private static void clusterizeAndSave(ClusteringMethod <Triplet <Word> > method, Metrica <Triplet <Word> > metrica, int count)
{
Console.WriteLine("Clustering method " + method.ToString());
Console.WriteLine("Metrica " + metrica.ToString());
var path = "..\\" + method.ToString() + "\\" + metrica.ToString() + "\\";
Directory.CreateDirectory(path);
var result = method.clusterize(data, metrica, count);
var format = new XmlSerializer(typeof(List <Triplet <Word> >));
var statFormat = new BinaryFormatter();
resultItemsCount = 0;
FileStream stream;
for (int i = 0; i < result.Count; i++)
{
resultItemsCount += result[i].Count;
var fileName = path + "clustersResult" + i + ".xml";
stream = File.Create(fileName);
format.Serialize(stream, result[i]);
stream.Close();
var fs = File.OpenWrite(path + "clusterStat" + i + ".txt");
var stat = getStatForCluster(result[i], fs);
fs.Close();
}
var dataStat = getStatForCluster(data, File.OpenWrite(path + "dataStat.txt"));
}
private void ClusteringMethodChanged(object sender, RoutedEventArgs e)
{
string text = (sender as ComboBox).SelectedValue as string;
ClusteringMethod m = (ClusteringMethod)Enum.Parse(typeof(ClusteringMethod), text);
(this.DataContext as ViewModel).MyBatchDesignView.BatchDesign.Method = m;
(this.DataContext as ViewModel).MyBatchDesignView.DisplayAllClusters();
}
/// <summary>
/// Cluster the data set into groups of similar instances.
/// </summary>
/// <param name="matrix">Input matrix</param>
/// <param name="clusteringMethod">Algorithm to use for clustering</param>
/// <returns>Results of the cluster analysis</returns>
public static IClusteringResults Cluster(this InsightMatrix matrix, ClusteringMethod clusteringMethod)
{
switch(clusteringMethod)
{
case ClusteringMethod.KMeans:
return new KMeansClustering().Cluster(matrix);
default:
return new MetaKMeansClustering().Cluster(matrix);
}
}
/// <summary>
/// Cluster the data set into groups of similar instances.
/// </summary>
/// <param name="matrix">Input matrix</param>
/// <param name="clusteringMethod">Algorithm to use for clustering</param>
/// <param name="distanceMethod">Distance measure used to compare instances</param>
/// <param name="clusters">Number of desired clusters</param>
/// <returns>Results of the cluster analysis</returns>
public static IClusteringResults Cluster(this InsightMatrix matrix, ClusteringMethod clusteringMethod,
DistanceMethod comparisonMethod, int clusters)
{
switch (clusteringMethod)
{
case ClusteringMethod.KMeans:
return new KMeansClustering().Cluster(matrix, comparisonMethod, clusters);
default:
return new MetaKMeansClustering().Cluster(matrix, comparisonMethod, clusters);
}
}
public bool ExportHierarchicalGraph(object ownerViewModel, HierarchicalGraphType graphType, ClusteringMethod clusteringMethod, SimilarityMetric similarityMetric)
{
FileDialogResult result = _dialogService.ShowSaveFileDialog("Export Hierarchical Graph", ownerViewModel, new FileType("PNG image", ".png"));
if (result.IsValid)
{
IBidirectionalGraph <HierarchicalGraphVertex, HierarchicalGraphEdge> graph = _graphService.GenerateHierarchicalGraph(graphType, clusteringMethod, similarityMetric);
Action <double> scaleUpdate = null;
FrameworkElement graphLayout = null;
switch (graphType)
{
case HierarchicalGraphType.Dendrogram:
graphLayout = new DendrogramLayout {
Graph = graph, Background = Brushes.White
};
break;
case HierarchicalGraphType.Tree:
var hgl = new HierarchicalGraphLayout
{
IsAnimationEnabled = false,
CreationTransition = null,
DestructionTransition = null,
LayoutAlgorithmType = "RadialTree",
LayoutParameters = new RadialTreeLayoutParameters {
BranchLengthScaling = BranchLengthScaling.MinimizeLabelOverlapMinimum
},
Graph = graph,
Background = Brushes.White,
ScaleLabelsToZoom = 1.0
};
hgl.Resources[typeof(VertexControl)] = System.Windows.Application.Current.Resources["HierarchicalVertexControlStyle"];
hgl.Resources[typeof(EdgeControl)] = System.Windows.Application.Current.Resources["HierarchicalEdgeControlStyle"];
graphLayout = hgl;
scaleUpdate = scale => hgl.ScaleLabelsToZoom = scale;
break;
}
Debug.Assert(graphLayout != null);
SaveElement(graphLayout, result.FileName, scaleUpdate);
return(true);
}
return(false);
}
/// <summary>
/// Cluster the data set into groups of similar instances.
/// </summary>
/// <param name="matrix">Input matrix</param>
/// <param name="clusteringMethod">Algorithm to use for clustering</param>
/// <param name="distanceMethod">Distance measure used to compare instances</param>
/// <param name="clusters">Number of desired clusters</param>
/// <param name="iterations">Number of times to run the algorithm</param>
/// <returns>Results of the cluster analysis</returns>
public static IClusteringResults Cluster(this InsightMatrix matrix, ClusteringMethod clusteringMethod,
DistanceMethod comparisonMethod, int clusters, int iterations)
{
return new MetaKMeansClustering().Cluster(matrix, comparisonMethod, clusters, iterations);
}
public IBidirectionalGraph<HierarchicalGraphVertex, HierarchicalGraphEdge> GenerateHierarchicalGraph(HierarchicalGraphType graphType,
ClusteringMethod clusteringMethod, SimilarityMetric similarityMetric)
{
switch (clusteringMethod)
{
case ClusteringMethod.Upgma:
Func<Variety, Variety, double> upgmaGetDistance = null;
switch (similarityMetric)
{
case SimilarityMetric.Lexical:
upgmaGetDistance = (v1, v2) => 1.0 - v1.VarietyPairs[v2].LexicalSimilarityScore;
break;
case SimilarityMetric.Phonetic:
upgmaGetDistance = (v1, v2) => 1.0 - v1.VarietyPairs[v2].PhoneticSimilarityScore;
break;
}
var upgma = new UpgmaClusterer<Variety>(upgmaGetDistance);
IBidirectionalGraph<Cluster<Variety>, ClusterEdge<Variety>> upgmaTree = upgma.GenerateClusters(_projectService.Project.Varieties);
return BuildHierarchicalGraph(upgmaTree);
case ClusteringMethod.NeighborJoining:
Func<Variety, Variety, double> njGetDistance = null;
switch (similarityMetric)
{
case SimilarityMetric.Lexical:
njGetDistance = (v1, v2) => 1.0 - v1.VarietyPairs[v2].LexicalSimilarityScore;
break;
case SimilarityMetric.Phonetic:
njGetDistance = (v1, v2) => 1.0 - v1.VarietyPairs[v2].PhoneticSimilarityScore;
break;
}
var nj = new NeighborJoiningClusterer<Variety>(njGetDistance);
IUndirectedGraph<Cluster<Variety>, ClusterEdge<Variety>> njTree = nj.GenerateClusters(_projectService.Project.Varieties);
switch (graphType)
{
case HierarchicalGraphType.Dendrogram:
IBidirectionalGraph<Cluster<Variety>, ClusterEdge<Variety>> rootedTree = njTree.ToRootedTree();
return BuildHierarchicalGraph(rootedTree);
case HierarchicalGraphType.Tree:
return BuildHierarchicalGraph(njTree);
}
break;
}
return null;
}
public IBidirectionalGraph <HierarchicalGraphVertex, HierarchicalGraphEdge> GenerateHierarchicalGraph(HierarchicalGraphType graphType,
ClusteringMethod clusteringMethod, SimilarityMetric similarityMetric)
{
switch (clusteringMethod)
{
case ClusteringMethod.Upgma:
Func <Variety, Variety, double> upgmaGetDistance = null;
switch (similarityMetric)
{
case SimilarityMetric.Lexical:
upgmaGetDistance = (v1, v2) => 1.0 - v1.VarietyPairs[v2].LexicalSimilarityScore;
break;
case SimilarityMetric.Phonetic:
upgmaGetDistance = (v1, v2) => 1.0 - v1.VarietyPairs[v2].PhoneticSimilarityScore;
break;
}
var upgma = new UpgmaClusterer <Variety>(upgmaGetDistance);
IBidirectionalGraph <Cluster <Variety>, ClusterEdge <Variety> > upgmaTree = upgma.GenerateClusters(_projectService.Project.Varieties);
return(BuildHierarchicalGraph(upgmaTree));
case ClusteringMethod.NeighborJoining:
Func <Variety, Variety, double> njGetDistance = null;
switch (similarityMetric)
{
case SimilarityMetric.Lexical:
njGetDistance = (v1, v2) => 1.0 - v1.VarietyPairs[v2].LexicalSimilarityScore;
break;
case SimilarityMetric.Phonetic:
njGetDistance = (v1, v2) => 1.0 - v1.VarietyPairs[v2].PhoneticSimilarityScore;
break;
}
var nj = new NeighborJoiningClusterer <Variety>(njGetDistance);
IUndirectedGraph <Cluster <Variety>, ClusterEdge <Variety> > njTree = nj.GenerateClusters(_projectService.Project.Varieties);
switch (graphType)
{
case HierarchicalGraphType.Dendrogram:
IBidirectionalGraph <Cluster <Variety>, ClusterEdge <Variety> > rootedTree = njTree.ToRootedTree();
return(BuildHierarchicalGraph(rootedTree));
case HierarchicalGraphType.Tree:
return(BuildHierarchicalGraph(njTree));
}
break;
}
return(null);
}