name: CorrelationIntegral : public AlgorithmBase

synopsis:

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

#include <CorrelationIntegral.h>

CorrelationIntegral(ALGORITHM algorithm = DEF_ALGORITHM, IMPLEMENTATION implementation = DEF_IMPLEMENTATION, SCALING scaling = DEF_SCALING, SPACING_MODE spacing_mode = DEF_SPACING_MODE, float epsilon_min = DEF_EPSILON_MIN, float epsilon_max = DEF_EPSILON_MAX, long epsilon_res = DEF_EPSILON_RES, long theiler_correction = DEF_THEILER_CORRECTION);
boolean eq(const CorrelationIntegral& arg) const;
boolean compute(VectorFloat& output, const MatrixFloat& input, AlgorithmData::COEF_TYPE coef_type = DEF_COEF_TYPE, long channel_index = DEF_CHANNEL_INDEX);
quick start:

MatrixFloat input;
input.assign(3, 3, L"1.0, 2.0, 3.0, 6.0, 5.0, 4.0, 7.0, 8.0, 9.0", Integral::FULL");
VectorFloat output;
CorrelationIntegral corrint;
corrint.compute(output, input);
description:

The CorrelationIntegral class is designed to estimate the correlation integral of an attractor's trajectory. It takes as input the Reconstructed Phase Space matrix of an attractor and outputs a vector of correlation integrals, corresponding to a range of neighborhood radii. A good overview of methods to estimate the correlation integral can be found in: The mathematical descriptions of the algorithm is provided below for reference purposes:


Four types of scaling are supported: LINSCALE, LOGSCALE, LOG2SCALE, LOG10SCALE. The choice of scaling will ensure that the output vector of corresponding correlation integrals is on a Linear or Logarithmic (with appropriate base) scale.

dependencies:

public constants:

error codes:

protected data:

required public methods:

class-specific public methods:

private methods:

examples:

notes: