IT++ Logo Newcom Logo
Main Page | Modules | Class Hierarchy | Alphabetical List | Class List | File List | Class Members | File Members | Related Pages | Search for 

itpp::Fast_ICA Class Reference
[Fast Independent Component Analysis]

Fast_ICA Fast Independent Component Analysis (Fast ICA). More...

#include <itpp/base/fastica.h>

List of all members.

Public Member Functions

 Fast_ICA (mat ma_mixed_sig)
 Constructor.
void separate (void)
 Explicit launch of main FastICA function.
void set_approach (int in_approach)
 Set approach : FICA_APPROACH_DEFL or FICA_APPROACH_SYMM (default).
void set_nrof_independent_components (int in_nrIC)
 Set number of independent components to separate.
void set_non_linearity (int in_g)
 Set non-linearity.
void set_fine_tune (bool in_finetune)
 Set fine tuning.
void set_a1 (double fl_a1)
 Set $a_1$ parameter.
void set_a2 (double fl_a2)
 Set $a_2$ parameter.
void set_mu (double fl_mu)
 Set $\mu$ parameter.
void set_epsilon (double fl_epsilon)
 Set convergence parameter $\epsilon$.
void set_sample_size (double fl_sampleSize)
 Set sample size.
void set_stabilization (bool in_stabilization)
 Set stabilization mode true or off.
void set_max_num_iterations (int in_maxNumIterations)
 Set maximum number of iterations.
void set_max_fine_tune (int in_maxFineTune)
 Set maximum number of iterations for fine tuning.
void set_first_eig (int in_firstEig)
 Set first eigenvalue index to take into account.
void set_last_eig (int in_lastEig)
 Set last eigenvalue index to take into account.
void set_pca_only (bool in_PCAonly)
 If true, only perform Principal Component Analysis (default = false).
void set_init_guess (mat ma_initGuess)
 Set initial guess matrix instead of random (default).
mat get_mixing_matrix ()
 Get mixing matrix.
mat get_separating_matrix ()
 Get separating matrix.
mat get_independent_components ()
 Get separated signals.
int get_nrof_independent_components ()
 Get number of independent components.
mat get_principal_eigenvectors ()
 Get nrIC first columns of the de-whitening matrix.
mat get_whitening_matrix ()
 Get the whitening matrix.
mat get_dewhitening_matrix ()
 Get the de-whitening matrix.
mat get_white_sig ()
 Get whitened signals.

Detailed Description

Fast_ICA Fast Independent Component Analysis (Fast ICA).

The software is based upon original FastICA for Matlab from A. Hyvarinen. Fast and Robust Fixed-Point Algorithms for Independent Component Analysis. IEEE Transactions on Neural Networks, 10(3), pp. 626-634, 1999.

Example: 

                FastICA fastica(sources);
                fastica.set_nrof_independent_components(sources.rows());
                fastica.set_non_linearity(  FICA_NONLIN_TANH );
                fastica.set_approach( FICA_APPROACH_DEFL );
                fastica.separate();
                mat ICs = fastica.get_independent_components();

Definition at line 122 of file fastica.h.

Constructor & Destructor Documentation

itpp::Fast_ICA::Fast_ICA (  mat  ma_mixed_sig  ) 
 

Constructor.

Construct a Fast_ICA object with mixed signals to separate.

Parameters:
ma_mixed_sig (Input) Mixed signals to separate

Definition at line 97 of file fastica.cpp.

References itpp::Mat< Num_T >::rows().

Member Function Documentation

void itpp::Fast_ICA::separate (  void   ) 
 

Explicit launch of main FastICA function.

Explicit launch of the Fast_ICA algorithm.

Definition at line 123 of file fastica.cpp.

References itpp::Mat< Num_T >::cols(), itpp::diag(), itpp::Mat< Num_T >::get_col(), itpp::max_index(), itpp::Mat< Num_T >::rows(), itpp::Mat< Num_T >::set_col(), itpp::Vec< Num_T >::size(), and itpp::zeros().

void itpp::Fast_ICA::set_approach (  int  in_approach  ) 
 

Set approach : FICA_APPROACH_DEFL or FICA_APPROACH_SYMM (default).

Set approach to use : FICA_APPROACH_SYMM (symmetric) or FICA_APPROACH_DEFL (deflation). The symmetric approach computes all ICs at a time, whereas the deflation approach computes them one by one.

Parameters:
in_approach (Input) Type of approach to use

Definition at line 169 of file fastica.cpp.

void itpp::Fast_ICA::set_nrof_independent_components (  int  in_nrIC  ) 
 

Set number of independent components to separate.

Set the number of ICs to compute.

Parameters:
in_nrIC (Input) Number of ICs to compute

Definition at line 171 of file fastica.cpp.

void itpp::Fast_ICA::set_non_linearity (  int  in_g  ) 
 

Set non-linearity.

Set non-linearity to use : FICA_NONLIN_POW3 (default), FICA_NONLIN_TANH, FICA_NONLIN_GAUSS, FICA_NONLIN_SKEW

Parameters:
in_g (Input) Non-linearity. Can be selected from FICA_NONLIN_POW3, FICA_NONLIN_TANH, FICA_NONLIN_GAUSS or FICA_NONLIN_SKEW

Definition at line 173 of file fastica.cpp.

void itpp::Fast_ICA::set_fine_tune (  bool  in_finetune  ) 
 

Set fine tuning.

Set fine tuning true or false.

Parameters:
in_finetune (Input) Boolean (true or false)

Definition at line 175 of file fastica.cpp.

void itpp::Fast_ICA::set_a1 (  double  fl_a1  ) 
 

Set $a_1$ parameter.

Set internal parameter $a_1$ of Fast_ICA (See reference paper).

Parameters:
fl_a1 (Input) Parameter $a_1$ from reference paper

Definition at line 177 of file fastica.cpp.

void itpp::Fast_ICA::set_a2 (  double  fl_a2  ) 
 

Set $a_2$ parameter.

Set internal parameter $a_2$ of Fast_ICA (See reference paper).

Parameters:
fl_a2 (Input) Parameter $a_2$ from reference paper

Definition at line 179 of file fastica.cpp.

void itpp::Fast_ICA::set_mu (  double  fl_mu  ) 
 

Set $\mu$ parameter.

Set internal parameter $\mu$ of Fast_ICA (See reference paper).

Parameters:
fl_mu (Input) Parameter $\mu$ from reference paper

Definition at line 181 of file fastica.cpp.

void itpp::Fast_ICA::set_epsilon (  double  fl_epsilon  ) 
 

Set convergence parameter $\epsilon$.

Set $\epsilon$ parameter for convergence precision.

Parameters:
fl_epsilon (Input) $\epsilon$ is convergence precision

Definition at line 183 of file fastica.cpp.

void itpp::Fast_ICA::set_sample_size (  double  fl_sampleSize  ) 
 

Set sample size.

Set the percentage of samples to take into account at every iteration.

Parameters:
fl_sampleSize (Input) Percentage of data to take into account at every iteration

Definition at line 185 of file fastica.cpp.

void itpp::Fast_ICA::set_stabilization (  bool  in_stabilization  ) 
 

Set stabilization mode true or off.

Set stabilization mode.

Parameters:
in_stabilization (Input) Set stabilization true or false

Definition at line 187 of file fastica.cpp.

void itpp::Fast_ICA::set_max_num_iterations (  int  in_maxNumIterations  ) 
 

Set maximum number of iterations.

Set maximum number of iterations for Fast_ICA.

Parameters:
in_maxNumIterations (Input) Maximum number of iterations to go through

Definition at line 189 of file fastica.cpp.

void itpp::Fast_ICA::set_max_fine_tune (  int  in_maxFineTune  ) 
 

Set maximum number of iterations for fine tuning.

Set maximum numberr of iterations for fine tuning.

Parameters:
in_maxFineTune (Input) Maximum number of iterations for fine tuning stage

Definition at line 191 of file fastica.cpp.

void itpp::Fast_ICA::set_first_eig (  int  in_firstEig  ) 
 

Set first eigenvalue index to take into account.

Set first eigenvalue index to take into account.

Parameters:
in_firstEig (Input) First eigenvalue index to take into account

Definition at line 193 of file fastica.cpp.

void itpp::Fast_ICA::set_last_eig (  int  in_lastEig  ) 
 

Set last eigenvalue index to take into account.

Set last eigenvalue index to take into account.

Parameters:
in_lastEig (Input) Last eigenvalue index to take into account

Definition at line 195 of file fastica.cpp.

void itpp::Fast_ICA::set_pca_only (  bool  in_PCAonly  ) 
 

If true, only perform Principal Component Analysis (default = false).

Wether to perform PCA only or PCA+ICA.

Parameters:
in_PCAonly (Input) True = PCA only, false = PCA+ICA (default)

Definition at line 197 of file fastica.cpp.

void itpp::Fast_ICA::set_init_guess (  mat  ma_initGuess  ) 
 

Set initial guess matrix instead of random (default).

Set initial matrix instead of random matrix.

Parameters:
ma_initGuess (Input) Initial guess matrix

Definition at line 199 of file fastica.cpp.

mat itpp::Fast_ICA::get_mixing_matrix (   ) 
 

Get mixing matrix.

Return mixing matrix.

Returns:
Mixing matrix

Definition at line 202 of file fastica.cpp.

References it_warning, and itpp::zeros().

mat itpp::Fast_ICA::get_separating_matrix (   ) 
 

Get separating matrix.

Return separating matrix.

Returns:
Separating matrix

Definition at line 204 of file fastica.cpp.

References it_warning, and itpp::zeros().

mat itpp::Fast_ICA::get_independent_components (   ) 
 

Get separated signals.

Return separated signals (Independent Components).

Returns:
ICs

Definition at line 206 of file fastica.cpp.

References it_warning, and itpp::zeros().

int itpp::Fast_ICA::get_nrof_independent_components (   ) 
 

Get number of independent components.

Return number of ICs.

Returns:
Number of ICs

Definition at line 208 of file fastica.cpp.

mat itpp::Fast_ICA::get_principal_eigenvectors (   ) 
 

Get nrIC first columns of the de-whitening matrix.

Return principal eigenvectors.

Returns:
Principal eigenvectors

Definition at line 210 of file fastica.cpp.

mat itpp::Fast_ICA::get_whitening_matrix (   ) 
 

Get the whitening matrix.

Return whitening matrix.

Returns:
Whitening matrix

Definition at line 212 of file fastica.cpp.

mat itpp::Fast_ICA::get_dewhitening_matrix (   ) 
 

Get the de-whitening matrix.

Return dewhitening matrix.

Returns:
Dewhitening matrix

Definition at line 214 of file fastica.cpp.

mat itpp::Fast_ICA::get_white_sig (   ) 
 

Get whitened signals.

Return whitened signals.

Returns:
Whitened signals

Definition at line 216 of file fastica.cpp.

The documentation for this class was generated from the following files:
SourceForge Logo

Generated on Fri Jan 11 08:17:31 2008 for IT++ by Doxygen 1.3.9.1