/// <summary>
/// Assignes vector to this cluster. </summary>
/// <param name="vector"> vector to assign </param>
public virtual void assignVector(KVector vector)
{
// if vector's cluster is allready set to this, save some cpu cycles
if (vector.Cluster != this)
{
vector.Cluster = this;
vectors.Add(vector);
}
}
public KMeansClustering(DataSet dataSet, int numberOfClusters)
{
this.dataSet = dataSet;
this.numberOfClusters = numberOfClusters;
this.dataVectors = new KVector[dataSet.size()];
// iterate dataset and create dataVectors field
int i = 0;
foreach (DataSetRow row in dataSet.Rows)
{
KVector vector = new KVector(row.Input);
this.dataVectors[i] = vector;
i++;
}
}
/// <summary>
/// Calculates and returns euclidean distance of this vector from the given cluster </summary>
/// <param name="cluster"> </param>
/// <returns> euclidean distance of this vector from given cluster </returns>
public virtual double distanceFrom(KVector otherVector)
{
// get values and do this in loop
double[] otherValues = otherVector.Values;
double distance = 0;
for (int i = 0; i < values.Length; i++)
{
distance += Math.Pow(otherValues[i] - values[i], 2);
}
distance = Math.Sqrt(distance);
return(distance);
}
/// <summary>
/// Find and return the nearest cluster for the specified vector </summary>
/// <param name="vector"> </param>
/// <returns> </returns>
private Cluster getNearestCluster(KVector vector)
{
Cluster nearestCluster = null;
double minimumDistanceFromCluster = double.MaxValue;
double distanceFromCluster = 0;
foreach (Cluster cluster in clusters)
{
distanceFromCluster = vector.distanceFrom(cluster.Centroid);
if (distanceFromCluster < minimumDistanceFromCluster)
{
minimumDistanceFromCluster = distanceFromCluster;
nearestCluster = cluster;
}
}
return(nearestCluster);
}
// find initial values for centroids/clusters
// dont need to return string this is just for debuging and output in dialog
// forgy and random partitions
// http://en.wikipedia.org/wiki/K-means_clustering
public virtual void initClusters()
{
List <int> idxList = new List <int>();
for (int i = 0; i < dataSet.size(); i++)
{
idxList.Add(i);
}
Shuffle(idxList);
// log.append("Clusters initialized at:\n\n");
clusters = new Cluster[numberOfClusters];
for (int i = 0; i < numberOfClusters; i++)
{
clusters[i] = new Cluster();
int randomIdx = idxList[i];
KVector randomVector = dataVectors[randomIdx];
clusters[i].Centroid = randomVector;
//log.append(randomVector.toString()+System.lineSeparator() );
}
//log.append(System.lineSeparator());
}
public virtual void removePoint(KVector point)
{
vectors.Remove(point);
}
