fork(1) download
copy 
  1. #include<iostream>
  2. #include<fstream>
  3. #include<vector>
  4. #include<map>
  5. #include<cmath>
  6. #include<algorithm>
  7. using namespace std;
  8.  
  9. void tf_idf_compute(map<string,int> &frequency,vector<vector<string> > &documents,vector<string> &terms)//function for creating a dictionery
  10. {
  11.      fstream file("regs.txt");//opens the file named regs.
  12.        if(!file)// reading file is not found
  13.        {
  14.        cout<<"file not found"<<endl;
  15.        }
  16.        else
  17.        {
  18.          while(!file.eof())//reading file is not found doesnot mark end of fuction
  19.          {
  20.             string hb;//variable of type string for holding a term
  21.             vector<string> words;//container for storing terms before storing the terms into a memory
  22.             while(file>>hb && hb!="#")//condition which direct raeding of documents with specified delimiter as the sign of an end of a document
  23.             {
  24.                    words.push_back(hb);//put terms into a temporary holding vector
  25.                    terms.push_back(hb);//keep terms in memory
  26.                     frequency[hb]++;
  27.                    sort(terms.begin(),terms.end());//sorting the terms
  28.                    terms.erase(unique(terms.begin(),terms.end()),terms.end());//remove term repeatation
  29.                    }
  30.                    if(!words.empty())//if the vector is not empty
  31.            {
  32.            documents.push_back(words);//push the words in temporary vector into the vector of vectors in order to be stored in a memory
  33.            }
  34.  
  35.         }
  36.           int a;
  37.           vector<int> ting;
  38.           for(int j=0;j<terms.size();j++)
  39.           {
  40.                  a=0;
  41.  
  42.          for(int i=0;i<documents.size();i++)
  43.          {
  44.                  //finds if a term occurs or doesn't occcur in the document
  45.           vector<string>::const_iterator p=find(documents[i].begin(),documents[i].end(),terms[j]);
  46.           if(p!=documents[i].end())
  47.           {
  48.             a=a+1;
  49.           }
  50.          }
  51.          ting.push_back(a);
  52.         //cout<<cnt<<endl;
  53.          }
  54.  
  55.  
  56.         //operation for calculating tf_idf
  57.         map<string,int>::iterator iter;
  58.         float tf_idf;
  59.         float cosine;
  60.         for(iter=frequency.begin();iter!=frequency.end();iter++)
  61.         {
  62.         tf_idf=(1+log10(iter->second))*log10(documents.size()/a);//formular to calculate tf_idf & cosine similarities
  63.         cosine=(tf_idf*iter->second)/(abs(tf_idf)*abs(iter->second));
  64.         cout<<iter->first<<"                  "<<iter->second<<"                       "<<tf_idf<<"              "<<cosine<<endl;
  65.         }
  66.  
  67.      cin.get();//holds the screen
  68. }
  69. }
  70. int main()
  71. {
  72. cout<<"Terms\t\t""Term frequency\t\t\t""TF_IDF\t\t""CosineSimilarity";
  73.     cout<<"\n";
  74.     vector<vector<string> > dc;// dictionary
  75.     vector<string> tokens;//container for documents which represent refered vector  in prototype
  76.     map<string,int>S;//A vector "s"that stores int type values.
  77.  
  78.     tf_idf_compute(S,dc,tokens);
  79. }
  80.  
  81.  
  82.  
Success #stdin #stdout 0s 2832KB
comments (?)
stdin
copy 
Standard input is empty
stdout
copy 
Terms		Term frequency			TF_IDF		CosineSimilarity
file not found
https://ideone.com/zgLJt8
language:
C++ 4.3.2 (gcc-4.3.2)
created:
9 years ago
visibility:
public

Share or Embed source code

Discover > Sphere Engine API

The brand new service which powers Ideone!

Discover > IDE Widget

Widget for compiling and running the source code in a web browser!