MosesTuning::MiraWeightVector Class Reference

#include <MiraWeightVector.h>

List of all members.

Public Member Functions
	MiraWeightVector ()
	MiraWeightVector (const std::vector< ValType > &init)
void	update (const MiraFeatureVector &fv, float tau)
void	tick ()
ValType	score (const MiraFeatureVector &fv) const
ValType	sqrNorm () const
AvgWeightVector	avg ()
void	ToSparse (SparseVector *sparse, size_t denseSize) const
Friends
class	AvgWeightVector
std::ostream &	operator<< (std::ostream &o, const MiraWeightVector &e)

---

Detailed Description

Definition at line 24 of file MiraWeightVector.h.

---

Constructor & Destructor Documentation

MosesTuning::MiraWeightVector::MiraWeightVector

(

)

Constructor, initializes to the zero vector

Definition at line 14 of file MiraWeightVector.cpp.

MosesTuning::MiraWeightVector::MiraWeightVector

(

const std::vector< ValType > &

init

)

Constructor with provided initial vector

Parameters:

init

Initial feature values

---

Member Function Documentation

AvgWeightVector MosesTuning::MiraWeightVector::avg

(

)

Return an averaged view of this weight vector

Definition at line 71 of file MiraWeightVector.cpp.

References AvgWeightVector.

ValType MosesTuning::MiraWeightVector::score

(

const MiraFeatureVector &

fv

)

const

Score a feature vector according to the model

Parameters:

fv

Feature vector to be scored

Definition at line 59 of file MiraWeightVector.cpp.

References MosesTuning::MiraFeatureVector::feat(), MosesTuning::MiraFeatureVector::size(), and MosesTuning::MiraFeatureVector::val().

Referenced by MosesTuning::NbestHopeFearDecoder::HopeFear().

Here is the call graph for this function:

Here is the caller graph for this function:

ValType MosesTuning::MiraWeightVector::sqrNorm

(

)

const

Squared norm of the weight vector

Definition at line 131 of file MiraWeightVector.cpp.

void MosesTuning::MiraWeightVector::tick

(

)

Perform an empty update (affects averaging)

Definition at line 50 of file MiraWeightVector.cpp.

void MosesTuning::MiraWeightVector::ToSparse	(	SparseVector *	sparse,
		size_t	denseSize
	)			const

Convert to sparse vector, interpreting all features as sparse. Only used by hgmira.

Definition at line 96 of file MiraWeightVector.cpp.

References MosesTuning::SparseVector::set().

Here is the call graph for this function:

void MosesTuning::MiraWeightVector::update	(	const MiraFeatureVector &	fv,
		float	tau
	)

Update a the model

Parameters:

	fv	Feature vector to be added to the weights
	tau	FV will be scaled by this value before update

Definition at line 39 of file MiraWeightVector.cpp.

References MosesTuning::MiraFeatureVector::feat(), MosesTuning::MiraFeatureVector::size(), update(), and MosesTuning::MiraFeatureVector::val().

Referenced by update().

Here is the call graph for this function:

Here is the caller graph for this function:

---

Friends And Related Function Documentation

friend class AvgWeightVector [friend]

Definition at line 71 of file MiraWeightVector.h.

Referenced by avg().

std::ostream& operator<<	(	std::ostream &	o,
		const MiraWeightVector &	e
	)			[friend]

---

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

• /disk4/html/www/moses/doxygen/mosesdecoder/mert/MiraWeightVector.h
• /disk4/html/www/moses/doxygen/mosesdecoder/mert/MiraWeightVector.cpp