#include <cstdlib>
#include <iostream>
#include <fstream>

using namespace std;

int main()
{
    const int ARRAY_SIZE = 64;
    const int TOP_PERCENTILE = 2;
    const int MIDDLE_PERCENTILE = 1;
    const int LOWER_PERCENTILE = 0;
    int i = 0; // generic counter variable
    int size;  // size of the dataset
    int max = 0;    // max value in dataset
    double upper_threshold, lower_threshold; // max * 0.9, max * 0.1
    int dataset[ARRAY_SIZE];

    cout << "Hello again, it feels so good to be executed (I'm a program, after all).\n";
    cout << "I will read in a data file and output discretized values, depending "
         << "on the lowest 10%, middle, and highest 10% thresholds.\n\n";
    ifstream data_file("DATA");

    if (!data_file) {
        cerr << "There was a problem opening the data file." << endl;
        system("PAUSE");
        exit(1);
    }
    
    // read size of dataset, which is the first number in the series
    if ( !(data_file >> size)) {
        cerr << "The size digit was not able to be read." << endl;
        system("PAUSE");
        exit(1);
    }
    
    // push each element in the datafile into an array location
    for (i = 0; i < size; ++i) {
        data_file >> dataset[i];
    }

    // DEBUG: Am I sane? Print the values in the array, to make sure.
    // for (i = 0; i < size; ++i) {
    //     cout << dataset[i] << endl;
    // }

    // what is the maximum value in the dataset?
    for (int i = 0; i < size; ++i) {
        if (dataset[i] > max) max = dataset[i];
    }

    // DEBUG: Am I sane? Print the values in the array, to make sure.    
    // cout << "MAX: " << max << endl;

    // determine the 90% and 10% thresholds, just using max (per the rubric)
    upper_threshold = max * 0.9;
    lower_threshold = max * 0.1;

    // Compare each value in the array to my thresholds.
    // Replace original value with the discretized value
    for (i = 0; i < size; ++i) {
        if (dataset[i] > upper_threshold) {
            dataset[i] = TOP_PERCENTILE;
        } else if (dataset[i] >= lower_threshold &&
                   dataset[i] <= upper_threshold) {
            dataset[i] = MIDDLE_PERCENTILE;
        } else {
            dataset[i] = LOWER_PERCENTILE;
        }
    }

    cout << "The resulting discretization of the dataset looks like this:\n\n";
    
    // Print the output to the screen
    for (i = 0; i < size; ++i) {
        cout << dataset[i];
    }    
    cout << endl << endl;
    
    system("PAUSE");
    return 0;
}