NullClassifier.cpp

#include "NullClassifier.hpp"
#include <iostream>
#include <algorithm>
#include <set>
using namespace std;

NullClassifier::NullClassifier() 
    : trueTotal(0)
{}

void NullClassifier::train(const vector<string> &words, bool label) {
    for (const auto &w : words) {
        trueCounts[w]++;
        trueTotal++;    
    }
}

void NullClassifier::trainingFinished() {
    // normalize counts based on # of words on true examples
    //  vs negative examples


    // remove this if you want to see debug info
    return;

    // print some debug info
    cout << "VocabClassifier debug info" << endl;
    cout << "==========================" << endl;
    cout << "trueTotal: " << trueTotal << endl;
    cout << endl;

    // most true and most false words
    const int topCount = 25;

    // create a set of all words. sets can't have duplicates
    set<string> vocabulary;
    for (auto it : trueCounts) {
        vocabulary.insert(it.first);
    }

    // compute score for each word
    map<string, double> differences;
    for (auto word : vocabulary) {
        differences[word] = trueCounts[word];
    }    

    // Find the words with the highest value. You don't need to understand
    //   how this works.
    vector<pair<string, double>> mostTrue;
    copy(differences.begin(), differences.end(),
       back_inserter<vector<pair<string, double> > >(mostTrue));
    sort(mostTrue.begin(), mostTrue.end(), 
        [=](auto& a, auto& b) {
            return a.second > b.second;
        });

    cout << "Most true words:" << endl;
    for (int i = 0; i < topCount; i++) {
        cout << mostTrue.at(i).second << ": " << mostTrue.at(i).first << endl;
    }

}

bool NullClassifier::classify(const vector<string> &text) {
    return true;
}