name: Rps : public AlgorithmBase

synopsis:

g++ [flags ...] file ...-l /isip/tools/lib/$ISIP_BINARY/lib_algo.a

#include <Rps.h>

Rps(ALGORITHM algorithm = DEF_ALGORITHM, IMPLEMENTATION implementation = DEF_IMPLEMENTATION, long embed_dimension = DEF_EMBED_DIMENSION, long svd_window_size = DEF_SVD_WINDOW_SIZE, float delay = DEF_DELAY, float svd_threshold = DEF_SVD_THRESHOLD);
boolean compute(MatrixFloat& output, const VectorFloat& input, AlgorithmData::COEF_TYPE input_coef_type = DEF_COEF_TYPE, long index = DEF_CHANNEL_INDEX);
boolean compute(MatrixFloat& output, const MatrixFloat& input, AlgorithmData::COEF_TYPE input_coef_type = DEF_COEF_TYPE, long index = DEF_CHANNEL_INDEX);

quick start:

Rps rps0;
Rps rps1(rps0);
VectorFloat input(L"0, 1, 2, 3, 4, 5");
MatrixFloat output;
rps1.setAlgorithm(Rps::SVD_EMBEDDING);
rps1.compute(output, input);
description:

The Rps (Reconstruced Phase Space) class reconstructs a pseudo phase space of a dynamical system from one (scalar) or more (vector) time-series measurements of the system by the process of embedding. Two ways (algorithms) of achieving embedding are TIME_DELAY_EMBEDDING and SVD_EMBEDDING. The details can be found in: Rps class works in FRAME-INTERNAL mode for both the algorithms and for both scalar and vector embedding but ACCUMULATE mode is possible only with scalar embedding presently. It should also be noted that the ouput RPS matrix has a last appended column with all its elements ones except for the last element which is zero. This is useful to mark discontinuities when concatenating Rps matrices.

dependencies:

public constants:

error codes:

protected data:

required public methods:

class-specific public methods:

private methods:

examples:

notes: