/// <summary>
/// Initializes a new instance of the Binary Split algorithm
/// </summary>
///
/// <param name="k">The number of clusters to divide the input data into.</param>
/// <param name="distance">The distance function to use. Default is to
/// use the <see cref="BestCS.Math.Distance.SquareEuclidean(double[], double[])"/> distance.</param>
///
public BinarySplit(int k, Func <double[], double[], double> distance)
{
if (k <= 0)
{
throw new ArgumentOutOfRangeException("k");
}
if (distance == null)
{
throw new ArgumentNullException("distance");
}
// Create the object-oriented structure to hold
// information about the k-means' clusters.
this.clusters = new KMeansClusterCollection(k, distance);
}
/// <summary>
/// Initializes a new instance of the KMeans algorithm
/// </summary>
///
/// <param name="k">The number of clusters to divide the input data into.</param>
/// <param name="distance">The distance function to use. Default is to
/// use the <see cref="BestCS.Math.Distance.SquareEuclidean(double[], double[])"/> distance.</param>
///
public KMeans(int k, Func <double[], double[], double> distance)
{
if (k <= 0)
{
throw new ArgumentOutOfRangeException("k");
}
if (distance == null)
{
throw new ArgumentNullException("distance");
}
this.Tolerance = 1e-5;
this.ComputeInformation = true;
this.UseCentroidSeeding = true;
// Create the object-oriented structure to hold
// information about the k-means' clusters.
this.clusters = new KMeansClusterCollection(k, distance);
}
/// <summary>
/// Initializes a new instance of the <see cref="KMeansCluster"/> class.
/// </summary>
///
/// <param name="owner">The owner collection.</param>
/// <param name="index">The cluster index.</param>
///
public KMeansCluster(KMeansClusterCollection owner, int index)
{
this.owner = owner;
this.index = index;
}