CCommUlongStringKernel Class Reference

Inheritance diagram for CCommUlongStringKernel:

Detailed Description

The CommUlongString kernel may be used to compute the spectrum kernel [ from strings that have been mapped into unsigned 64bit integers. These 64bit integers correspond to k-mers. To applicable in this kernel they need to be sorted (e.g. via the SortUlongString pre-processor).

It basically uses the algorithm in the unix "comm" command (hence the name) to compute:

$k({\bf x},({\bf x'})= \Phi_k({\bf x})\cdot \Phi_k({\bf x'})$

where $\Phi_k$ maps a sequence ${\bf x}$ that consists of letters in $\Sigma$ to a feature vector of size $|\Sigma|^k$ . In this feature vector each entry denotes how often the k-mer appears in that ${\bf x}$ .

Note that this representation enables spectrum kernels of order 8 for 8bit alphabets (like binaries) and order 32 for 2-bit alphabets like DNA.

For this kernel the linadd speedups are implemented (though there is room for improvement here when a whole set of sequences is ADDed) using sorted lists.

Definition at line 43 of file CommUlongStringKernel.h.


Public Member Functions
	CCommUlongStringKernel (INT size=10, bool use_sign=false, ENormalizationType normalization_=FULL_NORMALIZATION)
	CCommUlongStringKernel (CStringFeatures< ULONG > l, CStringFeatures< ULONG > r, bool use_sign=false, ENormalizationType normalization_=FULL_NORMALIZATION, INT size=10)
virtual	~CCommUlongStringKernel ()
virtual bool	init (CFeatures l, CFeatures r)
virtual void	cleanup ()
bool	load_init (FILE *src)
bool	save_init (FILE *dest)
virtual EKernelType	get_kernel_type ()
virtual const CHAR *	get_name ()
virtual bool	init_optimization (INT count, INT IDX, DREAL weights)
virtual bool	delete_optimization ()
virtual DREAL	compute_optimized (INT idx)
void	merge_dictionaries (INT &t, INT j, INT &k, ULONG vec, ULONG dic, DREAL *dic_weights, DREAL weight, INT vec_idx, INT len, ENormalizationType p_normalization)
virtual void	add_to_normal (INT idx, DREAL weight)
virtual void	clear_normal ()
virtual void	remove_lhs ()
virtual void	remove_rhs ()
virtual EFeatureType	get_feature_type ()
void	get_dictionary (INT &dsize, ULONG &dict, DREAL &dweights)
virtual EFeatureClass	get_feature_class ()
DREAL	kernel (INT idx_a, INT idx_b)
void	get_kernel_matrix (DREAL *dst, INT m, INT *n)
virtual DREAL *	get_kernel_matrix_real (int &m, int &n, DREAL *target)
virtual SHORTREAL *	get_kernel_matrix_shortreal (int &m, int &n, SHORTREAL *target)
bool	load (CHAR *fname)
bool	save (CHAR *fname)
CFeatures *	get_lhs ()
CFeatures *	get_rhs ()
INT	get_num_vec_lhs ()
INT	get_num_vec_rhs ()
bool	has_features ()
virtual void	remove_lhs_and_rhs ()
void	set_cache_size (INT size)
int	get_cache_size ()
void	list_kernel ()
bool	has_property (EKernelProperty p)
EOptimizationType	get_optimization_type ()
virtual void	set_optimization_type (EOptimizationType t)
bool	get_is_initialized ()
bool	init_optimization_svm (CSVM *svm)
virtual void	compute_batch (INT num_vec, INT vec_idx, DREAL target, INT num_suppvec, INT IDX, DREAL alphas, DREAL factor=1.0)
DREAL	get_combined_kernel_weight ()
void	set_combined_kernel_weight (double nw)
virtual INT	get_num_subkernels ()
virtual void	compute_by_subkernel (INT vector_idx, DREAL *subkernel_contrib)
virtual const DREAL *	get_subkernel_weights (INT &num_weights)
virtual void	set_subkernel_weights (DREAL *weights, INT num_weights)
bool	get_precompute_matrix ()
bool	get_precompute_subkernel_matrix ()
virtual void	set_precompute_matrix (bool flag, bool subkernel_flag)
Static Public Attributes
static CParallel	parallel
static CIO	io
static CVersion	version
Protected Member Functions
DREAL	compute (INT idx_a, INT idx_b)
DREAL	normalize_weight (DREAL value, INT seq_num, INT seq_len, ENormalizationType p_normalization)
void	set_property (EKernelProperty p)
void	unset_property (EKernelProperty p)
void	set_is_initialized (bool p_init)
void	do_precompute_matrix ()
	matrix precomputation
void	init_sqrt_diag (DREAL *v, INT num)
Protected Attributes
DREAL *	sqrtdiag_lhs
DREAL *	sqrtdiag_rhs
bool	initialized
CDynamicArray< ULONG >	dictionary
CDynamicArray< DREAL >	dictionary_weights
bool	use_sign
ENormalizationType	normalization
INT	cache_size
	cache_size in MB
KERNELCACHE_ELEM *	kernel_matrix
SHORTREAL *	precomputed_matrix
bool	precompute_subkernel_matrix
bool	precompute_matrix
CFeatures *	lhs
	feature vectors to occur on left hand side
CFeatures *	rhs
	feature vectors to occur on right hand side
DREAL	combined_kernel_weight
bool	optimization_initialized
EOptimizationType	opt_type
ULONG	properties

Constructor & Destructor Documentation

CCommUlongStringKernel::CCommUlongStringKernel	(	INT	size = `10`,
		bool	use_sign = `false`,
		ENormalizationType	normalization_ = `FULL_NORMALIZATION`
	)

constructor

Parameters:

	size	cache size
	use_sign	if sign shall be used
	normalization_	type of normalization

Definition at line 16 of file CommUlongStringKernel.cpp.

CCommUlongStringKernel::CCommUlongStringKernel	(	CStringFeatures< ULONG > *	l,
		CStringFeatures< ULONG > *	r,
		bool	use_sign = `false`,
		ENormalizationType	normalization_ = `FULL_NORMALIZATION`,
		INT	size = `10`
	)

constructor

Parameters:

	l	features of left-hand side
	r	features of right-hand side
	use_sign	if sign shall be used
	normalization_	type of normalization
	size	cache size

Definition at line 25 of file CommUlongStringKernel.cpp.

CCommUlongStringKernel::~CCommUlongStringKernel ( ) [virtual]

Definition at line 36 of file CommUlongStringKernel.cpp.

Member Function Documentation

bool CCommUlongStringKernel::init	(	CFeatures *	l,
		CFeatures *	r
	)			`[virtual]`

initialize kernel

Parameters:

	l	features of left-hand side
	r	features of right-hand side

Returns:: if initializing was successful

Reimplemented from CStringKernel< ST >.

Definition at line 74 of file CommUlongStringKernel.cpp.

void CCommUlongStringKernel::cleanup ( ) [virtual]

clean up kernel

Reimplemented from CKernel.

Definition at line 136 of file CommUlongStringKernel.cpp.

bool CCommUlongStringKernel::load_init ( FILE * src ) [virtual]

load kernel init_data

Parameters:

src

file to load from

Returns:: if loading was successful

Implements CKernel.

Definition at line 154 of file CommUlongStringKernel.cpp.

bool CCommUlongStringKernel::save_init ( FILE * dest ) [virtual]

save kernel init_data

Parameters:

dest

file to save to

Returns:: if saving was successful

Implements CKernel.

Definition at line 159 of file CommUlongStringKernel.cpp.

virtual EKernelType CCommUlongStringKernel::get_kernel_type ( ) [virtual]

return what type of kernel we are

Returns:: kernel type COMMULONGSTRING

Implements CKernel.

Definition at line 100 of file CommUlongStringKernel.h.

virtual const CHAR* CCommUlongStringKernel::get_name ( ) [virtual]

return the kernel's name

Returns:: name CommUlongString

Implements CKernel.

Definition at line 106 of file CommUlongStringKernel.h.

bool CCommUlongStringKernel::init_optimization	(	INT	count,
		INT *	IDX,
		DREAL *	weights
	)			`[virtual]`

initialize optimization

Parameters:

	count	count
	IDX	index
	weights	weights

Returns:: if initializing was successful

Reimplemented from CKernel.

Definition at line 319 of file CommUlongStringKernel.cpp.

bool CCommUlongStringKernel::delete_optimization ( ) [virtual]

delete optimization

Returns:: if deleting was successful

Reimplemented from CKernel.

Definition at line 346 of file CommUlongStringKernel.cpp.

DREAL CCommUlongStringKernel::compute_optimized ( INT idx ) [virtual]

compute optimized

Parameters:

idx

index to compute

Returns:: optimized value at given index

Reimplemented from CKernel.

Definition at line 355 of file CommUlongStringKernel.cpp.

void CCommUlongStringKernel::merge_dictionaries	(	INT &	t,
		INT	j,
		INT &	k,
		ULONG *	vec,
		ULONG *	dic,
		DREAL *	dic_weights,
		DREAL	weight,
		INT	vec_idx,
		INT	len,
		ENormalizationType	p_normalization
	)

merge dictionaries

Parameters:

	t	t
	j	j
	k	k
	vec	vector
	dic	dictionary
	dic_weights	dictionary weights
	weight	weight
	vec_idx	vector index
	len	length
	p_normalization	normalization

Definition at line 143 of file CommUlongStringKernel.h.

void CCommUlongStringKernel::add_to_normal	(	INT	idx,
		DREAL	weight
	)			`[virtual]`

add to normal

Parameters:

	idx	where to add
	weight	what to add

Reimplemented from CKernel.

Definition at line 248 of file CommUlongStringKernel.cpp.

void CCommUlongStringKernel::clear_normal ( ) [virtual]

clear normal

Reimplemented from CKernel.

Definition at line 312 of file CommUlongStringKernel.cpp.

void CCommUlongStringKernel::remove_lhs ( ) [virtual]

remove lhs from kernel

Reimplemented from CKernel.

Definition at line 41 of file CommUlongStringKernel.cpp.

void CCommUlongStringKernel::remove_rhs ( ) [virtual]

remove rhs from kernel

Reimplemented from CKernel.

Definition at line 61 of file CommUlongStringKernel.cpp.

virtual EFeatureType CCommUlongStringKernel::get_feature_type ( ) [virtual]

return feature type the kernel can deal with

Returns:: feature type ULONG

Reimplemented from CStringKernel< ST >.

Definition at line 189 of file CommUlongStringKernel.h.

void CCommUlongStringKernel::get_dictionary	(	INT &	dsize,
		ULONG *&	dict,
		DREAL *&	dweights
	)

get dictionary

Parameters:

	dsize	dictionary size will be stored in here
	dict	dictionary will be stored in here
	dweights	dictionary weights will be stored in here

Definition at line 197 of file CommUlongStringKernel.h.

DREAL CCommUlongStringKernel::compute	(	INT	idx_a,
		INT	idx_b
	)			`[protected, virtual]`

compute kernel function for features a and b idx_{a,b} denote the index of the feature vectors in the corresponding feature object

Parameters:

	idx_a	index a
	idx_b	index b

Returns:: computed kernel function at indices a,b

Implements CKernel.

Definition at line 164 of file CommUlongStringKernel.cpp.

DREAL CCommUlongStringKernel::normalize_weight	(	DREAL	value,
		INT	seq_num,
		INT	seq_len,
		ENormalizationType	p_normalization
	)			`[protected]`

normalize weight

Parameters:

	value	value
	seq_num	sequence number
	seq_len	length of sequence
	p_normalization	type of normalization

Definition at line 222 of file CommUlongStringKernel.h.

template<class ST>

virtual EFeatureClass CStringKernel< ST >::get_feature_class ( ) [virtual, inherited]

return feature class the kernel can deal with

Returns:: feature class STRING

Implements CKernel.

Definition at line 63 of file StringKernel.h.

DREAL CKernel::kernel	(	INT	idx_a,
		INT	idx_b
	)			`[inherited]`

get kernel function for lhs feature vector a and rhs feature vector b

Parameters:

	idx_a	index of feature vector a
	idx_b	index of feature vector b

Returns:: computed kernel function

Definition at line 48 of file Kernel.h.

void CKernel::get_kernel_matrix	(	DREAL **	dst,
		INT *	m,
		INT *	n
	)			`[inherited]`

get kernel matrix

Parameters:

	dst	destination where matrix will be stored
	m	dimension m of matrix
	n	dimension n of matrix

Definition at line 79 of file Kernel.cpp.

DREAL * CKernel::get_kernel_matrix_real	(	int &	m,
		int &	n,
		DREAL *	target
	)			`[virtual, inherited]`

get kernel matrix real

Parameters:

	m	dimension m of matrix
	n	dimension n of matrix
	target	the kernel matrix

Returns:: the kernel matrix

Definition at line 216 of file Kernel.cpp.

SHORTREAL * CKernel::get_kernel_matrix_shortreal	(	int &	m,
		int &	n,
		SHORTREAL *	target
	)			`[virtual, inherited]`

get kernel matrix shortreal

Parameters:

	m	dimension m of matrix
	n	dimension n of matrix
	target	target for kernel matrix

Returns:: the kernel matrix

Reimplemented in CCustomKernel.

Definition at line 146 of file Kernel.cpp.

bool CKernel::load ( CHAR * fname ) [inherited]

load the kernel matrix

Parameters:

fname

filename to load from

Returns:: if loading was succesful

Definition at line 322 of file Kernel.cpp.

bool CKernel::save ( CHAR * fname ) [inherited]

save kernel matrix

Parameters:

fname

filename to save to

Returns:: if saving was successful

Definition at line 327 of file Kernel.cpp.

CFeatures* CKernel::get_lhs ( ) [inherited]

get left-hand side of features used in kernel

Returns:: features of left-hand side

Definition at line 164 of file Kernel.h.

CFeatures* CKernel::get_rhs ( ) [inherited]

get right-hand side of features used in kernel

Returns:: features of right-hand side

Definition at line 170 of file Kernel.h.

INT CKernel::get_num_vec_lhs ( ) [inherited]

get number of vectors of lhs features

Returns:: number of vectors of left-hand side

Definition at line 176 of file Kernel.h.

INT CKernel::get_num_vec_rhs ( ) [inherited]

get number of vectors of rhs features

Returns:: number of vectors of right-hand side

Definition at line 188 of file Kernel.h.

bool CKernel::has_features ( ) [inherited]

test whether features have been assigned to lhs and rhs

Returns:: true if features are assigned

Definition at line 200 of file Kernel.h.

void CKernel::remove_lhs_and_rhs ( ) [virtual, inherited]

remove lhs and rhs from kernel

Definition at line 358 of file Kernel.cpp.

void CKernel::set_cache_size ( INT size ) [inherited]

set the size of the kernel cache

Parameters:

size

of kernel cache

Definition at line 249 of file Kernel.h.

int CKernel::get_cache_size ( ) [inherited]

return the size of the kernel cache

Returns:: size of kernel cache

Definition at line 259 of file Kernel.h.

void CKernel::list_kernel ( ) [inherited]

list kernel

Definition at line 389 of file Kernel.cpp.

bool CKernel::has_property ( EKernelProperty p ) [inherited]

check if kernel has given property

Parameters:

p

kernel property

Returns:: if kernel has given property

Definition at line 271 of file Kernel.h.

EOptimizationType CKernel::get_optimization_type ( ) [inherited]

get optimization type

Returns:: optimization type

Definition at line 289 of file Kernel.h.

virtual void CKernel::set_optimization_type ( EOptimizationType t ) [virtual, inherited]

set optimization type

Parameters:

t

optimization type to set

Reimplemented in CCombinedKernel.

Definition at line 295 of file Kernel.h.

bool CKernel::get_is_initialized ( ) [inherited]

check if optimization is initialized

Returns:: if optimization is initialized

Definition at line 301 of file Kernel.h.

bool CKernel::init_optimization_svm ( CSVM * svm ) [inherited]

initialize optimization

Parameters:

svm

svm model

Returns:: if initializing was successful

Definition at line 644 of file Kernel.cpp.

void CKernel::compute_batch	(	INT	num_vec,
		INT *	vec_idx,
		DREAL *	target,
		INT	num_suppvec,
		INT *	IDX,
		DREAL *	alphas,
		DREAL	factor = `1.0`
	)			`[virtual, inherited]`

computes output for a batch of examples in an optimized fashion (favorable if kernel supports it, i.e. has KP_BATCHEVALUATION. to the outputvector target (of length num_vec elements) the output for the examples enumerated in vec_idx are added. therefore make sure that it is initialized with ZERO. the following num_suppvec, IDX, alphas arguments are the number of support vectors, their indices and weights

Reimplemented in CCombinedKernel, CWeightedDegreePositionStringKernel, and CWeightedDegreeStringKernel.

Definition at line 568 of file Kernel.cpp.

DREAL CKernel::get_combined_kernel_weight ( ) [inherited]

get combined kernel weight

Returns:: combined kernel weight

Definition at line 346 of file Kernel.h.

void CKernel::set_combined_kernel_weight ( double nw ) [inherited]

set combined kernel weight

Parameters:

nw

Definition at line 352 of file Kernel.h.

INT CKernel::get_num_subkernels ( ) [virtual, inherited]

get number of subkernels

Returns:: number of subkernels

Reimplemented in CCombinedKernel, CWeightedDegreePositionStringKernel, and CWeightedDegreeStringKernel.

Definition at line 583 of file Kernel.cpp.

void CKernel::compute_by_subkernel	(	INT	vector_idx,
		DREAL *	subkernel_contrib
	)			`[virtual, inherited]`

compute by subkernel

Parameters:

	vector_idx	index
	subkernel_contrib	subkernel contribution

Reimplemented in CCombinedKernel, CWeightedDegreePositionStringKernel, and CWeightedDegreeStringKernel.

Definition at line 588 of file Kernel.cpp.

const DREAL * CKernel::get_subkernel_weights ( INT & num_weights ) [virtual, inherited]

get subkernel weights

Parameters:

num_weights

number of weights will be stored here

Returns:: subkernel weights

Reimplemented in CCombinedKernel, CWeightedDegreePositionStringKernel, and CWeightedDegreeStringKernel.

Definition at line 593 of file Kernel.cpp.

void CKernel::set_subkernel_weights	(	DREAL *	weights,
		INT	num_weights
	)			`[virtual, inherited]`

set subkernel weights

Parameters:

	weights	subkernel weights
	num_weights	number of weights

Reimplemented in CCombinedKernel, CWeightedDegreePositionStringKernel, and CWeightedDegreeStringKernel.

Definition at line 599 of file Kernel.cpp.

bool CKernel::get_precompute_matrix ( ) [inherited]

get precompute matrix

Returns:: if matrix shall be precomputed

Definition at line 386 of file Kernel.h.

bool CKernel::get_precompute_subkernel_matrix ( ) [inherited]

get precompute subkernel matrix

Returns:: if subkernel matrix shall be precomputed

Definition at line 392 of file Kernel.h.

virtual void CKernel::set_precompute_matrix	(	bool	flag,
		bool	subkernel_flag
	)			`[virtual, inherited]`

set precompute matrix

Parameters:

	flag	flag
	subkernel_flag	subkernel flag

Reimplemented in CCombinedKernel.

Definition at line 399 of file Kernel.h.

void CKernel::set_property ( EKernelProperty p ) [protected, inherited]

set property

Parameters:

p

kernel property to set

Definition at line 416 of file Kernel.h.

void CKernel::unset_property ( EKernelProperty p ) [protected, inherited]

unset property

Parameters:

p

kernel property to unset

Definition at line 425 of file Kernel.h.

void CKernel::set_is_initialized ( bool p_init ) [protected, inherited]

set is initialized

Parameters:

p_init

if optimization shall be set to initialized

Definition at line 434 of file Kernel.h.

void CKernel::do_precompute_matrix ( ) [protected, inherited]

matrix precomputation

Definition at line 606 of file Kernel.cpp.

void CKernel::init_sqrt_diag	(	DREAL *	v,
		INT	num
	)			`[protected, inherited]`

initialize sqrt diagonal

Parameters:

	v	v
	num	num

Definition at line 634 of file Kernel.cpp.

Member Data Documentation

DREAL* CCommUlongStringKernel::sqrtdiag_lhs [protected]

sqrt diagonal of left-hand side

Definition at line 253 of file CommUlongStringKernel.h.

DREAL* CCommUlongStringKernel::sqrtdiag_rhs [protected]

sqrt diagonal of right-hand side

Definition at line 255 of file CommUlongStringKernel.h.

bool CCommUlongStringKernel::initialized [protected]

if kernel is initialized

Definition at line 257 of file CommUlongStringKernel.h.

CDynamicArray<ULONG> CCommUlongStringKernel::dictionary [protected]

dictionary

Definition at line 260 of file CommUlongStringKernel.h.

CDynamicArray<DREAL> CCommUlongStringKernel::dictionary_weights [protected]

dictionary weights

Definition at line 262 of file CommUlongStringKernel.h.

bool CCommUlongStringKernel::use_sign [protected]

if sign shall be used

Definition at line 265 of file CommUlongStringKernel.h.

ENormalizationType CCommUlongStringKernel::normalization [protected]

type of normalization

Definition at line 267 of file CommUlongStringKernel.h.

INT CKernel::cache_size [protected, inherited]

cache_size in MB

Definition at line 464 of file Kernel.h.

KERNELCACHE_ELEM* CKernel::kernel_matrix [protected, inherited]

this *COULD* store the whole kernel matrix usually not applicable / faster

Definition at line 470 of file Kernel.h.

SHORTREAL* CKernel::precomputed_matrix [protected, inherited]

precomputed matrix

Definition at line 473 of file Kernel.h.

bool CKernel::precompute_subkernel_matrix [protected, inherited]

if subkernel matrix shall be precomputed

Definition at line 475 of file Kernel.h.

bool CKernel::precompute_matrix [protected, inherited]

if matrix shall be precomputed

Definition at line 477 of file Kernel.h.

CFeatures* CKernel::lhs [protected, inherited]

feature vectors to occur on left hand side

Definition at line 480 of file Kernel.h.

CFeatures* CKernel::rhs [protected, inherited]

feature vectors to occur on right hand side

Definition at line 482 of file Kernel.h.

DREAL CKernel::combined_kernel_weight [protected, inherited]

combined kernel weight

Definition at line 485 of file Kernel.h.

bool CKernel::optimization_initialized [protected, inherited]

if optimization is initialized

Definition at line 488 of file Kernel.h.

EOptimizationType CKernel::opt_type [protected, inherited]

optimization type (currently FASTBUTMEMHUNGRY and SLOWBUTMEMEFFICIENT)

Definition at line 492 of file Kernel.h.

ULONG CKernel::properties [protected, inherited]

properties

Definition at line 495 of file Kernel.h.

CParallel CSGObject::parallel [static, inherited]

Definition at line 105 of file SGObject.h.

CIO CSGObject::io [static, inherited]

Definition at line 106 of file SGObject.h.

CVersion CSGObject::version [static, inherited]

Definition at line 107 of file SGObject.h.

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

CCommUlongStringKernel Class Reference

Detailed Description

Public Member Functions

Static Public Attributes

Protected Member Functions

Protected Attributes

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation