public ArrayList<TrainingExample> readInput() throws FileNotFoundException {
for (int k = 1; k < 6; k++) {
Scanner input = new Scanner(new File("file1.txt"));
int fold = input.nextInt();
int no_of_ex = input.nextInt();
int perm_no = input.nextInt();
String[] permutations;
ArrayList<TrainingExample> trainingData = new ArrayList<>();
trainingData = Utility.readData();
ArrayList<TrainingExample> reArrangedTrainingData;
// Used to find the errors for a given permutation, this array
// contains errors of all the permutations according to index
double errorEstimates[] = new double[perm_no];
// used to skip to next line to read the permutations from the next
// line
input.nextLine();
double num_perm = perm_no;
// Taking the permutation from the dataSet
while (input.hasNextLine()) {
permutations = input.nextLine().split(" ");
// Rearranging the permutation in the serial order- so that
// working becomes easy
reArrangedTrainingData = Utility.reArrange(trainingData, permutations);
// Calling the KFold Cross Validation on the permutation for a
// given fold, and finding the error estimate
errorEstimates[perm_no - 1] = kfoldCV(reArrangedTrainingData, fold, k);
perm_no--;
}
// Computing the error estimate
double errorEstimate = 0;
double temp = 0;
// Summation of the errors across various permutation
for (int i = 0; i < num_perm; i++) {
temp = temp + errorEstimates[i];
}
// This gives the 'accurate error' estimate
errorEstimate = temp / num_perm;
temp = 0;
// Calculates the summation of the values for computing variance
for (int i = 0; i < num_perm; i++) {
temp = temp + Math.pow((errorEstimate - errorEstimates[i]), 2);
}
// This gives the variance of all the permutations for a given 'k'
// value
double variance = 0;
variance = temp / (num_perm - 1);
// This is the standard deviation for a given 'k' value
double standardDeviation = Math.sqrt(variance);
// Prints the error and standard deviation
System.out.println("Error:" + errorEstimate + "Standard Deviation:" + standardDeviation);
// Calls gridLabelling.. wherein the function prints the label grid
labelGrids(k);
}
return trainingData;
}
