ar::predictor< Value, Index > Class Template Reference

Simulate an autoregressive model process with an InputIterator interface. More...

#include <ar.hpp>

Public Types
typedef std::input_iterator_tag	iterator_category
typedef Value	value_type
typedef std::ptrdiff_t	difference_type

Public Member Functions
	predictor (Index n=0)
	Singular instance marking prediction index n.
template<class RandomAccessIterator >
	predictor (RandomAccessIterator params_first, RandomAccessIterator params_last)
	Iterate on the process \(x_n + a_1 x_{n - 1} + \dots + a_p x_{n - p} = 0\). More...
template<class RandomAccessIterator , class NoiseGenerator >
	predictor (RandomAccessIterator params_first, RandomAccessIterator params_last, NoiseGenerator generator)
	Iterate on the process \(x_n + a_1 x_{n - 1} + \dots + a_p x_{n - p} = \epsilon_n\) given zero initial conditions. More...
	predictor (const predictor &other)
	Copy constructor.
predictor &	operator= (const predictor &other)
	Assignment operator.
	~predictor ()
	Destructor.
template<class InputIterator >
predictor &	initial_conditions (InputIterator initial_first, const Value x0adjust=0)
	Specify process initial conditions \(x_{n-1}, \dots, x_{n-p}\) where \(p\) is the process order fixed by the constructor. More...
predictor &	operator++ ()
	Prefix increment.
predictor	operator++ (int)
	Postfix increment.
reference	operator* () const
	Obtain the process prediction \(x_n\).
bool	operator== (const predictor &other) const
	Check if two iterators represent the same simulation time.
bool	operator!= (const predictor &other) const
	Check if two iterators represent different simulation times.

Public Attributes
const typedef Value *	pointer
const typedef Value &	reference

Detailed Description

template<typename Value, typename Index = std::size_t>
class ar::predictor< Value, Index >

Simulate an autoregressive model process with an InputIterator interface.

Definition at line 823 of file ar.hpp.

Constructor & Destructor Documentation

predictor() [1/2]

template<typename Value , typename Index = std::size_t>

template<class RandomAccessIterator >

ar::predictor< Value, Index >::predictor ( RandomAccessIterator  params_first, RandomAccessIterator  params_last  )

inline

Iterate on the process \(x_n + a_1 x_{n - 1} + \dots + a_p x_{n - p} = 0\).

Presumably initial_conditions will be used to specify some initial state as otherwise the process is identically zero. The process order \(p\) is set by std::distance(params_first, params_last).

Parameters

params_first	Beginning of the process parameter range starting with \(a_1\).
params_last	End of the process parameter range.

Definition at line 854 of file ar.hpp.

predictor() [2/2]

template<typename Value , typename Index = std::size_t>

template<class RandomAccessIterator , class NoiseGenerator >

ar::predictor< Value, Index >::predictor ( RandomAccessIterator  params_first, RandomAccessIterator  params_last, NoiseGenerator  generator  )

inline

Iterate on the process \(x_n + a_1 x_{n - 1} + \dots + a_p x_{n - p} = \epsilon_n\) given zero initial conditions.

The process order \(p\) is set by std::distance(params_first,params_last). The class std::tr1::variate_generator may be helpful in constructing normally distributed input.

Parameters

params_first	Beginning of the process parameter range starting with \(a_1\).
params_last	End of the process parameter range.
generator	A nullary callback for generating \(\epsilon_n\). For example, a random number generator distributed like \(N\left(0, \sigma^2_\epsilon\right)\).

Definition at line 885 of file ar.hpp.

Member Function Documentation

initial_conditions()

template<typename Value , typename Index = std::size_t>

template<class InputIterator >

predictor& ar::predictor< Value, Index >::initial_conditions ( InputIterator  initial_first, const Value  x0adjust = 0  )

inline

Specify process initial conditions \(x_{n-1}, \dots, x_{n-p}\) where \(p\) is the process order fixed by the constructor.

The simulation index \(n\) is reset to zero and, optionally, \(x_0\) is additively adjusted by x0adjust.

Parameters

initial_first	Beginning of the initial condition range \(x_{n-1}, \dots, x_{n-p}\) which must contain \(p\) values.
x0adjust	An additive adjustment made to \(\epsilon_0\).

Definition at line 951 of file ar.hpp.

---

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

• ar.hpp