NAV::PolynomialRegressor< Scalar > Class Template Reference

Polynomial Curve Fitting. More...

Public Types
enum class	Strategy : uint8_t {   IncrementalLeastSquares ,   LeastSquares ,   HouseholderQR ,   BDCSVD ,   COD ,   COUNT }
	Possible Fit strategies. More...

Public Member Functions
Polynomial< Scalar >	calcPolynomial () const
	Calculates the polynomial.
const ScrollingBuffer< std::pair< Scalar, Scalar > > &	data () const
	Gets the underlying buffer.
bool	empty () const
	Checks if the container has no elements.
	PolynomialRegressor (size_t polynomialDegree, size_t windowSize, Strategy strategy=Strategy::HouseholderQR)
	Constructor.
void	push_back (const Scalar &x, const Scalar &y)
	Add a data point to the polynomial.
void	push_back (const std::pair< Scalar, Scalar > &dataPoint)
	Add a data point to the polynomial.
void	reset ()
	Reset the polynomial coefficients and saved data.
void	setPolynomialDegree (size_t polynomialDegree)
	Set the Polynomial Degree and resets the data.
void	setStrategy (Strategy strategy)
	Set the strategy for the fit and resets the data.
void	setWindowSize (size_t windowSize)
	Sets the amount of points used for the fit (sliding window)
bool	windowSizeReached () const
	Checks if the amount of data points equals the window size.

Private Member Functions
void	pop_front ()
	Removes the first data point from the polynomial fit (sliding window)

Private Attributes
ScrollingBuffer< std::pair< Scalar, Scalar > >	_data
	Values added to the fit.
IncrementalLeastSquares< Scalar >	_incrementalLSQ
	Incremental LSQ Regressor.
size_t	_polyDegree
	Polynomial degree to fit.
Strategy	_strategy
	Strategy to use to fit the polynomial.
size_t	_windowSize
	Amount of points to store.

Detailed Description

template<typename Scalar = double>
class NAV::PolynomialRegressor< Scalar >

Polynomial Curve Fitting.

Template Parameters

Scalar

Data type to store

Member Enumeration Documentation

Strategy

template<typename Scalar = double>

enum class NAV::PolynomialRegressor::Strategy : uint8_t

strong

Possible Fit strategies.

Enumerator
IncrementalLeastSquares	Incremental Least Squares (only polynomials of order <= 2)
LeastSquares	Least Squares (bas if even mildly ill-conditioned)
HouseholderQR	Householder QR decomposition.
BDCSVD	Bidiagonal Divide and Conquer SVD.
COD	Complete Orthogonal Decomposition.
COUNT	Amount of items in the enum.

Constructor & Destructor Documentation

PolynomialRegressor()

template<typename Scalar = double>

NAV::PolynomialRegressor< Scalar >::PolynomialRegressor ( size_t polynomialDegree, size_t windowSize, Strategy strategy = Strategy::HouseholderQR )

inline

Constructor.

Parameters

[in]	polynomialDegree	Degree of the polynomial to fit
[in]	windowSize	Amount of points to use for the fit (sliding window)
[in]	strategy	Strategy to use

Member Function Documentation

push_back() [1/2]

template<typename Scalar = double>

void NAV::PolynomialRegressor< Scalar >::push_back ( const Scalar & x, const Scalar & y )

inline

Add a data point to the polynomial.

Parameters

[in]	x	X Value
[in]	y	Y Value

push_back() [2/2]

template<typename Scalar = double>

void NAV::PolynomialRegressor< Scalar >::push_back ( const std::pair< Scalar, Scalar > & dataPoint )

inline

Add a data point to the polynomial.

Parameters

[in]

dataPoint

Data point

setPolynomialDegree()

template<typename Scalar = double>

void NAV::PolynomialRegressor< Scalar >::setPolynomialDegree ( size_t polynomialDegree )

inline

Set the Polynomial Degree and resets the data.

Parameters

[in]

polynomialDegree

Degree of the polynomial to fit

setStrategy()

template<typename Scalar = double>

void NAV::PolynomialRegressor< Scalar >::setStrategy ( Strategy strategy )

inline

Set the strategy for the fit and resets the data.

Parameters

strategy

Strategy to use for fitting data

setWindowSize()

template<typename Scalar = double>

void NAV::PolynomialRegressor< Scalar >::setWindowSize ( size_t windowSize )

inline

Sets the amount of points used for the fit (sliding window)

Parameters

[in]

windowSize

Amount of points to use for the fit

---

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

• /home/runner/work/INSTINCT/INSTINCT/src/Navigation/Math/PolynomialRegressor.hpp