/DROP-Product-Core/
  26.  *  - DROP Portfolio Core - https://lakshmidrip.github.io/DROP-Portfolio-Core/
  27.  *  - DROP Computational Core - https://lakshmidrip.github.io/DROP-Computational-Core/
  28.  * 
  29.  *  DROP Product Core implements libraries for the following:
  30.  *  - Fixed Income Analytics
  31.  *  - Loan Analytics
  32.  *  - Transaction Cost Analytics
  33.  * 
  34.  *  DROP Portfolio Core implements libraries for the following:
  35.  *  - Asset Allocation Analytics
  36.  *  - Asset Liability Management Analytics
  37.  *  - Capital Estimation Analytics
  38.  *  - Exposure Analytics
  39.  *  - Margin Analytics
  40.  *  - XVA Analytics
  41.  * 
  42.  *  DROP Computational Core implements libraries for the following:
  43.  *  - Algorithm Support
  44.  *  - Computation Support
  45.  *  - Function Analysis
  46.  *  - Model Validation
  47.  *  - Numerical Analysis
  48.  *  - Numerical Optimizer
  49.  *  - Spline Builder
  50.  *  - Statistical Learning
  51.  * 
  52.  *  Documentation for DROP is Spread Over:
  53.  * 
  54.  *  - Main                     => https://lakshmidrip.github.io/DROP/
  55.  *  - Wiki                     => https://github.com/lakshmiDRIP/DROP/wiki
  56.  *  - GitHub                   => https://github.com/lakshmiDRIP/DROP
  57.  *  - Repo Layout Taxonomy     => https://github.com/lakshmiDRIP/DROP/blob/master/Taxonomy.md
  58.  *  - Javadoc                  => https://lakshmidrip.github.io/DROP/Javadoc/index.html
  59.  *  - Technical Specifications => https://github.com/lakshmiDRIP/DROP/tree/master/Docs/Internal
  60.  *  - Release Versions         => https://lakshmidrip.github.io/DROP/version.html
  61.  *  - Community Credits        => https://lakshmidrip.github.io/DROP/credits.html
  62.  *  - Issues Catalog           => https://github.com/lakshmiDRIP/DROP/issues
  63.  *  - JUnit                    => https://lakshmidrip.github.io/DROP/junit/index.html
  64.  *  - Jacoco                   => https://lakshmidrip.github.io/DROP/jacoco/index.html
  65.  * 
  66.  *  Licensed under the Apache License, Version 2.0 (the "License");
  67.  *      you may not use this file except in compliance with the License.
  68.  *   
  69.  *  You may obtain a copy of the License at
  70.  *      http://www.apache.org/licenses/LICENSE-2.0
  71.  *  
  72.  *  Unless required by applicable law or agreed to in writing, software
  73.  *      distributed under the License is distributed on an "AS IS" BASIS,
  74.  *      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  75.  *  
  76.  *  See the License for the specific language governing permissions and
  77.  *      limitations under the License.
  78.  */

  80. /**
  81.  * <i>WolfeEvolutionVerifierMetrics</i> implements the Wolfe Criterion used for the Inexact Line Search
  82.  * Increment Generation. The References are:
  83.  * <br><br>
  84.  *  <ul>
  85.  *      <li>
  86.  *          Armijo, L. (1966): Minimization of Functions having Lipschitz-Continuous First Partial
  87.  *              Derivatives <i>Pacific Journal of Mathematics</i> <b>16 (1)</b> 1-3
  88.  *      </li>
  89.  *      <li>
  90.  *          Nocedal, J., and S. Wright (1999): <i>Numerical Optimization</i> <b>Wiley</b>
  91.  *      </li>
  92.  *      <li>
  93.  *          Wolfe, P. (1969): Convergence Conditions for Ascent Methods <i>SIAM Review</i> <b>11 (2)</b>
  94.  *              226-235
  95.  *      </li>
  96.  *      <li>
  97.  *          Wolfe, P. (1971): Convergence Conditions for Ascent Methods; II: Some Corrections <i>SIAM
  98.  *              Review</i> <b>13 (2)</b> 185-188
  99.  *      </li>
  100.  *  </ul>
  101.  *
  102.  *  <br><br>
  103.  *  <ul>
  104.  *      <li><b>Module </b> = <a href = "https://github.com/lakshmiDRIP/DROP/tree/master/ComputationalCore.md">Computational Core Module</a></li>
  105.  *      <li><b>Library</b> = <a href = "https://github.com/lakshmiDRIP/DROP/tree/master/NumericalAnalysisLibrary.md">Numerical Analysis Library</a></li>
  106.  *      <li><b>Project</b> = <a href = "https://github.com/lakshmiDRIP/DROP/tree/master/src/main/java/org/drip/function/README.md">R<sup>d</sup> To R<sup>d</sup> Function Analysis</a></li>
  107.  *      <li><b>Package</b> = <a href = "https://github.com/lakshmiDRIP/DROP/tree/master/src/main/java/org/drip/function/rdtor1descent/README.md">R<sup>d</sup> To R<sup>1</sup> Gradient Descent Techniques</a></li>
  108.  *  </ul>
  109.  *
  110.  * @author Lakshmi Krishnamurthy
  111.  */

  113. public class WolfeEvolutionVerifierMetrics
  114.     extends org.drip.function.rdtor1descent.LineEvolutionVerifierMetrics
  115. {
  116.     private boolean _maximizerCheck = false;
  117.     private boolean _strongCurvatureCriterion = false;
  118.     private double[] _nextVariateFunctionJacobian = null;
  119.     private double _armijoParameter = java.lang.Double.NaN;
  120.     private double _curvatureParameter = java.lang.Double.NaN;
  121.     private double _nextVariateFunctionValue = java.lang.Double.NaN;
  122.     private double _currentVariateFunctionValue = java.lang.Double.NaN;

  124.     /**
  125.      * WolfeEvolutionVerifierMetrics Constructor
  126.      * 
  127.      * @param armijoParameter The Armijo Criterion Parameter
  128.      * @param maximizerCheck TRUE - Perform a Check for the Function Maxima
  129.      * @param curvatureParameter The Curvature Criterion Parameter
  130.      * @param strongCurvatureCriterion TRUE - Apply the "Strong" Curvature Criterion
  131.      * @param targetDirection The Target Direction Unit Vector
  132.      * @param currentVariateArray Array of the Current Variate
  133.      * @param stepLength The Incremental Step Length
  134.      * @param currentVariateFunctionValue The Function Value at the Current Variate
  135.      * @param nextVariateFunctionValue The Function Value at the Next Variate
  136.      * @param currentVariateFunctionJacobian The Function Jacobian at the Current Variate
  137.      * @param nextVariateFunctionJacobian The Function Jacobian at the Next Variate
  138.      * 
  139.      * @throws java.lang.Exception Thrown if the Inputs are Invalid
  140.      */

  142.     public WolfeEvolutionVerifierMetrics (
  143.         final double armijoParameter,
  144.         final boolean maximizerCheck,
  145.         final double curvatureParameter,
  146.         final boolean strongCurvatureCriterion,
  147.         final org.drip.function.definition.UnitVector targetDirection,
  148.         final double[] currentVariateArray,
  149.         final double stepLength,
  150.         final double currentVariateFunctionValue,
  151.         final double nextVariateFunctionValue,
  152.         final double[] currentVariateFunctionJacobian,
  153.         final double[] nextVariateFunctionJacobian)
  154.         throws java.lang.Exception
  155.     {
  156.         super (
  157.             targetDirection,
  158.             currentVariateArray,
  159.             stepLength,
  160.             currentVariateFunctionJacobian
  161.         );

  163.         if (!org.drip.numerical.common.NumberUtil.IsValid (_armijoParameter = armijoParameter) ||
  164.             !org.drip.numerical.common.NumberUtil.IsValid (_curvatureParameter = curvatureParameter) ||
  165.             null == (_nextVariateFunctionJacobian = nextVariateFunctionJacobian) ||
  166.             !org.drip.numerical.common.NumberUtil.IsValid (_currentVariateFunctionValue =
  167.                 currentVariateFunctionValue) ||
  168.             !org.drip.numerical.common.NumberUtil.IsValid (_nextVariateFunctionValue =
  169.                 nextVariateFunctionValue) ||
  170.             currentVariateArray.length != _nextVariateFunctionJacobian.length)
  171.         {
  172.             throw new java.lang.Exception ("WolfeEvolutionVerifierMetrics Constructor => Invalid Inputs");
  173.         }

  175.         _maximizerCheck = maximizerCheck;
  176.         _strongCurvatureCriterion = strongCurvatureCriterion;
  177.     }

  179.     /**
  180.      * Retrieve the Armijo Parameter
  181.      * 
  182.      * @return The Armijo Parameter
  183.      */

  185.     public double armijoParameter()
  186.     {
  187.         return _armijoParameter;
  188.     }

  190.     /**
  191.      * Indicate if the Check is for Minimizer/Maximizer
  192.      * 
  193.      * @return TRUE - The Check is for Maximizer
  194.      */

  196.     public boolean maximizerCheck()
  197.     {
  198.         return _maximizerCheck;
  199.     }

  201.     /**
  202.      * Retrieve the Curvature Parameter
  203.      * 
  204.      * @return The Curvature Parameter
  205.      */

  207.     public double curvatureParameter()
  208.     {
  209.         return _curvatureParameter;
  210.     }

  212.     /**
  213.      * Retrieve Whether of not the "Strong" Curvature Criterion needs to be met
  214.      * 
  215.      * @return TRUE - The "Strong" Curvature Criterion needs to be met
  216.      */

  218.     public boolean strongCurvatureCriterion()
  219.     {
  220.         return _strongCurvatureCriterion;
  221.     }

  223.     /**
  224.      * Retrieve the Function Value at the Current Variate
  225.      * 
  226.      * @return The Function Value at the Current Variate
  227.      */

  229.     public double currentVariateFunctionValue()
  230.     {
  231.         return _currentVariateFunctionValue;
  232.     }

  234.     /**
  235.      * Retrieve the Function Value at the Next Variate
  236.      * 
  237.      * @return The Function Value at the Next Variate
  238.      */

  240.     public double nextVariateFunctionValue()
  241.     {
  242.         return _nextVariateFunctionValue;
  243.     }

  245.     /**
  246.      * Retrieve the Function Jacobian at the Next Variate
  247.      * 
  248.      * @return The Function Jacobian at the Next Variate
  249.      */

  251.     public double[] nextVariateFunctionJacobian()
  252.     {
  253.         return _nextVariateFunctionJacobian;
  254.     }

  256.     /**
  257.      * Indicate if the Wolfe Criterion has been met
  258.      * 
  259.      * @return TRUE - The Wolfe Criterion has been met
  260.      */

  262.     public boolean verify()
  263.     {
  264.         double[] targetDirectionVector = targetDirection().component();

  266.         double[] currentVariateFunctionJacobian = currentVariateFunctionJacobian();

  268.         try
  269.         {
  270.             double gradientUpdatedFunctionValue = _currentVariateFunctionValue +
  271.                 _armijoParameter * stepLength() * org.drip.numerical.linearalgebra.Matrix.DotProduct (
  272.                     targetDirectionVector,
  273.                     currentVariateFunctionJacobian
  274.                 );

  276.             if ((_maximizerCheck && _nextVariateFunctionValue < gradientUpdatedFunctionValue) ||
  277.                 (!_maximizerCheck && _nextVariateFunctionValue > gradientUpdatedFunctionValue))
  278.             {
  279.                 return false;
  280.             }

  282.             double nextFunctionIncrement = org.drip.numerical.linearalgebra.Matrix.DotProduct (
  283.                 targetDirectionVector,
  284.                 _nextVariateFunctionJacobian
  285.             );

  287.             double parametrizedCurrentFunctionIncrement = _curvatureParameter *
  288.                 org.drip.numerical.linearalgebra.Matrix.DotProduct (
  289.                     targetDirectionVector,
  290.                     currentVariateFunctionJacobian
  291.                 );

  293.             return _strongCurvatureCriterion ?
  294.                 java.lang.Math.abs (
  295.                     nextFunctionIncrement
  296.                 ) <= java.lang.Math.abs (
  297.                     parametrizedCurrentFunctionIncrement
  298.                 ) : nextFunctionIncrement >= parametrizedCurrentFunctionIncrement;
  299.         }
  300.         catch (java.lang.Exception e)
  301.         {
  302.             e.printStackTrace();
  303.         }

  305.         return false;
  306.     }
  307. }
Created with JaCoCo 0.7.9.201702052155