/// <summary>
/// Returns the number of data points that changed clusters.
/// </summary>
private int UpdateNearest()
{
// create balanced tree of cluster centers
var tree = KdTree.CreateBalanced(_clusterCenters, _clusterIndices);
// for each data point, find the nearest cluster center
int count = 0;
for (int i = 0; i < _points.Length; i++)
{
var nearest = tree.NearestL2(_points[i]);
// update if new nearest is different than previous
if (nearest != _nearestClusters[i])
{
_nearestClusters[i] = nearest;
count++;
}
}
return(count);
}
/// <summary>
/// Inserts point value pairs in a way that produces a balanced tree.
/// </summary>
public static KdTree <int> CreateBalanced(double[][] points)
{
return(KdTree <int> .CreateBalanced(points, Enumerable.Range(0, points.Length).ToArray()));
}
/// <summary>
/// Inserts point value pairs in a way that produces a balanced tree.
/// </summary>
public static KdTree <int> CreateBalanced(IEnumerable <double[]> points)
{
return(KdTree <int> .CreateBalanced(points, Sequences.CountFrom(0)));
}
/// <summary>
/// Inserts point value pairs in a way that produces a balanced tree.
/// </summary>
public static KdTree <T> CreateBalanced <T>(double[][] points, T[] values)
{
return(KdTree <T> .CreateBalanced(points, values));
}
/// <summary>
/// Inserts point value pairs in a way that produces a balanced tree.
/// </summary>
public static KdTree <T> CreateBalanced <T>(IEnumerable <double[]> points, IEnumerable <T> values)
{
return(KdTree <T> .CreateBalanced(points, values));
}
