train_hmm.cpp

来自「一个用于 词性标注的 HMM程序。 包含 训练和测试功能。」· C++ 代码 · 共 73 行

#include <stdio.h>
#include <stdlib.h>
#include <iostream.h>
#include <fstream.h>
#include <string.h>
#include "hmm.h"

void main(int argc, char* argv[])

{
  // get command line parms

  char* train_file=argv[1];
  char* initial_model;
  double min_delta_psum;
  int states,symbols,seed;

  // check args
  if (argc != 4 && argc != 6) {
    cerr << "ERROR: Too few arguments.\n\n" << 
    "Usage: "<< argv[0]<<" <train_file> <hmm_model> <min_delta_psum> or \n" <<
    "       "<< argv[0]<<" <train_file> <seed> <nstates> <nsymbols> <min_delta_psum>\n\n";
    exit (-1);
  }
  
  if (argc==4) {
    initial_model=argv[2];
    min_delta_psum=atof(argv[3]);
  }

  if (argc==6) {
    seed = atoi(argv[2]);
    states = atoi(argv[3]);
    symbols = atoi(argv[4]);
    min_delta_psum=atof(argv[5]);
  }

  HMM *hmm;

   //测试模型是否能够正确被保存 并读取
   //hmm=new HMM("test.hmm.seq.hmm");
   // hmm->dump_model("test.hmm.seq.hmm1");


  if (argc==4) {
    // initialize model from model file
    hmm = new HMM(initial_model);
    // train model on data
    hmm->batch_train(train_file,min_delta_psum);
  }
  else {
    // initialize model randomly
    hmm = new HMM(symbols, states, seed);
    // train model on data
    hmm->batch_train(train_file,min_delta_psum);
  }

  ///////////////////////////测试viterbi
   int *out =new int[hmm->string_len[0]];
   hmm->viterbi(hmm->strings[0],hmm->string_len[0],&out);
  ///////////////////////////测试viterbi结束
  
  // dump the resulting model to a file
  char newfilename[100];
  sprintf(newfilename,"%s.hmm",train_file);
  hmm->dump_model(newfilename);
  
  delete hmm;
}