read_graphml.cpp

一个用C语言编写的通用的隐性马尔可夫代码库

/*
 *
 * created: 26 Feb 2002 by Wasinee Rungsarityotin
 * authors: Wasinee Rungsarityotin ([e-mail])
 * file   : $Source: /cvsroot/ghmm/ghmm/ghmm++/examples/read_graphml.cpp,v $
 * $Id: read_graphml.cpp,v 1.5 2004/02/26 11:59:01 wasinee Exp $
 * revision date   : $Date: 2004/02/26 11:59:01 $
  _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
  _
 
 */

#include <stdio.h>
// #include <unistd.h>
#include <assert.h>
#include <ghmm/rng.h>
#include <ghmm/ghmm.h>
#include "ghmm++/GHMM.h"
#include "ghmm++/GHMM_GMLDoc.h"
#include <iostream>

#ifdef HAVE_NAMESPACES
using namespace std;
#endif

int seq_d_class(int index, int t) {
  printf(" c:%i ",index);
  if (index>5)
    return 1;
  else
    return 0; 
}

int main(int argc, char* argv[]) {
  //GHMM_Document      doc;
  GHMM_GraphMLDoc    doc;
  GHMM_SWDiscreteModel *dmo;
  GHMM_Sequences* training_seq = NULL;
  GHMM_Sequences* my_output = NULL;
  GHMM_Alphabet*  alphas = NULL;

  /* Important! initialise rng  */
  std::GHMM_Toolkit::gsl_rng_init();

  if (argc < 2)
    {
      fprintf(stderr, "Please input a filename. Usage: read_graphml file.gml\n");
      exit(-1);
    }
  doc.open(argv[1], "r");
  doc.readDocument();
  doc.close();

  dmo  = doc.getDiscreteModel();
  dmo->create_cmodel(0);
  //dmo->setTransition(0,0,0.0);
  //dmo->setTransition(0,1,1.0);

  if (dmo->check() == -1)
    return 1;

  fprintf(stdout, "initial distribution:\n");
  dmo->Pi_print(stdout,"  ",",","\n");
  fprintf(stdout,"transition matrix:\n");
  dmo->A_print(stdout,""," ","\n");
  fprintf(stdout,"observation symbol matrix:\n");
  dmo->B_print(stdout,""," ","\n");

  alphas = dmo->getAlphabet();
  assert( alphas != NULL );
  for(int i=0; i < alphas->size(); i++)
    {
      cout << "Symbol " << i << ":";
      cout << alphas->getSymbol(i) << endl;
    }

  if ( ( training_seq = doc.getSequences() ) != NULL )
    {
      training_seq->print(stdout);
      
      printf("reestimating with Baum-Welch-algorithm...");
      // dmo->reestimate_baum_welch( training_seq );

      /* print the result */
      printf("Done\nthe result is:\n");  
      dmo->A_print(stdout,""," ","\n");
      fprintf(stdout,"observation symbol matrix:\n");
      dmo->B_print(stdout,""," ","\n");
    }

  sequence_t* seqs;
  
  
  //my_output = dmo->generate_sequences(1,10,10,10);
  seqs = sdmodel_generate_sequences(dmo->c_model,1,10,10,20);  	
  sequence_print(stdout,seqs);
  
  printf("Pause....\nPlease check your sequences before continuing:\n");
  

  std::GHMM_GMLClass *dco = new std::GHMM_GMLClass( doc.getClass() );
  
  doc.open("gml2.xml","w"); // Try writing to a new format!!!!
  doc.writeElement(dco); 
  doc.writeEndl();
  doc.writeElement(dmo->getAlphabet());
  doc.writeEndl();
  doc.writeElement(dmo);
  doc.writeEndl();
  doc.close();
  SAFE_DELETE( dco );
  
  doc.open(stdout, "w");
  doc.writeEndl();
  // doc.writeElement(my_output);
  doc.writeEndl();
  doc.close();
  
  cout << "Wrote to gml2.xml" << endl;
  
  delete my_output;
  delete alphas;
  
  // SAFE_DELETE(dmo);
  
  return 0;
}