Module net.finmath.lib
Package net.finmath.marketdata.calibration

Class Solver

java.lang.Object
net.finmath.marketdata.calibration.Solver
public class Solver extends Object
Generates a calibrated model for a given set of calibrationProducts with respect to given CurveFromInterpolationPointss. The model and the curve are assumed to be immutable, i.e., the solver will return a calibrate clone, containing clones for every curve which is part of the set of curves to be calibrated. The calibration is performed as a multi-threaded global optimization. I will greatly profit from a multi-core architecture.
Version:
1.0
Author:
Christian Fries

  • Constructor Details

    • Solver

      public Solver(AnalyticModel model, Vector<AnalyticProduct> calibrationProducts, List<Double> calibrationTargetValues, ParameterTransformation parameterTransformation, double evaluationTime, OptimizerFactory optimizerFactory)
      Generate a solver for the given parameter objects (independents) and objective functions (dependents).
      Parameters:
      model - The model from which a calibrated clone should be created.
      calibrationProducts - The objective functions.
      calibrationTargetValues - Array of target values for the objective functions.
      parameterTransformation - A parameter transformation, if any, otherwise null.
      evaluationTime - Evaluation time applied to the calibration products.
      optimizerFactory - A factory providing the optimizer (for the given objective function)

    • Solver

      public Solver(AnalyticModel model, Vector<AnalyticProduct> calibrationProducts, List<Double> calibrationTargetValues, ParameterTransformation parameterTransformation, double evaluationTime, double calibrationAccuracy)
      Generate a solver for the given parameter objects (independents) and objective functions (dependents).
      Parameters:
      model - The model from which a calibrated clone should be created.
      calibrationProducts - The objective functions.
      calibrationTargetValues - Array of target values for the objective functions.
      parameterTransformation - A parameter transformation, if any, otherwise null.
      evaluationTime - Evaluation time applied to the calibration products.
      calibrationAccuracy - The error tolerance of the solver.

    • Solver

      public Solver(AnalyticModel model, Vector<AnalyticProduct> calibrationProducts, List<Double> calibrationTargetValues, double evaluationTime, double calibrationAccuracy)
      Generate a solver for the given parameter objects (independents) and objective functions (dependents).
      Parameters:
      model - The model from which a calibrated clone should be created.
      calibrationProducts - The objective functions.
      calibrationTargetValues - Array of target values for the objective functions.
      evaluationTime - Evaluation time applied to the calibration products.
      calibrationAccuracy - The error tolerance of the solver.

    • Solver

      public Solver(AnalyticModel model, Vector<AnalyticProduct> calibrationProducts, double evaluationTime, double calibrationAccuracy)
      Generate a solver for the given parameter objects (independents) and objective functions (dependents).
      Parameters:
      model - The model from which a calibrated clone should be created.
      calibrationProducts - The objective functions.
      evaluationTime - Evaluation time applied to the calibration products.
      calibrationAccuracy - The error tolerance of the solver.

    • Solver

      public Solver(AnalyticModel model, Vector<AnalyticProduct> calibrationProducts)
      Generate a solver for the given parameter objects (independents) and objective functions (dependents).
      Parameters:
      model - The model from which a calibrated clone should be created.
      calibrationProducts - The objective functions.

  • Method Details

    • getCalibratedModel

      public AnalyticModel getCalibratedModel(Set<ParameterObject> objectsToCalibrate) throws SolverException
      Find the model such that the equation objectiveFunctions.getValue(model) = 0 holds.
      Parameters:
      objectsToCalibrate - The set of parameterized objects to calibrate.
      Returns:
      A reference to a calibrated clone of the given model.
      Throws:
      SolverException - Thrown if the underlying optimizer does not find a solution.

    • getIterations

      public int getIterations()
      Returns the number of iterations required in the last solver step.
      Returns:
      The number of iterations required.

    • getAccuracy

      public double getAccuracy()
      Returns the accuracy achieved in the last solver run.
      Returns:
      The accuracy achieved in the last solver run.