#include 
     
      #include 
      
       #include 
       
        #include 
        
         #include 
         
          #include 
          
           #include 
           
            #include 
            
             #include 
             
              #include 
              
               #include 
               
                #include 
                
                 template 
                 
                  void ReadMatFromFile(std::string &filename, std::vector
                  
                   
                     > &lines_feat) { std::ifstream vm_info(filename.c_str()); std::string lines; DataType var; std::vector
                    
                      row; lines_feat.clear(); while(!vm_info.eof()) { getline(vm_info, lines); if(lines.empty()) break; std::stringstream stringin(lines); row.clear(); while(stringin >> var) { row.push_back(var); } lines_feat.push_back(row); }}void ReadStringFromFile(std::string &filename, std::vector
                     
                       &in_string) { std::ifstream vm_info(filename.c_str()); std::string lines, var; while(!vm_info.eof()) { getline(vm_info, lines); if(lines.empty()) break; std::stringstream stringin(lines); while(stringin >> var) { in_string.push_back(var); } }}std::string lowerCase(const std::string& s) { std::string lower(s); for(size_t i=0;i
                      
                       =65 && ch<=90)) { ch=ch+32; flag=true; } else if((ch>=97 && ch<=122) || (ch>=48 && ch<=57)) { flag=true; } else { ; } if(flag) { letter.push_back(ch); } } letter.push_back('\0'); return letter;}template 
                       
                        int MatMultiply(const std::vector
                        
                         
                           > &Mata, const std::vector
                          
                           
                             > &Matb, std::vector
                            
                             
                               > &MatOut) { if(Mata.at(0).size() != Matb.size()) { std::cout<<"not match!\n"; return -1; } for(size_t i=0; i
                              
                                col; col.clear(); for(size_t k=0; k
                               
                                void outer_product(const std::vector
                                
                                  &inst1, const std::vector
                                 
                                   &inst2, std::vector
                                  
                                   
                                     > &out) { std::vector
                                    
                                      temp_row(inst2.size()); for(typename::std::vector
                                     
                                      ::const_iterator it=inst1.begin();it!=inst1.end();++it) { transform(inst2.begin(), inst2.end(), temp_row.begin(), bind2nd(std::multiplies
                                      
                                       (), *it)); out.push_back(temp_row); }}void ReadDataFromFile(std::string &filename, std::vector
                                       
                                         &lines_feat) { std::ifstream vm_info(filename.c_str()); std::string lines; lines_feat.clear(); while(!vm_info.eof()) { getline(vm_info, lines); if(lines.empty()) break; lines_feat.push_back(lines); }}std::vector
                                        
                                          split(const std::string& s, char delimiter) { std::vector
                                         
                                           tokens; std::string token; std::istringstream tokenStream(s); while(std::getline(tokenStream, token, delimiter)) { tokens.push_back(token); } return tokens;}int stringtoint(const std::string& s) { std::istringstream iss(s); int num; return iss>>num?num:0;}void printip(const std::string& s) { std::vector
                                          
                                            temp, ip_segment; temp=split(s, '-'); ip_segment=split(temp.front(), '.'); std::string ip_start=ip_segment.back(), ip_end=temp.back(); int start, end; start=stringtoint(ip_start); end=stringtoint(ip_end); for(size_t i=start;i<=end;++i) { std::cout<
                                           
                                            <<"."<
                                            
                                             <<"."<
                                             
                                              <<"."<
                                              <<"\n"; }}template 
                                               
                                                void Display2DVector(std::vector
                                                
                                                 
                                                   > &vv) { for(size_t i=0;i
                                                  
                                                   ::const_iterator it=vv.at(i).begin();it!=vv.at(i).end();++it) { std::cout<<*it<<" "; } std::cout<<"\n"; } std::cout<<"--------the total of the 2DVector is "<
                                                   
                                                    <
                                                    
                                                      v_string; std::string words; std::set
                                                     
                                                       s_string; ReadStringFromFile(filename, v_string); for(std::vector
                                                      
                                                       ::const_iterator it=v_string.begin(); it!=v_string.end(); ++it) { std::cout<<*it<<" "; words=letters(*it); s_string.insert(words); } std::cout<
                                                       
                                                        ::const_iterator it=s_string.begin(); it!=s_string.end(); ++it) { std::cout<<*it<<" "; } std::cout<
                                                       
                                                      
                                                     
                                                    
                                                   
                                                  
                                                 
                                                
                                               
                                             
                                            
                                           
                                          
                                         
                                        
                                       
                                      
                                     
                                    
                                   
                                  
                                 
                                
                               
                              
                             
                            
                           
                          
                         
                        
                       
                      
                     
                    
                   
                  
                 
                
               
              
             
            
           
          
         
        
       
      
     

The bag of words model ignores grammar and order of words.

运行结果如下，第一行为原始数据，第二行为提取后的数据：