OpenMS  2.7.0
Public Types | Public Member Functions | Protected Member Functions | Protected Attributes | List of all members
ContinuousWaveletTransform Class Reference

This class is the base class of the continuous wavelet transformation. More...

#include <OpenMS/TRANSFORMATIONS/RAW2PEAK/ContinuousWaveletTransform.h>

Inheritance diagram for ContinuousWaveletTransform:
[legend]
Collaboration diagram for ContinuousWaveletTransform:
[legend]

Public Types

typedef std::vector< Peak1D >::const_iterator PeakConstIterator
 Raw data const iterator type. More...
 

Public Member Functions

 ContinuousWaveletTransform ()
 Constructor. More...
 
virtual ~ContinuousWaveletTransform ()
 Destructor. More...
 
const std::vector< Peak1D > & getSignal () const
 Non-mutable access to the wavelet transform of the signal. More...
 
std::vector< Peak1D > & getSignal ()
 Mutable access to the wavelet transform of the signal. More...
 
void setSignal (const std::vector< Peak1D > &signal)
 Mutable access to the wavelet transform of the signal. More...
 
const std::vector< double > & getWavelet () const
 Non-mutable access to the wavelet. More...
 
std::vector< double > & getWavelet ()
 Mutable access to the wavelet. More...
 
void setWavelet (const std::vector< double > &wavelet)
 Mutable access to the signal. More...
 
double getScale () const
 
double & getScale ()
 Mutable access to the spacing of raw data. More...
 
void setScale (double scale)
 Mutable access to the spacing of raw data. More...
 
double getSpacing () const
 
double & getSpacing ()
 Mutable access to the spacing of raw data. More...
 
void setSpacing (double spacing)
 Mutable access to the spacing of raw data. More...
 
SignedSize getLeftPaddingIndex () const
 Non-mutable access to the position where the signal starts (in the interval [0,end_left_padding_) are the padded zeros) More...
 
SignedSizegetLeftPaddingIndex ()
 Mutable access to the position where the signal starts. More...
 
void setLeftPaddingIndex (const SignedSize end_left_padding)
 Mutable access to position where the signal starts. More...
 
SignedSize getRightPaddingIndex () const
 Non-mutable access to the position where the signal ends (in the interval (begin_right_padding_,end] are the padded zeros) More...
 
SignedSizegetRightPaddingIndex ()
 Mutable access to the position where the signal starts. More...
 
void setRightPaddingIndex (const SignedSize begin_right_padding)
 Mutable access to position where the signal starts. More...
 
SignedSize getSignalLength () const
 Non-mutable access to signal length [end_left_padding,begin_right_padding]. More...
 
SignedSizegetSignalLength ()
 Mutable access to signal length [end_left_padding,begin_right_padding]. More...
 
void setSignalLength (const SignedSize signal_length)
 Mutable access to signal length [end_left_padding,begin_right_padding]. More...
 
int getSize () const
 Non-mutable access to signal length including padded zeros [0,end]. More...
 
virtual void init (double scale, double spacing)
 Perform possibly necessary preprocessing steps, like tabulating the Wavelet. More...
 
double operator[] (unsigned int i)
 Yields the signal (intensity) at position i. More...
 
double operator[] (unsigned int i) const
 

Protected Member Functions

template<typename InputPeakIterator >
double getInterpolatedValue_ (double x, InputPeakIterator it_left)
 

Protected Attributes

std::vector< Peak1Dsignal_
 The transformed signal. More...
 
std::vector< double > wavelet_
 The pre-tabulated wavelet used for the transform. More...
 
double scale_
 Spacing and scale of the wavelet and length of the signal. More...
 
double spacing_
 
SignedSize signal_length_
 
SignedSize end_left_padding_
 
SignedSize begin_right_padding_
 

Detailed Description

This class is the base class of the continuous wavelet transformation.

Member Typedef Documentation

◆ PeakConstIterator

typedef std::vector<Peak1D>::const_iterator PeakConstIterator

Raw data const iterator type.

Constructor & Destructor Documentation

◆ ContinuousWaveletTransform()

Constructor.

◆ ~ContinuousWaveletTransform()

virtual ~ContinuousWaveletTransform ( )
inlinevirtual

Destructor.

Member Function Documentation

◆ getInterpolatedValue_()

double getInterpolatedValue_ ( double  x,
InputPeakIterator  it_left 
)
inlineprotected

◆ getLeftPaddingIndex() [1/2]

SignedSize& getLeftPaddingIndex ( )
inline

Mutable access to the position where the signal starts.

◆ getLeftPaddingIndex() [2/2]

SignedSize getLeftPaddingIndex ( ) const
inline

Non-mutable access to the position where the signal starts (in the interval [0,end_left_padding_) are the padded zeros)

◆ getRightPaddingIndex() [1/2]

SignedSize& getRightPaddingIndex ( )
inline

Mutable access to the position where the signal starts.

◆ getRightPaddingIndex() [2/2]

SignedSize getRightPaddingIndex ( ) const
inline

Non-mutable access to the position where the signal ends (in the interval (begin_right_padding_,end] are the padded zeros)

◆ getScale() [1/2]

double& getScale ( )
inline

Mutable access to the spacing of raw data.

◆ getScale() [2/2]

double getScale ( ) const
inline

◆ getSignal() [1/2]

std::vector<Peak1D>& getSignal ( )
inline

Mutable access to the wavelet transform of the signal.

◆ getSignal() [2/2]

const std::vector<Peak1D>& getSignal ( ) const
inline

Non-mutable access to the wavelet transform of the signal.

◆ getSignalLength() [1/2]

SignedSize& getSignalLength ( )
inline

Mutable access to signal length [end_left_padding,begin_right_padding].

◆ getSignalLength() [2/2]

SignedSize getSignalLength ( ) const
inline

Non-mutable access to signal length [end_left_padding,begin_right_padding].

◆ getSize()

int getSize ( ) const
inline

Non-mutable access to signal length including padded zeros [0,end].

◆ getSpacing() [1/2]

double& getSpacing ( )
inline

Mutable access to the spacing of raw data.

◆ getSpacing() [2/2]

double getSpacing ( ) const
inline

◆ getWavelet() [1/2]

std::vector<double>& getWavelet ( )
inline

Mutable access to the wavelet.

◆ getWavelet() [2/2]

const std::vector<double>& getWavelet ( ) const
inline

Non-mutable access to the wavelet.

◆ init()

virtual void init ( double  scale,
double  spacing 
)
virtual

Perform possibly necessary preprocessing steps, like tabulating the Wavelet.

Reimplemented in ContinuousWaveletTransformNumIntegration.

◆ operator[]() [1/2]

double operator[] ( unsigned int  i)
inline

Yields the signal (intensity) at position i.

◆ operator[]() [2/2]

double operator[] ( unsigned int  i) const
inline

◆ setLeftPaddingIndex()

void setLeftPaddingIndex ( const SignedSize  end_left_padding)
inline

Mutable access to position where the signal starts.

◆ setRightPaddingIndex()

void setRightPaddingIndex ( const SignedSize  begin_right_padding)
inline

Mutable access to position where the signal starts.

◆ setScale()

void setScale ( double  scale)
inline

Mutable access to the spacing of raw data.

◆ setSignal()

void setSignal ( const std::vector< Peak1D > &  signal)
inline

Mutable access to the wavelet transform of the signal.

◆ setSignalLength()

void setSignalLength ( const SignedSize  signal_length)
inline

Mutable access to signal length [end_left_padding,begin_right_padding].

◆ setSpacing()

void setSpacing ( double  spacing)
inline

Mutable access to the spacing of raw data.

◆ setWavelet()

void setWavelet ( const std::vector< double > &  wavelet)
inline

Mutable access to the signal.

Member Data Documentation

◆ begin_right_padding_

SignedSize begin_right_padding_
protected

◆ end_left_padding_

SignedSize end_left_padding_
protected

We often have to pad the transform at the left and right with zeros. Since we don't want to iterate over those as well, we have to store their positions.

◆ scale_

double scale_
protected

Spacing and scale of the wavelet and length of the signal.

◆ signal_

std::vector<Peak1D> signal_
protected

The transformed signal.

◆ signal_length_

SignedSize signal_length_
protected

◆ spacing_

double spacing_
protected

◆ wavelet_

std::vector<double> wavelet_
protected

The pre-tabulated wavelet used for the transform.