meta::sequence::crf::scorer Class Reference

Internal class that holds scoring information for sequences under the current model. More...

#include <scorer.h>

Public Member Functions
void	score (const crf &model, const sequence &seq)
	Finds both transition and state scores, in logarithm and exponential domains. More...
void	transition_scores (const crf &model)
	Finds only the transition scores. More...
void	state_scores (const crf &model, const sequence &seq)
	Finds only the state scores. More...
void	forward ()
	Computes the forward trellis.
void	backward ()
	Computes the backward trellis.
void	marginals ()
	Computes the marginal probabilities of states and transitions by using the forward-backward algorithm results.
double	state (uint64_t time, label_id lbl) const
double	state_exp (uint64_t time, label_id lbl) const
double	trans (label_id from, label_id to) const
double	trans_exp (label_id from, label_id to) const
double	forward (uint64_t time, label_id lbl) const
double	backward (uint64_t time, label_id lbl) const
double	state_marginal (uint64_t time, label_id lbl) const
double	trans_marginal (label_id from, label_id to) const
double	loss (const sequence &seq) const
	Computes the loss function for a given sequence. More...

Private Member Functions
void	transition_marginals ()
	Computes the transition marginals.
void	state_marginals ()
	Computes the state marginals.

Private Attributes
double_matrix	state_
	Stores the state scores in log-domain.
double_matrix	state_exp_
	Stores the state scores.
double_matrix	trans_
	Stores the transition scores in log-domain.
double_matrix	trans_exp_
	Stores the transition scores.
util::optional< forward_trellis >	fwd_
	Stores the forward trellis, if computed.
util::optional< trellis >	bwd_
	Stores the backward trellis, if computed.
util::optional< double_matrix >	state_mrg_
	Stores the state marginals, if computed.
util::optional< double_matrix >	trans_mrg_
	Stores the transition marginals, if computed.

Detailed Description

Internal class that holds scoring information for sequences under the current model.

Member Function Documentation

void meta::sequence::crf::scorer::score	(	const crf &	model,
		const sequence &	seq
	)

Finds both transition and state scores, in logarithm and exponential domains.

Parameters

model	The model to score with
seq	The sequence to score

void meta::sequence::crf::scorer::transition_scores

(

const crf &

model

)

Finds only the transition scores.

Parameters

model

The model to score with

void meta::sequence::crf::scorer::state_scores	(	const crf &	model,
		const sequence &	seq
	)

Finds only the state scores.

Parameters

model	The model to score with
seq	The sequence to score

double meta::sequence::crf::scorer::state	(	uint64_t	time,
		label_id	lbl
	)		const

Parameters

time	The time step
lbl	The state

Returns

the log-domain score for being in this state at the given time

double meta::sequence::crf::scorer::state_exp	(	uint64_t	time,
		label_id	lbl
	)		const

Parameters

time	The time step
lbl	The state

Returns

the score for being in this state at the given time

double meta::sequence::crf::scorer::trans	(	label_id	from,
		label_id	to
	)		const

Parameters

from	The origin state
to	The destination state

Returns

the log-domain score for the given transition

double meta::sequence::crf::scorer::trans_exp	(	label_id	from,
		label_id	to
	)		const

Parameters

from	The origin state
to	The destination state

Returns

the score for the given transition

double meta::sequence::crf::scorer::forward	(	uint64_t	time,
		label_id	lbl
	)		const

Parameters

time	The time step
lbl	The state

Returns

the forward score for the given state at the given time

double meta::sequence::crf::scorer::backward	(	uint64_t	time,
		label_id	lbl
	)		const

Parameters

time	The time step
lbl	The state

Returns

the backward score for the given state at the given time

double meta::sequence::crf::scorer::state_marginal	(	uint64_t	time,
		label_id	lbl
	)		const

Parameters

time	The time step
lbl	The state

Returns

the marginal probability of being in the given state at the given time

double meta::sequence::crf::scorer::trans_marginal	(	label_id	from,
		label_id	to
	)		const

Parameters

from	The origin state
to	The destination state

Returns

the marginal probability of transitioning from the origin state to the destination state

double meta::sequence::crf::scorer::loss

(

const sequence &

seq

)

const

Computes the loss function for a given sequence.

Parameters

seq	The reference sequence

Returns

the value of the loss function with respect to this sequence

---

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

• /home/chase/projects/meta/include/sequence/crf/scorer.h
• /home/chase/projects/meta/src/sequence/crf/scorer.cpp