LSLfilter

class LSLfilter : public tsa::AlgoBase

Algorithm for the filter phase of the Adaptive Least Squares Lattice.

Setters

void SetData(Dmatrix &Data, double scale)

Operations

inline void operator<<(SeqViewDouble &Data)

to be used only when data are contiguos (offline analysis)

Declaration of execute operation

Parameters:

  • InputData – Matrix containing Time Series

  • WhitenedData – Matrix containing the WhitenedData

inline void operator>>(SeqViewDouble &outdata)
inline void operator()(SeqViewDouble &Data, SeqViewDouble &outdata)

for online process

Parameters:

  • Data –

  • outdata –

Getters

void GetData(Dmatrix &DWOutput)
inline Dmatrix *GetWhitenedMatrix()
Returns:

the whitened buffer

int GetDataNeeded()

Get the number of data needed in order to be able to call GetData successfully. If the returned value is less or equal than zero no data are needed.

Returns:

the needed data

inline unsigned int GetOrder()
inline double GetParcorForward(unsigned int j)
inline double GetParcorBackward(unsigned int j)
inline double GetEpf(unsigned int j)
inline double GetErrorForward(unsigned int j)
inline double GetErrorBackward(unsigned int j)
inline double GetSigma(unsigned int j)

Public Functions

LSLfilter(LSLLearning &LSL, double lambda = 1.0, unsigned int size = 1, bool Norm = false)

Constructor

Parameters:

  • Order – Order of the filter

  • sigma – guessed value for the initial sigma

  • Lwsp – Lenght of workspace

  • lambda – forgetting factor

~LSLfilter()

Destructor

inline void Load(const char *filename, const char *fmt = "txt")
inline void Save(const char *filename, const char *fmt = "txt")
inline void xml_serialize(eternity::xml_archive &xml, const char *p)