refl_02.cc

这是一个从音频信号里提取特征参量的程序

    }
    
    // compute the reflection coefficients using the autocorrelation
    // coefficients via the DURBAL algorithm with order M
    //
    reflc.set(AUTOCORRELATION, DURBIN, M, -200);
    reflc.compute(refl_coef, auto_coeff);    
    
    if (!refl_coef.almostEqual(res_refl_coef_autoc)) {
      refl_coef.debug(L"refl_coef");
      res_refl_coef_autoc.debug(L"res_refl_coef");
      return Error::handle(name(), L"compute", ERR, __FILE__,
			   __LINE__);
    }

    // since the order for computing the reflection coefficients is M
    // we sould get M reflection coefficients
    //
    if (refl_coef.length() != M) {
      return Error::handle(name(), L"diagnose", Error::TEST, __FILE__, __LINE__);
    }

    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    // case: Algorithm = AUTOCORRELATION, Implementation = DURBIN
    //       Input = zero CORRELATION vector
    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
        
    // test the algorithm for a vector of zeros
    //
    auto_coeff.assign(0.0);
    reflc.compute(refl_coef, auto_coeff);

    if (!refl_coef.almostEqual(zeros)) {
      return Error::handle(name(), L"diagnose", Error::TEST, __FILE__, __LINE__);
    }
    
    refl_coef.clear();

    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    // case: Algorithm = COVARIANCE, Implementation = CHOLESKY
    //       Input = non-zero COVARIANCE matrix
    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    
    // compute the covariance coefficients of order M
    //
    Covariance covar(Covariance::DEF_ALGORITHM, Covariance::DEF_IMPLEMENTATION,
		     Covariance::DEF_NORMALIZATION, M);
    MatrixFloat covar_coeff;
    covar.compute(covar_coeff, input);

    // compute the reflection coefficients using the covariance
    // coefficients via the CHOLESKY algorithm with order M
    //
    reflc.set(COVARIANCE, CHOLESKY, M, -200);
    reflc.compute(refl_coef, covar_coeff, AlgorithmData::COVARIANCE);
    
    if (!refl_coef.almostEqual(res_refl_coef_cov)) {
      res_pred_coef_cov.debug(L"res_pred_coef");
      refl_coef.debug(L"refl_coef");
      res_refl_coef_cov.debug(L"res_refl_coef");
      return Error::handle(name(), L"compute", ERR, __FILE__, __LINE__);
      }

    // since the order for computing the reflection coefficients is M
    // we sould get M reflection coefficients
    //
    if (refl_coef.length() != M) {
      return Error::handle(name(), L"diagnose", Error::TEST, __FILE__, __LINE__);
    }    

    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    // case: Algorithm = COVARIANCE, Implementation = CHOLESKY
    //       Input = zero COVARIANCE matrix
    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

    // test the algorithm for a vector of zeros
    //
    covar_coeff.assign(0);
    reflc.compute(refl_coef, covar_coeff, AlgorithmData::COVARIANCE);    

    if (!refl_coef.almostEqual(zeros)) {
      return Error::handle(name(), L"diagnose", Error::TEST, __FILE__, __LINE__);
    }

    refl_coef.clear();

    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    // case: Algorithm = LATTICE, Implementation = BURG
    //       Input = non-zero SIGNAL vector
    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

    // compute the reflection coefficients using the BURG algorithm via
    // a lattice with order M
    //
    reflc.set(LATTICE, BURG, M, -200);
    reflc.compute(refl_coef, input, AlgorithmData::SIGNAL);
  
    if (!refl_coef.almostEqual(res_refl_coef_burg)) {
      refl_coef.debug(L"pred_coef");
      res_refl_coef_burg.debug(L"res_refl_coef");
      return Error::handle(name(), L"compute", ERR, __FILE__, __LINE__);    
    }

    // since the order for computing the reflection coefficients is M
    // we sould get M reflection coefficients
    //
    if (refl_coef.length() != M) {
      return Error::handle(name(), L"diagnose", Error::TEST, __FILE__, __LINE__);
    }

    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    // case: Algorithm = LATTICE, Implementation = BURG
    //       Input = zero SIGNAL vector
    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

    // test the algorithm for a vector of zeros
    //
    input.assign(0.0);
    reflc.compute(refl_coef, input, AlgorithmData::SIGNAL);    

    if (!refl_coef.almostEqual(zeros)) {
      return Error::handle(name(), L"diagnose", Error::TEST, __FILE__, __LINE__);
    }
    
    refl_coef.clear();
    
    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    // case: Algorithm = PREDICTION, Implementation = STEP_DOWN
    //       Input = non-zero PREDICTION vector
    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    
    // compute the reflection coefficients from the predictor coefficients
    //
    pred_coef.assign(L"1.000000, -0.9666891, 0.00009353, 0.00009286241, 0.03525937");
    reflc.setAlgorithm(PREDICTION);
    reflc.setImplementation(STEP_UP);
    reflc.compute(refl_coef, pred_coef, AlgorithmData::PREDICTION);

    VectorFloat res_refl_coef;
    res_refl_coef.assign(L"-0.9378492, 0.03325092, 0.03422026, 0.03525937");
    
    if (!refl_coef.almostEqual(res_refl_coef, .3)) {
      refl_coef.debug(L"refl_coef from predictor");
      return Error::handle(name(), L"computeFromRefelction", ERR, __FILE__,
			   __LINE__);
    }

    // since we have used M + 1  predictor coefficients to compute the
    // reflection coefficients we should get M reflection coefficients
    //
    if (refl_coef.length() != M) {
      return Error::handle(name(), L"diagnose", Error::TEST, __FILE__, __LINE__);
    }

    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    // case: Algorithm = PREDICTION, Implementation = STEP_DOWN
    //       Input = zero PREDICTION vector
    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

    // test the algorithm for a vector of zeros
    //
    pred_coef.assign(0.0);
    reflc.compute(refl_coef, pred_coef, AlgorithmData::PREDICTION);

    if (!refl_coef.almostEqual(zeros)) {
      return Error::handle(name(), L"diagnose", Error::TEST, __FILE__, __LINE__);
    }
    
    refl_coef.clear();
    
    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    // case: Algorithm = LOG_AREA_RATIO, Implementation = KELLY_LOCHBAUM
    //       Input = non-zero LOG_AREA_RATIO vector
    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    
    // compute the reflection coefficients from the log area ratio coefficients
    //
    VectorFloat log_arr_coef(L"3.43977, -0.0665264, -0.0684673, -0.070548");

    reflc.setAlgorithm(LOG_AREA_RATIO);
    reflc.setImplementation(KELLY_LOCHBAUM);
    reflc.compute(refl_coef, log_arr_coef, AlgorithmData::LOG_AREA_RATIO);

    if (!refl_coef.almostEqual(refl_coef, .3)) {
      refl_coef.debug(L"refl_coef from log_arr");
      return Error::handle(name(), L"computeFromLogArr", ERR,
			   __FILE__, __LINE__);
    }

    // since we have used M log area ratio coefficients to compute the
    // reflection coefficients we should get M reflection coefficients
    //
    if (refl_coef.length() != M) {
      return Error::handle(name(), L"diagnose", Error::TEST, __FILE__, __LINE__);
    }

    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    // case: Algorithm = LOG_AREA_RATIO, Implementation = KELLY_LOCHBAUM
    //       Input = zero LOG_AREA_RATIO vector
    // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
    
    // test the algorithm for a vector of zeros
    //
    log_arr_coef.assign(0.0);
    reflc.compute(refl_coef, log_arr_coef, AlgorithmData::LOG_AREA_RATIO);    

    if (!refl_coef.almostEqual(zeros)) {
      return Error::handle(name(), L"diagnose", Error::TEST, __FILE__, __LINE__);
    }

    refl_coef.clear();
  }

  // reset indentation
  //
  if (level_a > Integral::NONE) {
    Console::decreaseIndention();
  }

  // --------------------------------------------------------------------
  //
  // 5. class-specific public methods
  //     apply methods
  //
  // --------------------------------------------------------------------

  // set indentation
  //
  if (level_a > Integral::NONE) {
    Console::put(L"testing class-specific public methods: apply methods...\n");
    Console::increaseIndention();
  }

  // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
  // case: Algorithm = LATTICE, Implementation = BURG
  //       Input = non-zero and zero SIGNAL at channel 0 and 1
  // =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

  // local variables
  //
  Vector< CircularBuffer<AlgorithmData> > in;
  Vector<AlgorithmData> out;
  AlgorithmData data;
  
  float z = 1;
  for (long i = 4; i < 20; i++) {
    input(i) = 2 * z - z * z;
    z = 0.99 * z;
  }    

  // number of channels
  //
  long N = 2;
  
  in.setLength(N);
  out.setLength(N);
  
  for (long i = 0; i < N; i++) {
    in(i).append(data);
    in(i)(0).makeVectorFloat();
    in(i)(0).setCoefType(AlgorithmData::SIGNAL);
  }
  in(0)(0).getVectorFloat().assign(input);
  input.assign((float)0);
  in(1)(0).getVectorFloat().assign(input);
    
  // compute the reflection coefficients using the BURG algorithm via
  // a lattice with order M
  //
  long M = 4;
  reflc.set(LATTICE, BURG, M, -200);
  reflc.apply(out, in);

  VectorFloat res_refl_coef_burg(L"-0.9378492, 0.03325069, 0.03421993, 0.03525940");
  VectorFloat zeros(L"0, 0, 0, 0");
  
  if (!out(0).getVectorFloat().almostEqual(res_refl_coef_burg)) {
    out(0).getVectorFloat().debug(L"refl(0)");
    res_refl_coef_burg.debug(L"res_refl_coef");
    return Error::handle(name(), L"diagnose", ERR, __FILE__, __LINE__);    
  }

  if (!out(1).getVectorFloat().almostEqual(zeros)) {
    out(1).getVectorFloat().debug(L"refl(1)");
    zeros.debug(L"zeros_refl_coef");
    return Error::handle(name(), L"diagnose", ERR, __FILE__, __LINE__);    
  }
  
  Error::reset();
  Error::set(Error::WARNING);

  // reset indentation
  //
  if (level_a > Integral::NONE) {
    Console::decreaseIndention();
  }

  //---------------------------------------------------------------------------
  //
  // 6. print completion message
  //
  //---------------------------------------------------------------------------

  // reset indentation
  //
  if (level_a > Integral::NONE) {
    Console::decreaseIndention();
  }
  
  if (level_a > Integral::NONE) {
    SysString output(L"diagnostics passed for class ");
    output.concat(name());
    output.concat(L"\n");
    Console::put(output);
  }
  
  // exit gracefully
  //
  return true;
}