public static double CalculateEuclideanDistance(AdultTrain a, AdultTrain b)
{
List <double> dist = new List <double>();
//here I am adding the squared difference between the instance variables in a (the test instance)
//and b (the training instance) to a list
dist.Add(Square(a.age, b.age));
dist.Add(Square(1, Match(a.workclass, b.workclass)));
dist.Add(Square(a.fnlwgt, b.fnlwgt));
dist.Add(Square(1, Match(a.education, b.education)));
dist.Add(Square(a.educationNum, b.educationNum));
dist.Add(Square(1, Match(a.maritalStatus, b.maritalStatus)));
dist.Add(Square(1, Match(a.occupation, b.occupation)));
dist.Add(Square(1, Match(a.relationship, b.relationship)));
dist.Add(Square(1, Match(a.race, b.race)));
dist.Add(Square(1, Match(a.sex, b.sex)));
dist.Add(Square(a.capitalGain, b.capitalGain));
dist.Add(Square(a.capitalLoss, b.capitalLoss));
dist.Add(Square(a.hoursPerWeek, b.hoursPerWeek));
dist.Add(Square(1, Match(a.nativeCountry, b.nativeCountry)));
double sum = 0;
//here I am summing all the squared distances together
foreach (var dis in dist)
{
sum += dis;
}
//here I am calculating the square root of the sum and returning it which is the
//euclidean distance
return(Math.Sqrt(sum));
}
public static double CalculateEuclideanDistance(AdultTest a, AdultTrain b)
{
List <double> dist = new List <double>();
dist.Add(Square(a.age, b.age));
dist.Add(Square(1, Match(a.workclass, b.workclass)));
dist.Add(Square(a.fnlwgt, b.fnlwgt));
dist.Add(Square(1, Match(a.education, b.education)));
dist.Add(Square(a.educationNum, b.educationNum));
dist.Add(Square(1, Match(a.maritalStatus, b.maritalStatus)));
dist.Add(Square(1, Match(a.occupation, b.occupation)));
dist.Add(Square(1, Match(a.relationship, b.relationship)));
dist.Add(Square(1, Match(a.race, b.race)));
dist.Add(Square(1, Match(a.sex, b.sex)));
dist.Add(Square(a.capitalGain, b.capitalGain));
dist.Add(Square(a.capitalLoss, b.capitalLoss));
dist.Add(Square(a.hoursPerWeek, b.hoursPerWeek));
dist.Add(Square(1, Match(a.nativeCountry, b.nativeCountry)));
double sum = 0;
foreach (var dis in dist)
{
sum += dis;
}
return(Math.Sqrt(sum));
}
public static void CountError(AdultTrain input, TupleList <double, int> distLabels, int inputLabel, int split, int k)
{
//in this class I loop through the sorted distances and add all distances labelled <=50K and >50K
//to seperate lists, then when i equals a desired tested K, I classify the test instance as the class list
//with more distances, I then increment the number of errors in the current fold for the K if the test instance was incorrectly classified
int i = 1;
List <double> lThan50 = new List <double>();
List <double> gThan50 = new List <double>();
foreach (var dl in distLabels)
{
if (dl.Item2 == 0)
{
lThan50.Add(dl.Item1);
}
else
{
gThan50.Add(dl.Item1);
}
if (i % 2 != 0 && i != 0)
{
int label = 1;
if (lThan50.Count > gThan50.Count)
{
label = 0;
}
if (inputLabel != label)
{
kAccs[i][split - 1]++;
}
}
if (i == 39)
{
break;
}
i++;
}
}
