Computes a modified sinc smoothing filter for profile data. More...

#include <OpenMS/PROCESSING/SMOOTHING/ModifiedSincSmoother.h>

Inheritance diagram for ModifiedSincSmoother:

Collaboration diagram for ModifiedSincSmoother:

Classes
struct	LinearRegression
	Helper class for weighted linear regression. More...

Public Member Functions
	ModifiedSincSmoother ()
	Default constructor with parameter registration.

	ModifiedSincSmoother (bool isMS1, int degree, int m)
	Constructor initializing the modified sinc smoother.

std::vector< double >	smooth (const std::vector< double > &data)
	Smooth a vector of data values with boundary extrapolation.

std::vector< double >	smoothExceptBoundaries (const std::vector< double > &data)
	Smooth data without boundary extension (interior points only).

void	filter (MSSpectrum &spectrum)
	Apply modified sinc smoothing to an MSSpectrum.

void	filter (MSChromatogram &chromatogram)
	Apply modified sinc smoothing to an MSChromatogram.

void	filter (Mobilogram &mobilogram)
	Apply modified sinc smoothing to a Mobilogram.

void	filterExperiment (PeakMap &map)
	Apply modified sinc smoothing to all spectra and chromatograms in an experiment.

Public Member Functions inherited from ProgressLogger
	ProgressLogger ()
	Constructor.

virtual	~ProgressLogger ()
	Destructor.

	ProgressLogger (const ProgressLogger &other)
	Copy constructor.

ProgressLogger &	operator= (const ProgressLogger &other)
	Assignment Operator.

void	setLogType (LogType type) const
	Sets the progress log that should be used. The default type is NONE!

LogType	getLogType () const
	Returns the type of progress log being used.

void	setLogger (ProgressLoggerImpl *logger)
	Sets the logger to be used for progress logging.

void	startProgress (SignedSize begin, SignedSize end, const String &label) const
	Initializes the progress display.

void	setProgress (SignedSize value) const
	Sets the current progress.

void	endProgress (UInt64 bytes_processed=0) const

void	nextProgress () const
	increment progress by 1 (according to range begin-end)

Public Member Functions inherited from DefaultParamHandler
	DefaultParamHandler (const String &name)
	Constructor with name that is displayed in error messages.

	DefaultParamHandler (const DefaultParamHandler &rhs)
	Copy constructor.

virtual	~DefaultParamHandler ()
	Destructor.

DefaultParamHandler &	operator= (const DefaultParamHandler &rhs)
	Assignment operator.

virtual bool	operator== (const DefaultParamHandler &rhs) const
	Equality operator.

void	setParameters (const Param &param)
	Sets the parameters.

const Param &	getParameters () const
	Non-mutable access to the parameters.

const Param &	getDefaults () const
	Non-mutable access to the default parameters.

const String &	getName () const
	Non-mutable access to the name.

void	setName (const String &name)
	Mutable access to the name.

const std::vector< String > &	getSubsections () const
	Non-mutable access to the registered subsections.

Static Public Member Functions
static int	bandwidthToM (bool isMS1, int degree, double bandwidth)
	Convert frequency domain bandwidth to spatial domain parameter m.

static int	noiseGainToM (bool isMS1, int degree, double noiseGain)
	Convert desired noise gain to optimal parameter m.

static double	savitzkyGolayBandwidth (int degree, int m)
	Compute equivalent Savitzky-Golay bandwidth for comparison.

Static Public Member Functions inherited from DefaultParamHandler
static void	writeParametersToMetaValues (const Param &write_this, MetaInfoInterface &write_here, const String &key_prefix="")
	Writes all parameters to meta values.

Protected Member Functions
void	updateMembers_ () override
	Sync internal state from parameters.

Protected Member Functions inherited from DefaultParamHandler
void	defaultsToParam_ ()
	Updates the parameters after the defaults have been set in the constructor.

Private Member Functions
void	registerDefaults_ ()
	Register default parameters (shared by both constructors)

std::vector< double >	makeKernel (bool isMS1, int degree, int m, const std::vector< double > &coeffs)
	Construct the modified sinc convolution kernel.

std::vector< double >	getCoefficients (bool isMS1, int degree, int m)
	Get correction coefficients for kernel modification.

std::vector< double >	makeFitWeights (bool isMS1, int degree, int m)
	Construct weights for boundary linear regression.

std::vector< double >	extendData (const std::vector< double > &data, int m)
	Extend data boundaries using weighted linear regression.

Static Private Member Functions
static double	sqr (double x)
	Helper: square function.

Private Attributes
bool	isMS1_
	MS1 vs MS filter variant selection.

int	degree_
	Polynomial degree for sinc modification (even, 2-10)

int	m_
	Kernel half-width parameter (spatial domain)

std::vector< double >	kernel_
	Symmetric convolution kernel coefficients [k_0, k_1, ..., k_m].

std::vector< double >	fit_weights_
	Weights for boundary linear regression fitting.

Additional Inherited Members
Public Types inherited from ProgressLogger
enum	LogType { CMD , GUI , NONE }
	Possible log types. More...

Protected Attributes inherited from ProgressLogger
LogType	type_

time_t	last_invoke_

ProgressLoggerImpl *	current_logger_

Protected Attributes inherited from DefaultParamHandler
Param	param_
	Container for current parameters.

Param	defaults_
	Container for default parameters. This member should be filled in the constructor of derived classes!

std::vector< String >	subsections_
	Container for registered subsections. This member should be filled in the constructor of derived classes!

String	error_name_
	Name that is displayed in error messages during the parameter checking.

bool	check_defaults_
	If this member is set to false no checking if parameters in done;.

bool	warn_empty_defaults_
	If this member is set to false no warning is emitted when defaults are empty;.

Static Protected Attributes inherited from ProgressLogger
static int	recursion_depth_

Detailed Description

Computes a modified sinc smoothing filter for profile data.

This class implements the modified sinc (MS) filter from Schmid, Rath & Diebold, "Why and How Savitzky-Golay Filters Should Be Replaced", ACS Measurement Science Au, 2022, 2(2), 185-196.

The filter supports two variants from the paper:

MS variant (isMS1 = false): wider kernel with decay alpha=4, providing excellent stopband suppression (~-70 dB). Requires m >= degree/2 + 2.
MS1 variant (isMS1 = true): narrower kernel with decay alpha=2, trading some stopband suppression for a smaller kernel size. Requires m >= degree/2 + 1.

Note: These are two filter design variants, NOT mass spectrometry scan level designations. The naming follows the original paper's convention.

The algorithm constructs a symmetric convolution kernel based on a sinc function multiplied by a modified Gaussian window, with optional passband correction terms for polynomial degrees >= 6 (MS) or >= 4 (MS1).

Parameters:

degree: polynomial degree (must be even, 2-10). Higher degree gives sharper cutoff.
m: kernel half-width (determines smoothing strength). Use bandwidthToM() or noiseGainToM() to compute from desired filter characteristics.
isMS1: selects MS1 (smaller) vs MS (wider) kernel variant.

Boundary Handling: For finite data, the filter extends boundaries using weighted linear regression on the nearest available data points to minimize edge artifacts.

Container Support: The filter can process individual data vectors or OpenMS container types (MSSpectrum, MSChromatogram, Mobilogram, PeakMap).

Note: This filter works optimally with uniform profile data.; Data should be sorted by position (m/z, RT, or IM) before filtering.

Constructor & Destructor Documentation

◆ ModifiedSincSmoother() [1/2]

ModifiedSincSmoother ( )

Default constructor with parameter registration.

◆ ModifiedSincSmoother() [2/2]

ModifiedSincSmoother	(	bool	isMS1,
		int	degree,
		int	m
	)

Constructor initializing the modified sinc smoother.

Parameters

isMS1	true for MS1 variant (smaller kernel), false for MS variant (wider kernel)
degree	polynomial degree for sinc modification (must be even, 2-10)
m	kernel half-width parameter (spatial domain)

Exceptions

std::invalid_argument	if degree is not even or outside [2,10]
std::invalid_argument	if m is too small for the given degree (minimum: m >= degree/2 + 1 for MS1 variant, m >= degree/2 + 2 for MS variant)

Member Function Documentation

◆ bandwidthToM()

static int bandwidthToM	(	bool	isMS1,
		int	degree,
		double	bandwidth
	)

static

Convert frequency domain bandwidth to spatial domain parameter m.

Parameters

isMS1	true for MS1 variant (smaller kernel), false for MS variant (wider kernel)
degree	polynomial degree used in the filter
bandwidth	normalized frequency bandwidth in (0, 0.5)

Returns: corresponding kernel half-width parameter m

Exceptions

std::invalid_argument if bandwidth ∉ (0, 0.5)

Note: Bandwidth interpretation: fraction of Nyquist frequency; Smaller bandwidth → larger m → more smoothing

◆ extendData()

std::vector< double > extendData	(	const std::vector< double > &	data,
		int	m
	)

private

Extend data boundaries using weighted linear regression.

Fits linear models to boundary regions using cosine-squared weights, then extrapolates to extend the data symmetrically by m points on each side.

Parameters

data	original data vector
m	extension half-width (points to add on each side)

Returns: extended data: [left_ext..., original_data..., right_ext...]

Note: Left extension: fit forward from start, extrapolate backward; Right extension: fit backward from end, extrapolate forward; Uses fit_weights_ for regression weighting

◆ filter() [1/3]

void filter ( Mobilogram & mobilogram )

Apply modified sinc smoothing to a Mobilogram.

Smooths the intensity values of an ion mobility spectrum while preserving all mobility positions and metadata.

Parameters

mobilogram input/output mobilogram to be smoothed in-place

Note: IM positions are preserved exactly; only intensities modified; Negative intensities are clipped to zero after smoothing

◆ filter() [2/3]

void filter ( MSChromatogram & chromatogram )

Apply modified sinc smoothing to an MSChromatogram.

Smooths the intensity values of a chromatogram while preserving all retention time positions and metadata.

Parameters

chromatogram input/output chromatogram to be smoothed in-place

Note: RT positions are preserved exactly; only intensities modified; Negative intensities are clipped to zero after smoothing

◆ filter() [3/3]

void filter ( MSSpectrum & spectrum )

Apply modified sinc smoothing to an MSSpectrum.

Smooths the intensity values of a mass spectrum while preserving all m/z positions and metadata. Uses copy-filter-swap pattern for exception safety.

Parameters

spectrum input/output spectrum to be smoothed in-place

Note: Requires uniform profile data for optimal results; m/z positions are preserved exactly; only intensities modified; Negative intensities are clipped to zero after smoothing

◆ filterExperiment()

void filterExperiment ( PeakMap & map )

Apply modified sinc smoothing to all spectra and chromatograms in an experiment.

Batch processes all MSSpectra and MSChromatograms in the peak map, applying the filter to each container. Progress is reported via the ProgressLogger interface.

Parameters

map	input/output experiment to be smoothed in-place

Note: Processes both spectra and chromatograms; Progress is reported as "smoothing data" with step counts; All metadata and positions are preserved across containers

◆ getCoefficients()

std::vector< double > getCoefficients	(	bool	isMS1,
		int	degree,
		int	m
	)

private

Get correction coefficients for kernel modification.

Parameters

isMS1	MS1 vs MS filter variant selection
degree	polynomial degree
m	kernel half-width

Returns: correction coefficients (currently empty placeholder)

Note: Currently returns empty vector; reserved for future enhancements

◆ makeFitWeights()

std::vector< double > makeFitWeights	(	bool	isMS1,
		int	degree,
		int	m
	)

private

Construct weights for boundary linear regression.

Derives fitting window size and cosine-squared weights based on kernel first zero location and empirical beta parameter.

Parameters

isMS1	MS1 vs MS filter variant selection
degree	polynomial degree
m	kernel half-width

Returns: weight vector for boundary fitting

Note: Weight derivation: first_zero → beta → fit_length → cos²(·) weights; Optimized separately for MS1 and MS filter variant boundary characteristics

◆ makeKernel()

std::vector< double > makeKernel	(	bool	isMS1,
		int	degree,
		int	m,
		const std::vector< double > &	coeffs
	)

private

Construct the modified sinc convolution kernel.

Parameters

isMS1	MS1 vs MS filter variant selection
degree	polynomial degree for modifications
m	kernel half-width
coeffs	optional correction coefficients (currently unused)

Returns: normalized symmetric kernel [k_0, k_1, ..., k_m]

Note: Kernel is normalized to unity gain at DC (frequency 0); Incorporates exponential decay terms for improved characteristics

◆ noiseGainToM()

static int noiseGainToM	(	bool	isMS1,
		int	degree,
		double	noiseGain
	)

static

Convert desired noise gain to optimal parameter m.

Computes the kernel half-width m that achieves the specified noise gain factor for the given filter configuration.

Parameters

isMS1	true for MS1 variant (smaller kernel), false for MS variant (wider kernel)
degree	polynomial degree used in the filter
noiseGain	desired noise amplification factor (typically < 1)

Returns: corresponding kernel half-width parameter m

Note: noiseGain = 1 means noise unchanged, < 1 means noise reduction; Formula based on empirical characterization of filter response

◆ registerDefaults_()

void registerDefaults_ ( )

private

Register default parameters (shared by both constructors)

◆ savitzkyGolayBandwidth()

static double savitzkyGolayBandwidth	(	int	degree,
		int	m
	)

static

Compute equivalent Savitzky-Golay bandwidth for comparison.

Given modified sinc parameters, computes the frequency bandwidth that a Savitzky-Golay filter of the same degree would need to achieve similar smoothing characteristics.

Parameters

degree	polynomial degree
m	kernel half-width parameter

Returns: equivalent Savitzky-Golay bandwidth

Note: Useful for comparing filter characteristics across algorithms; Based on empirical matching of filter frequency responses

◆ smooth()

std::vector< double > smooth ( const std::vector< double > & data )

Smooth a vector of data values with boundary extrapolation.

Applies the modified sinc filter to the input data. Boundaries are extended using weighted linear regression before filtering, then the result is trimmed to match the original input size.

Parameters

data	input intensity values to be smoothed

Returns: smoothed data vector of same size as input

Note: Returns empty vector if input is empty; All values in returned vector correspond to positions in original data

◆ smoothExceptBoundaries()

std::vector< double > smoothExceptBoundaries ( const std::vector< double > & data )

Smooth data without boundary extension (interior points only).

Applies the modified sinc filter only where a full symmetric neighborhood exists. Boundary points are left as zeros. This method is typically used with pre-extended data.

Parameters

data	input data (potentially pre-extended)

Returns: smoothed data vector of same size, with boundary zeros

Note: Use this method when you have already extended the data externally; Boundary points (within kernel radius) will be zero in output

◆ sqr()

static double sqr ( double x )

inlinestaticprivate

Helper: square function.

◆ updateMembers_()

void updateMembers_ ( )

overrideprotectedvirtual

Sync internal state from parameters.

Reimplemented from DefaultParamHandler.

Member Data Documentation

◆ degree_

int degree_

private

Polynomial degree for sinc modification (even, 2-10)

◆ fit_weights_

std::vector<double> fit_weights_

private

Weights for boundary linear regression fitting.

◆ isMS1_

bool isMS1_

private

MS1 vs MS filter variant selection.

◆ kernel_

std::vector<double> kernel_

private

Symmetric convolution kernel coefficients [k_0, k_1, ..., k_m].

◆ m_

int m_

private

Kernel half-width parameter (spatial domain)

Classes

Public Member Functions

Static Public Member Functions

Protected Member Functions

Private Member Functions

Static Private Member Functions

Private Attributes

Additional Inherited Members

Detailed Description

Constructor & Destructor Documentation

◆ ModifiedSincSmoother() [1/2]

◆ ModifiedSincSmoother() [2/2]

Member Function Documentation

◆ bandwidthToM()

◆ extendData()

◆ filter() [1/3]

◆ filter() [2/3]

◆ filter() [3/3]

◆ filterExperiment()

◆ getCoefficients()

◆ makeFitWeights()

◆ makeKernel()

◆ noiseGainToM()

◆ registerDefaults_()

◆ savitzkyGolayBandwidth()

◆ smooth()

◆ smoothExceptBoundaries()

◆ sqr()

◆ updateMembers_()

Member Data Documentation

◆ degree_

◆ fit_weights_

◆ isMS1_

◆ kernel_

◆ m_