CSIM: LinearPreprocessor Class Reference

LinearPreprocessor Class Reference

#include <linearpreprocessor.h>

Inheritance diagram for LinearPreprocessor:

List of all members.

Detailed Description

Implementation of a linear transformation of the input. Every row x_i of the input vector is transformed into x_i' = a_i * x_i + b_i.

Public Member Functions

LinearPreprocessor (unsigned int rows=1)
~LinearPreprocessor (void)
int process (const double *S, double *X)
void reset ()
int importRepresentation (const double *rep, int rep_length)
double * exportRepresentation (int *rep_length)
string getFormatDescription ()

Private Attributes

double * linear_coefficients
The parameters of the linear transformation.

Friends

class csimClassInfoDB

Constructor & Destructor Documentation

LinearPreprocessor::LinearPreprocessor ( unsigned int rows = 1 )

Constructs a new linear transformation. Initially the transformation is an identical transformation x'=x.
Parameters:

rows Number of rows in each input vector.

LinearPreprocessor::~LinearPreprocessor ( void )

Frees the memory.

Member Function Documentation

double * LinearPreprocessor::exportRepresentation ( int * rep_length ) [virtual]

Exports the representation of this preprocessor for use in external objects. Format: first number gives the number of rows for input vectors, the following 2*rows elements are in the format [a_1, b_1, a_2, b_2, ..., a_n, b_n], where the transformations are $x_i \cdot a_i + b_i$ .
Parameters:

rep_length Length of the representation vector.

Returns:
A list of parameters that represent the preprocessor.

Warning:
Do not forget to free the memory reserved for the representation!

Implements Preprocessor.

string LinearPreprocessor::getFormatDescription ( ) [virtual]

Returns a textual description of the representation format for import/export - Representation.
Implements Preprocessor.

int LinearPreprocessor::importRepresentation ( const double * rep,

int rep_length

) [virtual]

Imports the data from an externally (e.g. Matlab) trained preprocessor.
Parameters:

rep Representation of the preprocessor as a double vector. Format: first number gives the number of rows for input vectors, the following 2*rows elements are in the format [a_1, b_1, a_2, b_2, ..., a_n, b_n], where the transformations are $x_i \cdot a_i + b_i$ .

rep_length Length of the representation vector.

Returns:
-1 if an error occured, 1 for success.

Implements Preprocessor.

int LinearPreprocessor::process ( const double * S,

double * X

) [virtual]

Preprocess a state representation.
Parameters:

S State of the liquid (= filtered response of the neural microcircuit).

X Target vector where to save the results.

Returns:
-1 if an error occured, 1 for success.

Implements Preprocessor.

void LinearPreprocessor::reset ( ) [virtual]

Resets the information stored within the preprocessor.
Implements Preprocessor.

last modified 07/10/2006