protected CentroidsKMeansPPKP FillRandomCluster(int dimensions)
{
this.dimensions = dimensions;
CentroidsKMeansPPKP cl = new CentroidsKMeansPPKP(dimensions);
Random random = new Random();
//tutaj chodzi nie o randomowych liczbach w TFIDF a o randomowych dokumentach w kolekcji.
for (var i = 0; i < dimensions; i++)
{
cl.TFIDF[i] = (float)random.Next(0, Int32.MaxValue) / (float)Int32.MaxValue;
}
return cl;
}
protected CentroidsKMeansPPKP FindNearestClusterCenter(DocumentVector doc)
{
var minDistance = (double)dimensions;
CentroidsKMeansPPKP bestClusterCenter = clusters.First();
foreach (var cluster in clusters)
{
var distance = cluster.ComputeTFIDFDistance(doc);
if (distance < minDistance)
{
bestClusterCenter = cluster;
minDistance = distance;
}
}
return bestClusterCenter;
}
//changes provided 29.10.2017
protected void Iteration(int current, int max)
{
documentMoved = false;
CentroidsKMeansPPKP cluster = null;
foreach (var doc in DocCollection)
{
cluster = FindNearestClusterCenter(doc);
cluster.AssignedDocuments.Add(doc);
}
if (current == max - 1)
foreach (var clusterr in clusters)
{
clusterr.Update(true);
}
cluster.AssignedDocuments.Clear();
}