third_codes/cppjieba
1
0
Fork 0
mirror of https://github.com/yanyiwu/cppjieba.git synced 2025-07-18 00:00:12 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix bugs

Browse Source
This commit is contained in:
gwdwyy 2013-08-26 01:30:35 +08:00
parent 05a4ba3a22
commit dce3755d79
2 changed files with 7 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
src/HMMSegment.cpp
View File

@ -120,24 +120,22 @@ namespace CppJieba
return false; return false;
} }
cout<<__FILE__<<__LINE__<<endl;
if(!viterbi(unico, status)) if(!viterbi(unico, status))
{ {
LogError("viterbi failed."); LogError("viterbi failed.");
return false; return false;
} }
begin = unico.begin(); begin = unico.begin();
left = begin;
res.clear(); res.clear();
for(uint i =0; i< status.size(); i++) for(uint i =0; i< status.size(); i++)
{ {
switch(status[i]) switch(status[i])
{ {
case B:
left = begin + i;
break;
case E: case E:
right = begin + i + 1; right = begin + i + 1;
res.push_back(TransCode::vecToStr(left, right)); res.push_back(TransCode::vecToStr(left, right));
left = right;
break; break;
case S: case S:
res.push_back(TransCode::vecToStr(begin + i, begin + i +1)); res.push_back(TransCode::vecToStr(begin + i, begin + i +1));
@ -187,7 +185,6 @@ namespace CppJieba
weight[0 + y * X] = _startProb[y] + _getEmitProb(_emitProbVec[y], unico[0], MIN_DOUBLE); weight[0 + y * X] = _startProb[y] + _getEmitProb(_emitProbVec[y], unico[0], MIN_DOUBLE);
path[0 + y * X] = -1; path[0 + y * X] = -1;
} }
cout<<__FILE__<<__LINE__<<endl;
//process //process
for(uint x = 1; x < X; x++) for(uint x = 1; x < X; x++)
@ -196,12 +193,11 @@ namespace CppJieba
{ {
now = x + y*X; now = x + y*X;
weight[now] = MIN_DOUBLE; weight[now] = MIN_DOUBLE;
path[now] = E; path[now] = E; // warning
for(uint preY = 0; preY < Y; preY++) for(uint preY = 0; preY < Y; preY++)
{ {
old = x - 1 + preY * X; old = x - 1 + preY * X;
tmp = weight[old] + _transProb[preY][y] + _getEmitProb(_emitProbVec[y], unico[x], MIN_DOUBLE); tmp = weight[old] + _transProb[preY][y] + _getEmitProb(_emitProbVec[y], unico[x], MIN_DOUBLE);
//cout<<__FILE__<<__LINE__<<endl;
//cout<<MIN_DOUBLE+MIN_DOUBLE+MIN_DOUBLE<<endl; //cout<<MIN_DOUBLE+MIN_DOUBLE+MIN_DOUBLE<<endl;
//cout<<weight[old]<<":"<<_transProb[preY][y]<<":"<<_getEmitProb(_emitProbVec[y], unico[x], MIN_DOUBLE)<<endl; //cout<<weight[old]<<":"<<_transProb[preY][y]<<":"<<_getEmitProb(_emitProbVec[y], unico[x], MIN_DOUBLE)<<endl;
//cout<<tmp<<endl; //cout<<tmp<<endl;
@ -211,7 +207,6 @@ namespace CppJieba
path[now] = preY; path[now] = preY;
} }
} }
//cout<<__FILE__<<__LINE__<<endl;
//cout<<x<<":"<<y<<":"<<weight[now]<<endl; //cout<<x<<":"<<y<<":"<<weight[now]<<endl;
//getchar(); //getchar();
} }
@ -228,19 +223,14 @@ namespace CppJieba
{ {
stat = S; stat = S;
} }
cout<<__FILE__<<__LINE__<<endl;
status.assign(X, 0); status.assign(X, 0);
cout<<__FILE__<<__LINE__<<endl;
for(int x = X -1 ; x >= 0; x--) for(int x = X -1 ; x >= 0; x--)
{ {
status[x] = stat; status[x] = stat;
cout<<__FILE__<<__LINE__<<endl;
cout<<stat<<endl;
stat = path[x + stat*X]; stat = path[x + stat*X];
} }
cout<<__FILE__<<__LINE__<<endl;
delete [] path; delete [] path;
delete [] weight; delete [] weight;
return true; return true;
@ -302,6 +292,7 @@ namespace CppJieba
return true; return true;
} }
/*
double HMMSegment::_getEmitProb(const EmitProbMap& mp, uint16_t key, double defVal) double HMMSegment::_getEmitProb(const EmitProbMap& mp, uint16_t key, double defVal)
{ {
EmitProbMap::const_iterator cit = mp.find(key); EmitProbMap::const_iterator cit = mp.find(key);
@ -311,6 +302,7 @@ namespace CppJieba
} }
return cit->second; return cit->second;
} }
*/
double HMMSegment::_getEmitProb(const EmitProbMap* ptMp, uint16_t key, double defVal) double HMMSegment::_getEmitProb(const EmitProbMap* ptMp, uint16_t key, double defVal)
{ {
@ -339,7 +331,7 @@ int main()
HMMSegment hmm; HMMSegment hmm;
hmm.loadModel("../dicts/hmm_model.utf8"); hmm.loadModel("../dicts/hmm_model.utf8");
vector<string> res; vector<string> res;
hmm.cut("小明硕士毕业于北邮网络研究院，然", res); hmm.cut("小明硕士毕业于北邮网络研究院。。.", res);
cout<<joinStr(res, "/")<<endl; cout<<joinStr(res, "/")<<endl;

2
src/HMMSegment.h
View File

@ -42,7 +42,7 @@ namespace CppJieba
bool _getLine(ifstream& ifile, string& line); bool _getLine(ifstream& ifile, string& line);
bool _loadEmitProb(const string& line, EmitProbMap& mp); bool _loadEmitProb(const string& line, EmitProbMap& mp);
bool _decodeOne(const string& str, uint16_t& res); bool _decodeOne(const string& str, uint16_t& res);
double _getEmitProb(const EmitProbMap& mp, uint16_t key, double defVal); //double _getEmitProb(const EmitProbMap& mp, uint16_t key, double defVal);
double _getEmitProb(const EmitProbMap* ptMp, uint16_t key, double defVal); double _getEmitProb(const EmitProbMap* ptMp, uint16_t key, double defVal);