public void KMeansClusteringWorksOnTwoDimensions()
{
var dataSet = new IdentifiableDataPointCollection();
var p1 = new IdentifiableDataPoint(0, 2);
var p2 = new IdentifiableDataPoint(1, 2);
var p3 = new IdentifiableDataPoint(2, 2);
var p4 = new IdentifiableDataPoint(3, 2);
p1.AddAttribute("Gender", 1);
p1.AddAttribute("Income", 1);
p2.AddAttribute("Gender", 0);
p2.AddAttribute("Income", 0.1429);
p3.AddAttribute("Gender", 1);
p3.AddAttribute("Income", 0.2858);
p4.AddAttribute("Gender", 1);
p4.AddAttribute("Income", 1);
dataSet.AddItem(p1);
dataSet.AddItem(p2);
dataSet.AddItem(p3);
dataSet.AddItem(p4);
var kmeans2 = new KMeans(dataSet, new [] { 0, 1, 2 }, new EuclideanMetric());
var result2 = kmeans2.Calculate();
Assert.AreEqual(2, result2.Clusters[0].Members.Count);
Assert.AreEqual(1, result2.Clusters[1].Members.Count);
Assert.AreEqual(1, result2.Clusters[2].Members.Count);
}
public void KMeansShouldCalculateCorrectCentroidsInOneDimension()
{
var dataSet = new IdentifiableDataPointCollection();
var p1 = new IdentifiableDataPoint(0, 1);
var p2 = new IdentifiableDataPoint(1, 1);
var p3 = new IdentifiableDataPoint(2, 1);
var p4 = new IdentifiableDataPoint(3, 1);
p1.AddAttribute("Gender", 1);
p2.AddAttribute("Gender", 0);
p3.AddAttribute("Gender", 1);
p4.AddAttribute("Gender", 1);
dataSet.AddItem(p1);
dataSet.AddItem(p2);
dataSet.AddItem(p3);
dataSet.AddItem(p4);
var kmeans2 = new KMeans(dataSet, new [] { 0, 1 }, new EuclideanMetric());
var result2 = kmeans2.Calculate();
double[] centroid1 = { 1 };
double[] centroid2 = { 0 };
Assert.AreEqual(centroid1, result2.Clusters[0].Centroid.Coordinates);
Assert.AreEqual(centroid2, result2.Clusters[1].Centroid.Coordinates);
}
public void KMeansShouldClusterDataPointsInOneDimensionCorrectly()
{
var dataSet = new IdentifiableDataPointCollection();
var p1 = new IdentifiableDataPoint(0, 1);
var p2 = new IdentifiableDataPoint(1, 1);
var p3 = new IdentifiableDataPoint(2, 1);
var p4 = new IdentifiableDataPoint(3, 1);
p1.AddAttribute("Gender", 1);
p2.AddAttribute("Gender", 0);
p3.AddAttribute("Gender", 1);
p4.AddAttribute("Gender", 1);
dataSet.AddItem(p1);
dataSet.AddItem(p2);
dataSet.AddItem(p3);
dataSet.AddItem(p4);
var kmeans2 = new KMeans(dataSet, new [] { 0, 1 }, new EuclideanMetric());
var result2 = kmeans2.Calculate();
var cluster1Members = result2.Clusters[0].Members.Select(e => e.Member).ToArray();
var cluster2Members = result2.Clusters[1].Members.Select(e => e.Member).ToArray();
Assert.AreEqual(3, cluster1Members.Length);
Assert.Contains(p1, cluster1Members);
Assert.Contains(p3, cluster1Members);
Assert.Contains(p4, cluster1Members);
Assert.AreEqual(1, cluster2Members.Length);
Assert.Contains(p2, cluster2Members);
}
static void RunProgram()
{
//Config
string codeName = "Out-Cikade";
bool printKMeans = true;
bool printKMeansMembers = false;
bool distanceMatrix = true;
bool multiDimensionalScaling = true;
bool localOutlierFactor = true;
IConfiguration configuration =
(ConfigurationSectionHandler)ConfigurationManager.GetSection("csvDataImport");
var reader = new StreamReader("UserData.csv");
var importer = new CsvDataImporter(reader, configuration);
var dataCollection = importer.Run();
//Algo//
//KMeans
var k = 5;
var distanceMetric = new EuclideanMetric();
var kmeans = new KMeans(dataCollection, k, distanceMetric);
var start = DateTime.Now;
var result = kmeans.Calculate();
//Other alogs
//File printing//
var fileName = string.Format("{0}-{1:MM-dd_HH_mm_ss}.txt", codeName, DateTime.UtcNow);
FileStream fs = new FileStream(fileName, FileMode.CreateNew);
StreamWriter writer = new StreamWriter(fs);
writer.Write("Testrun under CodeName {0}. \r\n\r\n", codeName, DateTime.UtcNow);
if (printKMeans)
{
PrintKMeans(writer, result, printKMeansMembers);
}
if (distanceMatrix)
{
DistanceMatrix(writer, dataCollection, distanceMetric);
}
if (multiDimensionalScaling)
{
MultiDimensionalScaling(writer, dataCollection, distanceMetric);
}
if (localOutlierFactor)
{
LocalOutlierFactorD(4, dataCollection, distanceMetric);
}
writer.Close();
fs.Close();
}
public void KMeansCentroidsArePlacedRightInFiveDimensions()
{
var dataSet = new IdentifiableDataPointCollection();
var p1 = new IdentifiableDataPoint(0, 5);
var p2 = new IdentifiableDataPoint(1, 5);
var p3 = new IdentifiableDataPoint(2, 5);
var p4 = new IdentifiableDataPoint(3, 5);
p1.AddAttribute("Gender", 1);
p1.AddAttribute("Income", 1);
p1.AddAttribute("Age", 0.16);
p1.AddAttribute("Purchase", 0.5);
p1.AddAttribute("Control", 1);
p2.AddAttribute("Gender", 0);
p2.AddAttribute("Income", 0.1429);
p2.AddAttribute("Age", 0.16);
p2.AddAttribute("Purchase", 1);
p2.AddAttribute("Control", 0);
p3.AddAttribute("Gender", 1);
p3.AddAttribute("Income", 0.2858);
p3.AddAttribute("Age", 0.16);
p3.AddAttribute("Purchase", 1);
p3.AddAttribute("Control", 1);
p4.AddAttribute("Gender", 1);
p4.AddAttribute("Income", 1);
p4.AddAttribute("Age", 0.16);
p4.AddAttribute("Purchase", 1);
p4.AddAttribute("Control", 0.5);
dataSet.AddItem(p1);
dataSet.AddItem(p2);
dataSet.AddItem(p3);
dataSet.AddItem(p4);
var kmeans2 = new KMeans(dataSet, new [] { 0, 1, 2 }, new EuclideanMetric());
var result2 = kmeans2.Calculate();
double[] _centroid1 = { 1, 0.2858, 0.16, 1, 1 };
double[] _centroid2 = { 1, 1, 0.16, 0.75, 0.75 };
double[] _centroid3 = { 0, 0.1429, 0.16, 1, 0 };
Assert.AreEqual(_centroid2, result2.Clusters[0].Centroid.Coordinates);
Assert.AreEqual(_centroid3, result2.Clusters[1].Centroid.Coordinates);
Assert.AreEqual(_centroid1, result2.Clusters[2].Centroid.Coordinates);
}
private ClusteringResult ClusterCalculate()
{
int inputClusterNumber = Convert.ToInt32(numberOfClusters.Value);
string inputClusterAlgo = cbClusteringAlgo.Text;
currentClusterNumber = inputClusterNumber;
/* Choosing clustering algorithm without name filter */
switch (inputClusterAlgo)
{
case "K-means":
currentClusterAlgorithm = inputClusterAlgo;
kmeans = new KMeans(this.dataSet, currentClusterNumber, distanceMetric);
return(kmeans.Calculate());
default:
currentClusterAlgorithm = inputClusterAlgo;
kmeans = new KMeans(this.dataSet, currentClusterNumber, distanceMetric);
return(kmeans.Calculate());
}
}
private void bw_DoWork(object sender, DoWorkEventArgs e)
{
bool changed;
bool drawLine = checkBox1.Checked;
uint currentIteration = 0;
BackgroundWorker worker = sender as BackgroundWorker;
KMeans kmeans = new KMeans(_maxmin._points, _maxmin._clusters);
do
{
currentIteration++;
label5.Invoke(new Action(delegate()
{
label5.Text = currentIteration.ToString();
label5.Update();
}));
lock (_sync)
{
_graphicsBuffer.Graphics.Clear(BackColor);
kmeans.DrawClusters(_graphicsBuffer.Graphics, drawLine);
_graphicsBuffer.Render();
}
if (_bw.CancellationPending == true)
{
button2.Invoke(new Action(delegate()
{
button2.Text = "Стоп";
button2.Update();
}));
break;
}
changed = kmeans.Calculate();
} while (changed);
}
