scluster.h

一个通用的隐性马尔可夫C代码库 开发环境:C语言 简要说明:这是一个通用的隐性马尔可夫C代码库

/*-----------------------------------------------------------------------------
  author       : Bernhard Knab, Benjamin Georgi
  filename     : ghmm/ghmm/scluster.h
  created      : TIME: 15:53:53     DATE: Tue 16. November 1999
  $Id: scluster.h,v 1.9 2003/10/10 13:10:13 cic99 Exp $

Copyright (C) 1998-2001, ZAIK/ZPR, Universit鋞 zu K鰈n

This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA


------------------------------------------------------------------------------*/


#ifndef SCLUSTER_H
#define SCLUSTER_H
#include <ghmm/sequence.h>
#include <ghmm/smodel.h>
#include <ghmm/sreestimate.h>


#ifdef __cplusplus
extern "C" {
#endif

/**
   @name scluster
 */

/*@{
 */

#define CLASSIFY 0 /* Switch for Classificator: 0 == MD, 1 == MAW */

/**
   Cluster structure: All models and sequences. */
struct scluster_t{
  /** 
  Vector of SHMMs pointers */
  smodel **smo;
  /** 
    Vector of sequence_t pointers; to store the sequences, that 
    belong to the models */
  sequence_d_t **smo_seq;
  /** 
    Number of models to read in */
  int smo_number;
  /** 
    Number of sequences for each model */
  long *seq_counter;
  /** 
    log(P) for the model */
  double *smo_Z_MD;
  /** a posteriori probability for the Modelle to calculate the objective
      fucntion in case of a MAW classificator. Is calculated using smap_bayes */
  double *smo_Z_MAW;
};
/**
 */
typedef struct scluster_t scluster_t;


/**
   Frees the memory allocated for a scluster_t struct.
   @return 0 for success; -1 for error
   @param scl pointer to scl struct
*/
int scluster_t_free(scluster_t *scl);


/**
   Writes out the final models.
   @return 0 for success; -1 for error
   @param cl cluster of models to write
   @param sqd
   @param outfile output file
   @param out_filename name of the output file
 */
int scluster_out(scluster_t *cl, sequence_d_t *sqd, FILE *outfile,
		 char *argv[]);

/**
   Avoids empty models going out as outputs by assigning them a random 
   sequence. This may lead to a produce of a new empty model - therefore
   change out sequences until a non empty model is found. (Quit after 100 
   iterations to avoid an infinite loop). 
   @return 0 for success; -1 for error
   @param sqd sequences to generate the model from
   @param cl cluster for the models
 */
int scluster_avoid_empty_smodel(sequence_d_t *sqd, scluster_t *cl);

/**
   Makes a cluster and reestimates the HMMs.
   @return 0 for success; -1 for error
   @param argv vector of input files, one with sequences, one with models, 
   one for output and one with labels for the sequences - in this order.
 */
int scluster_hmm(char *argv[]);

/**
   Updates the cluster with additional sequences.
   @return 0 for success; -1 for error
   @param cl cluster to update
   @param sqd sequences to update the cluster with
 */
int scluster_update(scluster_t *cl, sequence_d_t *sqd);

/**
   Updates a label
   @return number of changes made
   @param oldlabel label to update
   @param up to date label for comparison
   @param seq_number number of sequences
   @param smo_changed tells, which labels have been changed
 */
long scluster_update_label(long *oldlabel, long *seq_label, long seq_number, 
			   long *smo_changed);

/**
   Prints out the likelihood values for the cluster.
   @param outfile output file
   @param cl cluster of models and sequences
 */
void scluster_print_likelihood(FILE *outfile, scluster_t *cl);

/** 
    Determines form an already calculated probability matrix, which model 
    fits best to a certain sequence. 
    @return index of best model if success, otherwize -1
    @param cl cluster 
    @param seq_id ID of the sequence in question
    @param all_log_p matrix containing the probability of each sequence
    for each model
    @param log_p the probability of the sequence in question for the 
    best fitting model
*/
int scluster_best_model(scluster_t *cl, long seq_id, double **all_log_p,
			double *log_p);

/**
   Calculates the logarithmic probability of sequences for a model.
   @param cs sequences and model 
 */
void scluster_prob(smosqd_t *cs);

/* int scluster_labels_from_kmeans(sequence_d_t *sqd, int smo_number); */

/**
   Creates random labels for a vector of sequences
   @return 0 for success; -1 for error
   @param sqd vector of sequences
   @param smo_number number of models (needed to determine the interval
   for the random numbers)
*/
int scluster_random_labels(sequence_d_t *sqd, int smo_number);

/** Calculates the aposteriori prob. $\log(p(\lambda_best | O[seq\_id]))$, 
    where $\lambda_best$ is the model with highest apost. prob.
    @return 0 for success; -1 for error
    @param cl cluster
    @param sqd the sequence in question
    @param seq_id the ID of the sequence
    @param lob_apo the results
*/
int  scluster_log_aposteriori(scluster_t *cl, sequence_d_t *sqd, int seq_id, 
			      double *log_apo);

/**
   Prints the input vector for scluster_hmm
 */
void scluster_print_header(FILE *file, char* argv[]);

#ifdef __cplusplus
}
#endif


/*@} */

#endif /* SCLUSTER_H */