net.finmath.montecarlo.process

Class LinearInterpolatedTimeDiscreteProcess

java.lang.Object

net.finmath.montecarlo.process.LinearInterpolatedTimeDiscreteProcess

All Implemented Interfaces:

ProcessInterface

public class LinearInterpolatedTimeDiscreteProcess
extends Object
implements ProcessInterface

A linear interpolated time discrete process, that is, given a collection of tuples (Double, RandomVariable) representing realizations \( X(t_{i}) \) this class implements the ProcessInterface and creates a stochastic process \( t \mapsto X(t) \) where \[ X(t) = \frac{t_{i+1} - t}{t_{i+1}-t_{i}} X(t_{i}) + \frac{t - t_{i}}{t_{i+1}-t_{i}} X(t_{i+1}) \] with \( t_{i} \leq t \leq t_{i+1} \). Note: this is the interpolation scheme used in the convergence of the Euler-Maruyama scheme.

Author:

Christian Fries

Constructor Summary

Constructors

Constructor and Description
LinearInterpolatedTimeDiscreteProcess(Map<Double,RandomVariableInterface> realizations) Create a time discrete process by linear interpolation of random variables.

Method Summary

Modifier and Type	Method and Description
LinearInterpolatedTimeDiscreteProcess	add(LinearInterpolatedTimeDiscreteProcess process) Create a new linear interpolated time discrete process by using the time discretization of this process and the sum of this process and the given one as its values.
LinearInterpolatedTimeDiscreteProcess	apply(UnivariateFunction function) Create a new process consisting of the interpolation of the random variables obtained by applying the given function to this process discrete set of random variables.
ProcessInterface	clone() Create and return a clone of this process.
RandomVariableInterface	getMonteCarloWeights(int timeIndex) This method returns the weights of a weighted Monte Carlo method (the probability density).
int	getNumberOfComponents()
RandomVariableInterface	getProcessValue(double time, int component) Returns the (possibly interpolated) value of this stochastic process at a given time \( t \).
RandomVariableInterface	getProcessValue(int timeIndex, int component) This method returns the realization of a component of the process for a given time index.
double	getTime(int timeIndex)
TimeDiscretizationInterface	getTimeDiscretization()
int	getTimeIndex(double time) Returns the time index for a given simulation time.

Methods inherited from class java.lang.Object

equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

LinearInterpolatedTimeDiscreteProcess

public LinearInterpolatedTimeDiscreteProcess(Map<Double,RandomVariableInterface> realizations)

Create a time discrete process by linear interpolation of random variables.

Parameters:

realizations - Given map from time to random variable. The map must not be modified.

Method Detail

add

public LinearInterpolatedTimeDiscreteProcess add(LinearInterpolatedTimeDiscreteProcess process)
                                          throws CalculationException

Create a new linear interpolated time discrete process by using the time discretization of this process and the sum of this process and the given one as its values.

Parameters:

process - A given process.

Returns:

A new process representing the of this and the given process.

Throws:

CalculationException - Thrown if the given process fails to evaluate at a certain time point.

apply

public LinearInterpolatedTimeDiscreteProcess apply(UnivariateFunction function)

Create a new process consisting of the interpolation of the random variables obtained by applying the given function to this process discrete set of random variables. That is \( t \mapsto Y(t) \) where \[ Y(t) = \frac{t_{i+1} - t}{t_{i+1}-t_{i}} f(X(t_{i})) + \frac{t - t_{i}}{t_{i+1}-t_{i}} f(X(t_{i+1})) \] with \( t_{i} \leq t \leq t_{i+1} \) and a given function \( f \).

Parameters:

function - The function \( f \), a univariate function.

Returns:

A new process consisting of the interpolation of the random variables obtained by applying the given function to this process discrete set of random variables.

getProcessValue

public RandomVariableInterface getProcessValue(double time,
                                               int component)
                                        throws CalculationException

Returns the (possibly interpolated) value of this stochastic process at a given time \( t \).

Parameters:

time - The time \( t \).

component - The component to be returned (if this is a vector valued process), otherwise 0.

Returns:

The random variable \( X(t) \).

Throws:

CalculationException - Thrown if this process fails to evaluate at the given time point.

getProcessValue

public RandomVariableInterface getProcessValue(int timeIndex,
                                               int component)
                                        throws CalculationException

Description copied from interface: ProcessInterface

This method returns the realization of a component of the process for a given time index.

Specified by:

getProcessValue in interface ProcessInterface

Parameters:

timeIndex - Time index at which the process should be observed

component - Component index of the process

Returns:

The process component realizations (given as RandomVariable)

Throws:

CalculationException - Thrown if the valuation fails, specific cause may be available via the cause() method.

getMonteCarloWeights

public RandomVariableInterface getMonteCarloWeights(int timeIndex)
                                             throws CalculationException

Description copied from interface: ProcessInterface

This method returns the weights of a weighted Monte Carlo method (the probability density).

Specified by:

getMonteCarloWeights in interface ProcessInterface

Parameters:

timeIndex - Time index at which the process should be observed

Returns:

A vector of positive weights which sums up to one

Throws:

CalculationException - Thrown if the valuation fails, specific cause may be available via the cause() method.

getNumberOfComponents

public int getNumberOfComponents()

Specified by:

getNumberOfComponents in interface ProcessInterface

Returns:

Returns the numberOfComponents.

getTimeDiscretization

public TimeDiscretizationInterface getTimeDiscretization()

Specified by:

getTimeDiscretization in interface ProcessInterface

Returns:

Returns the timeDiscretization.

getTime

public double getTime(int timeIndex)

Specified by:

getTime in interface ProcessInterface

Parameters:

timeIndex - Time index.

Returns:

Returns the time for a given time index.

getTimeIndex

public int getTimeIndex(double time)

Description copied from interface: ProcessInterface

Returns the time index for a given simulation time.

Specified by:

getTimeIndex in interface ProcessInterface

Parameters:

time - The given simulation time.

Returns:

Returns the time index for a given time

clone

public ProcessInterface clone()

Description copied from interface: ProcessInterface

Create and return a clone of this process. The clone is not tied to any model, but has the same process specification, that is, if the model is the same, it would generate the same paths.

Specified by:

clone in interface ProcessInterface

Overrides:

clone in class Object

Returns:

Clone of the process