DocumentazioneSezioni
MQL5 RiferimentoMetodi Matriciali e VettorialiOpenBLASSingular Spectrum AnalysisSingularSpectrumAnalysisReconstructComponents 

SingularSpectrumAnalysisReconstructComponents

A method function for calculating reconstructed components of the input time series and their contributions.

Calculations for vector<double> and matrix<double> types

bool  vector::SingularSpectrumAnalysisReconstructComponents(
   ulong    window_length,      // window size for constructing the trajectory matrix
   matrix&  components,         // matrix of reconstructed components
   vector&  contributions       // vector of component contributions to the input series
   );

Calculations for vector<complex> and matrix<complex> types

bool  vector::SingularSpectrumAnalysisReconstructComponents(
   ulong    window_length,      // window size for constructing the trajectory matrix
   matrixc& components,         // matrix of reconstructed components
   vectorc& contributions       // vector of component contributions to the input series
   );

Parameters

window_length

[in]  Window size for constructing the trajectory matrix, the number of components the input time series should be decomposed into.

components

[out]  A matrix of reconstructed components, where each column describes a separate component.

contributions

[out]  Vector of component contributions to the input series (eigenvalues of the covariance matrix of the input time series).

Return Value

The function returns 'true' on success or 'false' if an error occurs.

Note

The window_length parameter value should be less than the size of the input time series. To construct a full-fledged trajectory matrix, the optimal size is considered to be approximately equal to half the size of the input time series.