Serves as a wrapper for the libsvm. More...

#include <OpenMS/ANALYSIS/SVM/SVMWrapper.h>

Inheritance diagram for SVMWrapper:

Collaboration diagram for SVMWrapper:

Public Types
enum	SVM_parameter_type { SVM_TYPE , KERNEL_TYPE , DEGREE , C , NU , P , GAMMA , PROBABILITY , SIGMA , BORDER_LENGTH }
	Parameters for the svm to be set from outside. More...

enum	SVM_kernel_type { OLIGO = 19 , OLIGO_COMBINED }
	Kernel type. More...

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

Public Member Functions
	SVMWrapper ()
	standard constructor More...

virtual	~SVMWrapper ()
	destructor More...

void	setParameter (SVM_parameter_type type, Int value)
	You can set the parameters of the svm: More...

void	setParameter (SVM_parameter_type type, double value)
	sets the double parameters of the svm More...

Int	train (struct svm_problem *problem)
	trains the svm More...

Int	train (SVMData &problem)
	trains the svm More...

void	saveModel (std::string modelFilename) const
	saves the svm model More...

void	loadModel (std::string modelFilename)
	loads the model More...

void	predict (struct svm_problem *problem, std::vector< double > &predicted_labels)
	predicts the labels using the trained model More...

void	predict (const SVMData &problem, std::vector< double > &results)
	predicts the labels using the trained model More...

Int	getIntParameter (SVM_parameter_type type)
	You can get the actual int- parameters of the svm. More...

double	getDoubleParameter (SVM_parameter_type type)
	You can get the actual double- parameters of the svm. More...

void	createRandomPartitions (svm_problem problem, Size number, std::vector< svm_problem > &partitions)
	You can create 'number' equally sized random partitions. More...

void	createRandomPartitions (const SVMData &problem, Size number, std::vector< SVMData > &problems)
	You can create 'number' equally sized random partitions. More...

void	predict (const std::vector< svm_node * > &vectors, std::vector< double > &predicted_rts)
	predicts the labels using the trained model More...

double	performCrossValidation (svm_problem *problem_ul, const SVMData &problem_l, const bool is_labeled, const std::map< SVM_parameter_type, double > &start_values_map, const std::map< SVM_parameter_type, double > &step_sizes_map, const std::map< SVM_parameter_type, double > &end_values_map, Size number_of_partitions, Size number_of_runs, std::map< SVM_parameter_type, double > &best_parameters, bool additive_step_sizes=true, bool output=false, String performances_file_name="performances.txt", bool mcc_as_performance_measure=false)
	Performs a CV for the data given by 'problem'. More...

double	getSVRProbability ()
	Returns the probability parameter sigma of the fitted Laplace model. More...

void	getSignificanceBorders (svm_problem *data, std::pair< double, double > &borders, double confidence=0.95, Size number_of_runs=5, Size number_of_partitions=5, double step_size=0.01, Size max_iterations=1000000)
	calculates the significance borders of the error model and stores them in 'sigmas' More...

void	getSignificanceBorders (const SVMData &data, std::pair< double, double > &sigmas, double confidence=0.95, Size number_of_runs=5, Size number_of_partitions=5, double step_size=0.01, Size max_iterations=1000000)
	calculates the significance borders of the error model and stores them in 'sigmas' More...

double	getPValue (double sigma1, double sigma2, std::pair< double, double > point)
	calculates a p-value for a given data point using the model parameters More...

void	getDecisionValues (svm_problem *data, std::vector< double > &decision_values)
	stores the prediction values for the encoded data in 'decision_values' More...

void	scaleData (svm_problem *data, Int max_scale_value=-1)
	Scales the data such that every column is scaled to [-1, 1]. More...

svm_problem *	computeKernelMatrix (svm_problem problem1, svm_problem problem2)
	computes the kernel matrix using the actual svm parameters and the given data More...

svm_problem *	computeKernelMatrix (const SVMData &problem1, const SVMData &problem2)
	computes the kernel matrix using the actual svm parameters and the given data More...

void	setTrainingSample (svm_problem *training_sample)
	This is used for being able to perform predictions with non libsvm standard kernels. More...

void	setTrainingSample (SVMData &training_sample)
	This is used for being able to perform predictions with non libsvm standard kernels. More...

void	getSVCProbabilities (struct svm_problem *problem, std::vector< double > &probabilities, std::vector< double > &prediction_labels)
	This function fills probabilities with the probability estimates for the first class. More...

void	setWeights (const std::vector< Int > &weight_labels, const std::vector< double > &weights)
	Sets weights for the classes in C_SVC (see libsvm documentation for further details) More...

Public Member Functions inherited from ProgressLogger
	ProgressLogger ()
	Constructor. More...

virtual	~ProgressLogger ()
	Destructor. More...

	ProgressLogger (const ProgressLogger &other)
	Copy constructor. More...

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

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

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

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

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

void	endProgress () const
	Ends the progress display. More...

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

Static Public Member Functions
static svm_problem *	mergePartitions (const std::vector< svm_problem * > &problems, Size except)
	You can merge partitions excluding the partition with index 'except'. More...

static void	mergePartitions (const std::vector< SVMData > &problems, Size except, SVMData &merged_problem)
	You can merge partitions excluding the partition with index 'except'. More...

static void	getLabels (svm_problem *problem, std::vector< double > &labels)
	Stores the stored labels of the encoded SVM data at 'labels'. More...

static double	kernelOligo (const std::vector< std::pair< int, double > > &x, const std::vector< std::pair< int, double > > &y, const std::vector< double > &gauss_table, int max_distance=-1)
	returns the value of the oligo kernel for sequences 'x' and 'y' More...

static double	kernelOligo (const svm_node x, const svm_node y, const std::vector< double > &gauss_table, double sigma_square=0, Size max_distance=50)
	calculates the oligo kernel value for the encoded sequences 'x' and 'y' More...

static void	calculateGaussTable (Size border_length, double sigma, std::vector< double > &gauss_table)

Private Member Functions
bool	nextGrid_ (const std::vector< double > &start_values, const std::vector< double > &step_sizes, const std::vector< double > &end_values, const bool additive_step_sizes, std::vector< double > &actual_values)
	find next grid search parameter combination More...

Size	getNumberOfEnclosedPoints_ (double m1, double m2, const std::vector< std::pair< double, double > > &points)

void	initParameters_ ()
	Initializes the svm with standard parameters. More...

Static Private Member Functions
static void	printToVoid_ (const char *)
	This function is passed to lib svm for output control. More...

Private Attributes
svm_parameter *	param_
	the parameters for the svm More...

svm_model *	model_
	the learned svm discriminant More...

double	sigma_
	for the oligo kernel (amount of positional smearing) More...

std::vector< double >	sigmas_
	for the combined oligo kernel (amount of positional smearing) More...

std::vector< double >	gauss_table_
	lookup table for fast computation of the oligo kernel More...

std::vector< std::vector< double > >	gauss_tables_
	lookup table for fast computation of the combined oligo kernel More...

Size	kernel_type_
	the actual kernel type More...

Size	border_length_
	the actual kernel type More...

svm_problem *	training_set_ = nullptr
	the training set More...

svm_problem *	training_problem_ = nullptr
	the training set More...

SVMData	training_data_
	the training set (different encoding) More...

Math::RandomShuffler	shuffler_
	random shuffler to create training partitions More...

Additional Inherited Members
Static Protected Member Functions inherited from ProgressLogger
static String	logTypeToFactoryName_ (LogType type)
	Return the name of the factory product used for this log type. More...

Protected Attributes inherited from ProgressLogger
LogType	type_

time_t	last_invoke_

ProgressLoggerImpl *	current_logger_

Static Protected Attributes inherited from ProgressLogger
static int	recursion_depth_

Detailed Description

Serves as a wrapper for the libsvm.

This class can be used for svm predictions. You can either perform classification or regression and choose certain kernel functions and additional parameters. Furthermore the models can be saved and loaded and we support also a new kernel function that was specially designed for learning with small sequences of different lengths.

Member Enumeration Documentation

◆ SVM_kernel_type

enum SVM_kernel_type

Kernel type.

Enumerator
OLIGO
OLIGO_COMBINED

◆ SVM_parameter_type

enum SVM_parameter_type

Parameters for the svm to be set from outside.

This type is used to specify the kind of parameter that is to be set or retrieved by the set/getParameter methods.

Enumerator
SVM_TYPE	the svm type cab be NU_SVR or EPSILON_SVR
KERNEL_TYPE	the kernel type
DEGREE	the degree for the polynomial- kernel
C	the C parameter of the svm
NU	the nu parameter for nu-SVR
P	the epsilon parameter for epsilon-SVR
GAMMA	the gamma parameter of the POLY, RBF and SIGMOID kernel
PROBABILITY
SIGMA
BORDER_LENGTH

Constructor & Destructor Documentation

◆ SVMWrapper()

SVMWrapper ( )

standard constructor

◆ ~SVMWrapper()

virtual ~SVMWrapper ( )

virtual

destructor

Member Function Documentation

◆ calculateGaussTable()

static void calculateGaussTable	(	Size	border_length,
		double	sigma,
		std::vector< double > &	gauss_table
	)

static

◆ computeKernelMatrix() [1/2]

svm_problem* computeKernelMatrix	(	const SVMData &	problem1,
		const SVMData &	problem2
	)

computes the kernel matrix using the actual svm parameters and the given data

This function can be used to compute a kernel matrix. 'problem1' and 'problem2' are used together wit the oligo kernel function (could be extended if you want to use your own kernel functions).

◆ computeKernelMatrix() [2/2]

svm_problem* computeKernelMatrix	(	svm_problem *	problem1,
		svm_problem *	problem2
	)

computes the kernel matrix using the actual svm parameters and the given data

This function can be used to compute a kernel matrix. 'problem1' and 'problem2' are used together wit the oligo kernel function (could be extended if you want to use your own kernel functions).

◆ createRandomPartitions() [1/2]

void createRandomPartitions	(	const SVMData &	problem,
		Size	number,
		std::vector< SVMData > &	problems
	)

You can create 'number' equally sized random partitions.

This function creates 'number' equally sized random partitions and stores them in 'partitions'.

◆ createRandomPartitions() [2/2]

void createRandomPartitions	(	svm_problem *	problem,
		Size	number,
		std::vector< svm_problem * > &	partitions
	)

You can create 'number' equally sized random partitions.

This function creates 'number' equally sized random partitions and stores them in 'partitions'.

◆ getDecisionValues()

void getDecisionValues	(	svm_problem *	data,
		std::vector< double > &	decision_values
	)

stores the prediction values for the encoded data in 'decision_values'

This function can be used to get the prediction values of the data if a model is already trained by the train() method. For regression the result is the same as for the method predict. For classification this function returns the distance from the separating hyperplane. For multiclass classification the decision_values vector will be empty.

◆ getDoubleParameter()

double getDoubleParameter ( SVM_parameter_type type )

You can get the actual double- parameters of the svm.

Parameter types
C	the C parameter of the svm
P	the P parameter of the svm (sets the epsilon in epsilon-svr)
NU	the nu parameter in nu-SVR
GAMMA	for POLY, RBF and SIGMOID

Parameters

type	The parameter that should be returned.

◆ getIntParameter()

Int getIntParameter ( SVM_parameter_type type )

You can get the actual int- parameters of the svm.

Parameter types
KERNEL_TYPE	LINEAR	for the linear kernel
	RBF	for the rbf kernel
	POLY	for the polynomial kernel
	SIGMOID	for the sigmoid kernel
DEGREE	the degree for the polynomial- kernel and the locality- improved kernel
SVM_TYPE	he SVM type of the svm: can be NU_SVR or EPSILON_SVR

Parameters

type	The parameter that should be returned.

◆ getLabels()

static void getLabels	(	svm_problem *	problem,
		std::vector< double > &	labels
	)

static

Stores the stored labels of the encoded SVM data at 'labels'.

◆ getNumberOfEnclosedPoints_()

Size getNumberOfEnclosedPoints_	(	double	m1,
		double	m2,
		const std::vector< std::pair< double, double > > &	points
	)

private

◆ getPValue()

double getPValue	(	double	sigma1,
		double	sigma2,
		std::pair< double, double >	point
	)

calculates a p-value for a given data point using the model parameters

Uses the model parameters to calculate the p-value for 'point' which has the data entries: measured, predicted retention time.

◆ getSignificanceBorders() [1/2]

void getSignificanceBorders	(	const SVMData &	data,
		std::pair< double, double > &	sigmas,
		double	confidence = `0.95`,
		Size	number_of_runs = `5`,
		Size	number_of_partitions = `5`,
		double	step_size = `0.01`,
		Size	max_iterations = `1000000`
	)

calculates the significance borders of the error model and stores them in 'sigmas'

◆ getSignificanceBorders() [2/2]

void getSignificanceBorders	(	svm_problem *	data,
		std::pair< double, double > &	borders,
		double	confidence = `0.95`,
		Size	number_of_runs = `5`,
		Size	number_of_partitions = `5`,
		double	step_size = `0.01`,
		Size	max_iterations = `1000000`
	)

calculates the significance borders of the error model and stores them in 'sigmas'

◆ getSVCProbabilities()

void getSVCProbabilities	(	struct svm_problem *	problem,
		std::vector< double > &	probabilities,
		std::vector< double > &	prediction_labels
	)

This function fills probabilities with the probability estimates for the first class.

The libSVM function svm_predict_probability is called to get probability estimates for the positive class. Since this is only used for binary classification it is sufficient for every test example to report the probability of the test example belonging to the positive class. Probability estimates have to be turned on during training (svm.setParameter(PROBABILITY, 1)), otherwise this method will fill the 'probabilities' vector with -1s.

◆ getSVRProbability()

double getSVRProbability ( )

Returns the probability parameter sigma of the fitted Laplace model.

The libsvm is used to fit a Laplace model to the prediction values by performing an internal cv using the training set if setParameter(PROBABILITY, 1) was invoked before using train. Look for your libsvm documentation for more details. The model parameter sigma is returned by this method. If no model was fitted during training zero is returned.

◆ initParameters_()

void initParameters_ ( )

private

Initializes the svm with standard parameters.

◆ kernelOligo() [1/2]

static double kernelOligo	(	const std::vector< std::pair< int, double > > &	x,
		const std::vector< std::pair< int, double > > &	y,
		const std::vector< double > &	gauss_table,
		int	max_distance = `-1`
	)

static

returns the value of the oligo kernel for sequences 'x' and 'y'

This function computes the kernel value of the oligo kernel, which was introduced by Meinicke et al. in 2004. 'x' and 'y' are encoded by encodeOligo and 'gauss_table' has to be constructed by calculateGaussTable.

'max_distance' can be used to speed up the computation even further by restricting the maximum distance between a k_mer at position i in sequence 'x' and a k_mer at position j in sequence 'y'. If i - j > 'max_distance' the value is not added to the kernel value. This approximation is switched off by default (max_distance < 0).

◆ kernelOligo() [2/2]

static double kernelOligo	(	const svm_node *	x,
		const svm_node *	y,
		const std::vector< double > &	gauss_table,
		double	sigma_square = `0`,
		Size	max_distance = `50`
	)

static

calculates the oligo kernel value for the encoded sequences 'x' and 'y'

This kernel function calculates the oligo kernel value [Meinicke 04] for the sequences 'x' and 'y' that had been encoded by the encodeOligoBorder... function of the LibSVMEncoder class.

◆ loadModel()

void loadModel ( std::string modelFilename )

loads the model

The svm- model is loaded. After this, the svm is ready for prediction.

Parameters

modelFilename The name of the model file that should be loaded.

◆ mergePartitions() [1/2]

static svm_problem* mergePartitions	(	const std::vector< svm_problem * > &	problems,
		Size	except
	)

static

You can merge partitions excluding the partition with index 'except'.

◆ mergePartitions() [2/2]

static void mergePartitions	(	const std::vector< SVMData > &	problems,
		Size	except,
		SVMData &	merged_problem
	)

static

You can merge partitions excluding the partition with index 'except'.

◆ nextGrid_()

bool nextGrid_	(	const std::vector< double > &	start_values,
		const std::vector< double > &	step_sizes,
		const std::vector< double > &	end_values,
		const bool	additive_step_sizes,
		std::vector< double > &	actual_values
	)

private

find next grid search parameter combination

The current grid cell is given in actual_values. The result is returned in actual_values.

◆ performCrossValidation()

double performCrossValidation	(	svm_problem *	problem_ul,
		const SVMData &	problem_l,
		const bool	is_labeled,
		const std::map< SVM_parameter_type, double > &	start_values_map,
		const std::map< SVM_parameter_type, double > &	step_sizes_map,
		const std::map< SVM_parameter_type, double > &	end_values_map,
		Size	number_of_partitions,
		Size	number_of_runs,
		std::map< SVM_parameter_type, double > &	best_parameters,
		bool	additive_step_sizes = `true`,
		bool	output = `false`,
		String	performances_file_name = `"performances.txt"`,
		bool	mcc_as_performance_measure = `false`
	)

Performs a CV for the data given by 'problem'.

◆ predict() [1/3]

void predict	(	const std::vector< svm_node * > &	vectors,
		std::vector< double > &	predicted_rts
	)

predicts the labels using the trained model

The prediction process is started and the results are stored in 'predicted_rts'.

◆ predict() [2/3]

void predict	(	const SVMData &	problem,
		std::vector< double > &	results
	)

predicts the labels using the trained model

The prediction process is started and the results are stored in 'predicted_labels'.

◆ predict() [3/3]

void predict	(	struct svm_problem *	problem,
		std::vector< double > &	predicted_labels
	)

predicts the labels using the trained model

The prediction process is started and the results are stored in 'predicted_labels'.

Referenced by LibSVMEncoder::predictPeptideRT().

◆ printToVoid_()

static void printToVoid_ ( const char * )

staticprivate

This function is passed to lib svm for output control.

The intention is to discard the output, as we don't need it.

◆ saveModel()

void saveModel ( std::string modelFilename ) const

saves the svm model

The model of the trained svm is saved into 'modelFilename'. Throws an exception if the model cannot be saved.

Exceptions

Exception::UnableToCreateFile

Parameters

modelFilename The file name where the model will be saved.

◆ scaleData()

void scaleData	(	svm_problem *	data,
		Int	max_scale_value = `-1`
	)

Scales the data such that every column is scaled to [-1, 1].

Scales the x[][].value values of the svm_problem* structure. If the second parameter is omitted, the data is scaled to [-1, 1]. Otherwise the data is scaled to [0, max_scale_value]

◆ setParameter() [1/2]

void setParameter	(	SVM_parameter_type	type,
		double	value
	)

sets the double parameters of the svm

Parameters

type	The type of parameter to set.
value	The new value for parameter `type`.

◆ setParameter() [2/2]

void setParameter	(	SVM_parameter_type	type,
		Int	value
	)

You can set the parameters of the svm:

Parameter types
KERNEL_TYPE	LINEAR	for the linear kernel
	RBF	for the rbf kernel
	POLY	for the polynomial kernel
	SIGMOID	for the sigmoid kernel
DEGREE	the degree for the polynomial- kernel and the locality- improved kernel
C	the C parameter of the svm

Parameters

type	The type of parameter to set.
value	The new value for parameter `type`.

◆ setTrainingSample() [1/2]

void setTrainingSample ( svm_problem * training_sample )

This is used for being able to perform predictions with non libsvm standard kernels.

◆ setTrainingSample() [2/2]

void setTrainingSample ( SVMData & training_sample )

This is used for being able to perform predictions with non libsvm standard kernels.

◆ setWeights()

void setWeights	(	const std::vector< Int > &	weight_labels,
		const std::vector< double > &	weights
	)

Sets weights for the classes in C_SVC (see libsvm documentation for further details)

◆ train() [1/2]

Int train ( struct svm_problem * problem )

trains the svm

The svm is trained with the data stored in the 'svm_problem' structure.

◆ train() [2/2]

Int train ( SVMData & problem )

trains the svm

The svm is trained with the data stored in the 'SVMData' structure.

Member Data Documentation

◆ border_length_

Size border_length_

private

the actual kernel type

◆ gauss_table_

std::vector<double> gauss_table_

private

lookup table for fast computation of the oligo kernel

◆ gauss_tables_

std::vector<std::vector<double> > gauss_tables_

private

lookup table for fast computation of the combined oligo kernel

◆ kernel_type_

Size kernel_type_

private

the actual kernel type

◆ model_

svm_model* model_

private

the learned svm discriminant

◆ param_

svm_parameter* param_

private

the parameters for the svm

◆ shuffler_

Math::RandomShuffler shuffler_

private

random shuffler to create training partitions

◆ sigma_

double sigma_

private

for the oligo kernel (amount of positional smearing)

◆ sigmas_

std::vector<double> sigmas_

private

for the combined oligo kernel (amount of positional smearing)

◆ training_data_

SVMData training_data_

private

the training set (different encoding)

◆ training_problem_

svm_problem* training_problem_ = nullptr

private

the training set

◆ training_set_

svm_problem* training_set_ = nullptr

private

the training set

Public Types

Public Member Functions

Static Public Member Functions

Private Member Functions

Static Private Member Functions

Private Attributes

Additional Inherited Members

Detailed Description

Member Enumeration Documentation

◆ SVM_kernel_type

◆ SVM_parameter_type

Constructor & Destructor Documentation

◆ SVMWrapper()

◆ ~SVMWrapper()

Member Function Documentation

◆ calculateGaussTable()

◆ computeKernelMatrix() [1/2]

◆ computeKernelMatrix() [2/2]

◆ createRandomPartitions() [1/2]

◆ createRandomPartitions() [2/2]

◆ getDecisionValues()

◆ getDoubleParameter()

◆ getIntParameter()

◆ getLabels()

◆ getNumberOfEnclosedPoints_()

◆ getPValue()

◆ getSignificanceBorders() [1/2]

◆ getSignificanceBorders() [2/2]

◆ getSVCProbabilities()

◆ getSVRProbability()

◆ initParameters_()

◆ kernelOligo() [1/2]

◆ kernelOligo() [2/2]

◆ loadModel()

◆ mergePartitions() [1/2]

◆ mergePartitions() [2/2]

◆ nextGrid_()

◆ performCrossValidation()

◆ predict() [1/3]

◆ predict() [2/3]

◆ predict() [3/3]

◆ printToVoid_()

◆ saveModel()

◆ scaleData()

◆ setParameter() [1/2]

◆ setParameter() [2/2]

◆ setTrainingSample() [1/2]

◆ setTrainingSample() [2/2]

◆ setWeights()

◆ train() [1/2]

◆ train() [2/2]

Member Data Documentation

◆ border_length_

◆ gauss_table_

◆ gauss_tables_

◆ kernel_type_

◆ model_

◆ param_

◆ shuffler_

◆ sigma_

◆ sigmas_

◆ training_data_

◆ training_problem_

◆ training_set_