Class implements. More...

#include <constrainedSmoother.hpp>

Inheritance diagram for julian::ir::ConstrainedSmoother:

Public Member Functions
	ConstrainedSmoother (SmartPointer< SmootherCostFunction > cf, const std::vector< double > &lT, const std::vector< double > &uT, int number_of_iterations)
	Constructor. More...

	ConstrainedSmoother (SmartPointer< SmootherCostFunction > cf, double tolerance, int number_of_iterations)
	Constructor. More...

virtual void	calculate (const std::vector< SmartPointer< BuildingBlock > > &instruments, const CurveSettings &settings, SmartPointer< ir::Curve > &discounting_curve, SmartPointer< ir::Curve > &projection_curve)
	Estimates the curve. More...

virtual std::vector< double >	getDF () const
	Returns estimated DF. More...

virtual std::vector< Date >	getDates () const
	Returns grid dates of the curve. More...

virtual ConstrainedSmoother *	clone () const
	Virtual Copy Constructor. More...

virtual std::string	info () const
	Returns estimator's name. More...

Public Member Functions inherited from julian::ir::Estimator
	Estimator ()
	default constructor More...

virtual	~Estimator ()
	destructor More...

Private Member Functions
arma::mat	calculateJacobian (SmartPointer< ir::Curve >, SmartPointer< ir::Curve >, InterpolatedCurve, const std::vector< SmartPointer< BuildingBlock > > &)
	Method used by SQP algorithm to calculate Jacobian for constrains' approximation. More...

arma::mat	calculateParRates (const SmartPointer< ir::Curve > &, const SmartPointer< ir::Curve > &, const SmartPointer< ir::Curve > &, const std::vector< SmartPointer< BuildingBlock > > &)
	Calculates the vector of par rates for instruments. More...

Private Attributes
SmartPointer< SmootherCostFunction >	cost_function_
	Cost Function. More...

std::vector< double >	lower_tolerance_
	Lower tolerances for par rate generated by curve. More...

std::vector< double >	upper_tolerance_
	Upper tolerances for par rate generated by curve. More...

int	number_of_iterations_
	Number of iterations of Sequential Quadratic Programming. More...

double	tolerance_
	Tolerance for par rate generated by curve. More...

std::vector< double >	DFs_
	Vector holding the Discount Factors being result of estimation. More...

std::vector< Date >	dates_
	Grid dates of resulting curve. More...

Detailed Description

Class implements.

The idea of this estimator is to find the minimum of function (see SmootherCostFunction)

$Cost_{f} = a \int^{tN}_{t0} \Big(\frac{df(x)}{dx}\Big)^2 dx + b \int^{tN}_{t0} \Big(\frac{df^2(x)}{dx^2}\Big)^2 dx$

with following constrains:

$q_i -tolerante_{lower} < r_i < q_i + tolerance_{upper}$

where:

To document:: Describe the algorithm more precisly

Examples:: constrainedBootstrapperExample.cpp.

Constructor & Destructor Documentation

julian::ir::ConstrainedSmoother::ConstrainedSmoother	(	SmartPointer< SmootherCostFunction >	cf,
		const std::vector< double > &	lT,
		const std::vector< double > &	uT,
		int	number_of_iterations
	)

Constructor.

Parameters

cf	Cost function
lT	lower tolerance vector
uT	upper tolerance vector
number_of_iterations	Number of iteration of Sequential Quadratic Programming algorithm

julian::ir::ConstrainedSmoother::ConstrainedSmoother	(	SmartPointer< SmootherCostFunction >	cf,
		double	tolerance,
		int	number_of_iterations
	)

Constructor.

Parameters

cf	Cost function
tolerance	Maximal deviation from market quote
number_of_iterations	Number of iteration of Sequential Quadratic Programming algorithm

Member Function Documentation

void julian::ir::ConstrainedSmoother::calculate	(	const std::vector< SmartPointer< BuildingBlock > > &	instruments,
		const CurveSettings &	settings,
		SmartPointer< ir::Curve > &	discounting_curve,
		SmartPointer< ir::Curve > &	projection_curve
	)

virtual

Estimates the curve.

Algorithm uses the SQP algorithm to estimate the curve.

Implements julian::ir::Estimator.

arma::mat julian::ir::ConstrainedSmoother::calculateJacobian	(	SmartPointer< ir::Curve >	disc,
		SmartPointer< ir::Curve >	proj,
		InterpolatedCurve	smoothed,
		const std::vector< SmartPointer< BuildingBlock > > &	instruments
	)

private

Method used by SQP algorithm to calculate Jacobian for constrains' approximation.

arma::mat julian::ir::ConstrainedSmoother::calculateParRates	(	const SmartPointer< ir::Curve > &	disc,
		const SmartPointer< ir::Curve > &	proj,
		const SmartPointer< ir::Curve > &	smoothed,
		const std::vector< SmartPointer< BuildingBlock > > &	instruments
	)

private

Calculates the vector of par rates for instruments.

ConstrainedSmoother * julian::ir::ConstrainedSmoother::clone ( ) const

virtual

Virtual Copy Constructor.

Implements julian::ir::Estimator.

std::vector< Date > julian::ir::ConstrainedSmoother::getDates ( ) const

virtual

Returns grid dates of the curve.

Implements julian::ir::Estimator.

std::vector< double > julian::ir::ConstrainedSmoother::getDF ( ) const

virtual

Returns estimated DF.

Implements julian::ir::Estimator.

std::string julian::ir::ConstrainedSmoother::info ( ) const

virtual

Returns estimator's name.

Implements julian::ir::Estimator.

Member Data Documentation

SmartPointer<SmootherCostFunction> julian::ir::ConstrainedSmoother::cost_function_

private

Cost Function.

std::vector<Date> julian::ir::ConstrainedSmoother::dates_

private

Grid dates of resulting curve.

std::vector<double> julian::ir::ConstrainedSmoother::DFs_

private

Vector holding the Discount Factors being result of estimation.

std::vector<double> julian::ir::ConstrainedSmoother::lower_tolerance_

private

Lower tolerances for par rate generated by curve.

The tolerances are set for each instrument separately. The n-th number in vector is applied to n-th instrument

int julian::ir::ConstrainedSmoother::number_of_iterations_

private

Number of iterations of Sequential Quadratic Programming.

double julian::ir::ConstrainedSmoother::tolerance_

private

Tolerance for par rate generated by curve.

Applied to all instruments. The par rate implied from the curve will be in range (quote - tolerance, quote + tolerance)

std::vector<double> julian::ir::ConstrainedSmoother::upper_tolerance_

private

Upper tolerances for par rate generated by curve.

The documentation for this class was generated from the following files:

C:/Unix/home/OEM/jULIAN/src/marketData/interestRateCurves/estimators/constrainedSmoother.hpp
C:/Unix/home/OEM/jULIAN/src/marketData/interestRateCurves/estimators/constrainedSmoother.cpp

Public Member Functions

Private Member Functions

Private Attributes

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation