Markov Class Reference

Markov modelling, estimation and simulation. More...

#include <seqpp/Markov.h>

Inheritance diagram for Markov:

List of all members.

Public Member Functions
	Markov (const char *ConfFile, bool calc_rank=false)
	Constructor 1 : read a configuration file.
	Markov (const SequenceSet &seqset, bool calc_rank=false, const string &prior_alpha_file=string())
	Constructor 2 : Estimate the transition matrices on the sequences of seqset.
	Markov (const Sequence &seq, bool calc_rank=false, const string &prior_alpha_file=string())
	Constructor 3 : Estimate the transition matrices on the sequence seq.
	Markov (const Markov &)
	Constructor 4 : Copy constructor.
	Markov ()
	Constructor 5 : Default constructor.
	Markov (short size, short order, bool alloc=true, const string &prior_alpha_file=string())
	Constructor 6 : Minimal Constructor.
	Markov (const Markov &M1, const Markov &M2, const float p)
	Constructor 7 : Creation of a mixed Markov chain M = p*M1 + (1-p)*M2 */.
	Markov (const gsl_rng *r, short size, short order, bool calc_rank=false)
	Constructor 8 : random markov matrices.
	Markov (unsigned long *count, short size, short order, const string &prior_alpha_file=string(), bool calc_rank=false)
	Constructor 9 : Estimate the transition matrices on a word-count.
virtual	~Markov ()
	Destructor.
template<class TSeq>
void	estimate (const TSeq &tseq, unsigned long beg, unsigned long end, bool calc_rank)
	Estimate the transition matrices on the sequence/sequenceset tseq.
void	estimate (unsigned long *count, bool decal_required, bool calc_rank=false)
	Estimate the transition matrices from a word count.
void	estimate (const string &count_file, bool calc_rank=false)
	Estimate the transition matrices from a file containing the count.
const double *	markov_matrix () const
	access to the markov matrix
void	draw_markov_matrix (const gsl_rng *r)
	draw at random the markov matrix
void	free_markov_matrix ()
	free the memory allocated for markov matrix
void	compute_stat_law (bool force=false)
	Compute the stationnary laws.
void	free_stat_law ()
	free the memory allocated for stationnary law
const double *	stat_law () const
	access to the stationnary distrib
virtual int	compute_rank ()
	Computes the rank of convergence of the Markov Chain.
void	compute_power ()
	Initialisation the _rank powers of the markov matrix.
int	free_power ()
	free the memory allocated for the power
double	proba_step (long w1, long w2, int step)
	transition from word w1 to word w2 in step steps
bool	isPi () const
bool	isPow () const
bool	isMu () const
double &	operator() (int index)
	() operator for Markov matrix Pi elements
double	Mu (int index) const
	Access to stationnary vector Mu elements.
Protected Attributes
double *	_Pi
	Pointer to "Matrix" (in a vector format) of transition probabilities (=_Pis[0]).
double *	_container
	Container of "Matrix" (in a vector format) of transition probabilities (=_containers[0]).
double *	_Mu
	Vector of stationnary probabilities (=_Mus[0]).
double ***	_PowPi
	Power of the _Pi matrix.

---

Detailed Description

Markov modelling, estimation and simulation.

This is a special case of a phased Markov [PhasedMarkov] model when only one phase is considered.

---

Constructor & Destructor Documentation

Markov::Markov	(	const SequenceSet &	seqset,
		bool	calc_rank = false,
		const string &	prior_alpha_file = string()
	)

Constructor 2 : Estimate the transition matrices on the sequences of seqset.

Parameters:

	seqset	set of sequences for estimation
	calc_rank	calculus of the convergence rank if true
	prior_alpha_file	file containing the alpha for the a priori law, one value per alphabet element

References _container, PhasedMarkov::_containers, _Mu, PhasedMarkov::_Mus, _Pi, PhasedMarkov::_Pis, and _PowPi.

Markov::Markov	(	const Sequence &	seq,
		bool	calc_rank = false,
		const string &	prior_alpha_file = string()
	)

Constructor 3 : Estimate the transition matrices on the sequence seq.

Parameters:

	seq	sequence for estimation
	calc_rank	calculus of the convergence rank if true
	prior_alpha_file	file containing the alpha for the a priori law, one value per alphabet element

References _container, PhasedMarkov::_containers, _Mu, PhasedMarkov::_Mus, _Pi, PhasedMarkov::_Pis, and _PowPi.

Markov::Markov	(	short	size,
		short	order,
		bool	alloc = true,
		const string &	prior_alpha_file = string()
	)			[inline]

Constructor 6 : Minimal Constructor.

Initialises the constants of the model but not the matrix nor the stat law

Parameters:

	size	alphabet size
	order	markovian order
	alloc	true for matrices memory allocation
	prior_alpha_file	file containing the alpha for the a priori law, one value per alphabet element

References _container, PhasedMarkov::_containers, _Mu, _Pi, PhasedMarkov::_Pis, and _PowPi.

Markov::Markov	(	const Markov &	M1,
		const Markov &	M2,
		const float	p
	)

Constructor 7 : Creation of a mixed Markov chain M = p*M1 + (1-p)*M2 */.

Parameters:

	M1	first Markov chain object
	M2	second Markov chain object
	p	weight of M1 in the resulting M(with 0<=p<=1)

References _container, PhasedMarkov::_containers, _Mu, PhasedMarkov::_Mus, _Pi, PhasedMarkov::_Pis, and _PowPi.

Markov::Markov	(	const gsl_rng *	r,
		short	size,
		short	order,
		bool	calc_rank = false
	)			[inline]

Constructor 8 : random markov matrices.

Parameters:

	r	gsl random generator
	size	alphabet size
	order	markovian order
	calc_rank	calculus of the convergence rank if true GSL use exple: const gsl_rng_type * T; // Choice a default generator and seed // from environment variables gsl_rng_env_setup(); // New created instance of the generator T = gsl_rng_default; gsl_rng * r = gsl_rng_alloc (T); // Initialize/Seeds the random number generator gsl_rng_set( r, (long)getpid() );

References _container, PhasedMarkov::_containers, _Mu, _Pi, PhasedMarkov::_Pis, and _PowPi.

Markov::Markov	(	unsigned long *	count,
		short	size,
		short	order,
		const string &	prior_alpha_file = string(),
		bool	calc_rank = false
	)

Constructor 9 : Estimate the transition matrices on a word-count.

Parameters:

	count	count of all the coded word(base size) of size order+1 for estimation
	size	alphabet size
	order	markovian order
	calc_rank	calculus of the convergence rank if true
	prior_alpha_file	file containing the alpha for the a priori law, one value per alphabet element

References _container, PhasedMarkov::_containers, _Mu, PhasedMarkov::_Mus, _Pi, PhasedMarkov::_Pis, and _PowPi.

---

Member Function Documentation

template<class TSeq>

void Markov::estimate	(	const TSeq &	tseq,
		unsigned long	beg,
		unsigned long	end,
		bool	calc_rank
	)			[inline]

Estimate the transition matrices on the sequence/sequenceset tseq.

Parameters:

	tseq	sequence/sequenceset for estimation
	beg	begin position in sequence(s) if subsequences
	end	end position in sequence(s) if subsequences
	calc_rank	calculus of the convergence rank if true

References PhasedMarkov::estimate().

void Markov::estimate	(	unsigned long *	count,
		bool	decal_required,
		bool	calc_rank = false
	)			[inline]

Estimate the transition matrices from a word count.

Parameters:

	count	count of all the coded word(base size) of size order+1, for estimation
	decal_required	necessary when using a count of word from 1-word to (_order+1)-word
	calc_rank	calculus of the convergence rank if true

References PhasedMarkov::estimate().

void Markov::estimate	(	const string &	count_file,
		bool	calc_rank = false
	)			[inline]

Estimate the transition matrices from a file containing the count.

Parameters:

	count_file	file containing the count in the adapted format
	calc_rank	calculus of the convergence rank if true

Reimplemented from PhasedMarkov.

References PhasedMarkov::estimate().

void Markov::draw_markov_matrix

(

const gsl_rng *

r

)

[inline]

draw at random the markov matrix

Parameters:

r

gsl random generator

GSL use exple:

    const gsl_rng_type * T;
    // Choice a default generator and seed
    // from environment variables
    gsl_rng_env_setup();
    // New created instance of the generator
    T = gsl_rng_default;
    gsl_rng * r = gsl_rng_alloc (T);
    // Initialize/Seeds the random number generator
    gsl_rng_set( r, (long)(time( NULL )) );
    ...
    ...
    gsl_rng_free( r );

References PhasedMarkov::draw_markov_matrices().

double Markov::proba_step	(	long	w1,
		long	w2,
		int	step
	)			[inline]

transition from word w1 to word w2 in step steps

-a Translator object is required to process string to int, use link_to_translator- if step > _rank, give stat_law(w2). You must have w1, w2 of length order

Parameters:

	w1	first word (of length order) as a string
	w2	second word (of length order) as a string transition from word w1 to word w2 in step steps

if step > _rank, give stat_law(w2). You must have w1, w2 of length order

Parameters:

	w1	first word (of length order) as a vector of short
	w2	second word (of length order) as a vector of short transition from word w1 to word w2 in step steps

if step > _rank, give stat_law(w2). You must have w1-jump, w2-jump < _nMu

Parameters:

	w1	first word (of length order) as a Sequence-coded-like integer (see Sequence)
	w2	second word (of length order) as a Sequence-coded-like integer (see Sequence)
	step	number of steps between w1 and w2

References _Mu, _PowPi, PhasedMarkov::_rank, compute_power(), and compute_rank().

bool Markov::isPi

(

)

const [inline]

_Pi == NULL ?

References _Pi.

bool Markov::isPow

(

)

const [inline]

_PowPi == NULL ?

References _PowPi.

Referenced by free_power(), and Markov().

bool Markov::isMu

(

)

const [inline]

_Mu == NULL ?

References _Mu.

double& Markov::operator()

(

int

index

)

[inline]

() operator for Markov matrix Pi elements

Parameters:

index

index of the word

References _Pi.

double Markov::Mu

(

int

index

)

const [inline]

Access to stationnary vector Mu elements.

Parameters:

index

index of the word

References _Mu.

---

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

• Markov.h
• Markov.cc