static void PrintKMeans(StreamWriter writer, ClusteringResult result, bool printMembers)
{
writer.WriteLine("KMeans clusters");
for (int i = 0; i < result.Clusters.Count; i++)
{
var cluster = result.Clusters[i];
var sumOfId = cluster.Members.Sum(e => e.Member.Id);
var firstId = "None";
var lastId = "None";
if (cluster.Members.Count > 0)
{
firstId = cluster.Members[0].Member.Id.ToString();
lastId = cluster.Members.Last().Member.Id.ToString();
}
writer.WriteLine("Cluster {0}: member(s) {4,4} || #{1,5}, first {2,4}, last {3,4}", i + 1, sumOfId, firstId, lastId, cluster.Members.Count);
}
writer.Write("\r\n\r\n");
if (printMembers)
{
for (int valCluster = 0; valCluster < result.Clusters.Count; valCluster++)
{
writer.WriteLine("Cluster {0} members:", valCluster);
for (int valMember = 0; valMember < result.Clusters[valCluster].Members.Count; valMember++)
{
writer.WriteLine("Member {0,4}, id {1,3}", valMember, result.Clusters[valCluster].Members[valMember].Member.Id);
}
writer.Write("\r\n\r\n");
}
}
}
private void RunClusteringAndGraph()
{
if (chartDataSource != null)
{
Cluster cluster;
distanceMetric = DistanceMetric(currentDistanceMatrix);
try
{
clusterResult = ClusterCalculate();
}
catch (InvalidOperationException)
{
MessageBox.Show("Please try again.");
return;
}
/* Executing scatterplot */
foreach (var dataPoint in chartDataSource)
{
cluster = clusterResult.FindCluster(dataPoint.Origin);
if (cluster != null)
{
dataPoint.Group = string.Format("Cluster {0}", cluster.Id);
}
}
chartDataSource = chartDataSource.OrderBy(item => item.Group).ToList();
scatterPlotControl1.BuildScatterPlot(chartDataSource);
}
}
private static void GetBestPartition(
ClusteringResult <DataPoint> clustering,
IInternalEvaluationCriterion <DataPoint> criterion, string criterionName)
{
// gets coeffs for all cluster-sets
var evals = clustering.EvaluateClustering(criterion);
// saves cluster-sets indexes to CSV file
SaveToCsv(evals, Path.GetFullPath(Path.Combine(RESULTS_PATH, $"{criterionName}.csv")), criterionName);
// gets max coeff
var maxEval = new ClusterSetEvaluation <DataPoint>(null, double.MinValue);
foreach (var eval in evals)
{
if (eval.EvaluationValue > maxEval.EvaluationValue)
{
maxEval = eval;
}
}
// prints cluster set info
Console.WriteLine("======================================");
Console.WriteLine($"Max {criterionName}: {maxEval.EvaluationValue:0.00}");
if (maxEval.ClusterSet == null)
{
return;
}
Console.WriteLine(
$"Clusters at distance: {maxEval.ClusterSet.Dissimilarity:0.00} ({maxEval.ClusterSet.Count})");
foreach (var cluster in maxEval.ClusterSet)
{
Console.WriteLine($" - {cluster}");
}
}
/// <summary>
/// Saves the given <see cref="ClusteringResult{TInstance}" /> to a d3.js dendrogram file.
/// </summary>
/// <typeparam name="TInstance">The type of instance considered.</typeparam>
/// <param name="clustering">The clustering result to be saved to a dendrogram file.</param>
/// <param name="filePath">The path to the file in which to save the clustering dendrogram.</param>
/// <param name="printNames">Whether to include clusters' string representation in their nodes.</param>
/// <param name="formatting">The Json file formatting.</param>
public static void SaveD3DendrogramFile <TInstance>(
this ClusteringResult <TInstance> clustering, string filePath,
bool printNames = true, Formatting formatting = Formatting.None)
where TInstance : IComparable <TInstance>
{
using (var fs = File.Create(filePath))
using (var sw = new StreamWriter(fs, Encoding.UTF8))
{
var writer = new JsonTextWriter(sw)
{
Formatting = formatting
};
WriteJson(clustering.SingleCluster, writer, printNames);
}
}
public static string GetDendrogramJson <TInstance>(
this ClusteringResult <TInstance> clustering,
bool printNames = true,
Formatting formatting = Formatting.None)
where TInstance : IComparable <TInstance>
{
var sb = new StringBuilder();
var sbw = new StringWriter(sb);
var writer = new JsonTextWriter(sbw)
{
Formatting = formatting
};
WriteJson(clustering.SingleCluster, writer, printNames);
return(sb.ToString());
}
private void ClusteringWorkerDoWork(object sender, DoWorkEventArgs e)
{
// checks data points
if (this._dataPoints == null || this._dataPoints.Count == 0)
{
return;
}
// selects linkage criterion
ILinkageCriterion <DataPoint> linkage;
var selectedIndex = e.Argument;
switch (selectedIndex)
{
case 1:
linkage = new CompleteLinkage <DataPoint>(this._dissimilarityMetric);
break;
case 2:
linkage = new SingleLinkage <DataPoint>(this._dissimilarityMetric);
break;
case 3:
linkage = new MinimumEnergyLinkage <DataPoint>(this._dissimilarityMetric);
break;
case 4:
linkage = new CentroidLinkage <DataPoint>(this._dissimilarityMetric, DataPoint.GetMedoid);
break;
case 5:
linkage = new WardsMinimumVarianceLinkage <DataPoint>(
this._dissimilarityMetric, DataPoint.GetMedoid);
break;
default:
linkage = new AverageLinkage <DataPoint>(this._dissimilarityMetric);
break;
}
// clusters data-points
var clusteringAlg = new AgglomerativeClusteringAlgorithm <DataPoint>(linkage);
this._clusteringResult = clusteringAlg.GetClustering(this._dataPoints);
}
private void LoadDataSet()
{
// loads data-points
var parser = new CsvParser();
this._dataPoints = parser.Load(Path.GetFullPath(this.openFileDialog.FileName));
// clears series
this._chartDataPoints.Clear();
this.ChartPoints.Clear();
// adds points to series
var maxX = double.MinValue;
var minX = double.MaxValue;
foreach (var dataPoint in this._dataPoints)
{
var chartDataPoint = new ChartDataPoint(dataPoint.Value[0], dataPoint.Value[1])
{
Label = dataPoint.ID
};
this.ChartPoints.Add(chartDataPoint);
this._chartDataPoints.Add(dataPoint, chartDataPoint);
maxX = Math.Max(maxX, dataPoint.Value[0]);
minX = Math.Min(minX, dataPoint.Value[0]);
}
// resets
this._numClusters = int.MinValue;
this._clusteringResult = null;
this._dissimilarityMetric = new CachedDissimilarityMetric <DataPoint>(new DataPoint(), this._dataPoints);
// adjusts track-bar according to num clusters
this.numClustersTrackBar.SmallChange = (uint)(this.numClustersTrackBar.Maximum / this._dataPoints.Count);
this.numClustersTrackBar.LargeChange = this.numClustersTrackBar.SmallChange * 5;
this.datasetChart.ChartAreas[0].AxisX.Maximum = Math.Ceiling(maxX);
this.datasetChart.ChartAreas[0].AxisX.Minimum = Math.Floor(minX);
}
public void Train(IUnlabeledExampleCollection <SparseVector <double> > dataset, ClusteringResult hierarchy)
{
Utils.ThrowException(dataset == null ? new ArgumentNullException("dataset") : null);
Utils.ThrowException(dataset.Count == 0 ? new ArgumentValueException("dataset") : null);
Utils.ThrowException(hierarchy == null ? new ArgumentNullException("hierarchy") : null);
Utils.ThrowException(hierarchy.Roots.Count == 0 ? new ArgumentValueException("hierarchy") : null);
mModel = new Dictionary <Cluster, ClusterInfo>();
mDataset = dataset;
foreach (Cluster root in hierarchy.Roots)
{
ComputeCentroid(root);
}
mDataset = null;
}
void IHierarchicalModel.Train(IUnlabeledExampleCollection dataset, ClusteringResult hierarchy)
{
Utils.ThrowException(dataset == null ? new ArgumentNullException("dataset") : null);
Utils.ThrowException(!(dataset is IUnlabeledExampleCollection <SparseVector <double> >) ? new ArgumentTypeException("dataset") : null);
Train((IUnlabeledExampleCollection <SparseVector <double> >)dataset, hierarchy); // throws ArgumentNullException, ArgumentValueException
}
static void Main(string[] args)
{
// Get the stop words and stemmer for English.
IStemmer stemmer;
Set <string> .ReadOnly stopWords;
TextMiningUtils.GetLanguageTools(Language.English,
out stopWords, out stemmer);
// Create a tokenizer.
UnicodeTokenizer tokenizer = new UnicodeTokenizer();
tokenizer.MinTokenLen = 2; // Each token must be at least 2
// characters long.
tokenizer.Filter = TokenizerFilter.AlphaStrict; // Tokens
// can consist of alphabetic characters only.
// Load a document corpus from a file. Each line in the file
// represents one document.
string[] docs
= File.ReadAllLines("..\\..\\Data\\YahooFinance.txt");
// Create a bag-of-words space.
BowSpace bowSpc = new BowSpace();
bowSpc.Tokenizer = tokenizer; // Assign the tokenizer.
bowSpc.StopWords = stopWords; // Assign the stop words.
bowSpc.Stemmer = stemmer; // Assign the stemmer.
bowSpc.MinWordFreq = 3; // A term must appear at least 3
// times in the corpus for it to be part of the
// vocabulary.
bowSpc.MaxNGramLen = 3; // Terms consisting of at most 3
// consecutive words will be considered.
bowSpc.WordWeightType = WordWeightType.TfIdf; // Set the
// weighting scheme for the bag-of-words vectors to
// TF-IDF.
bowSpc.NormalizeVectors = true; // The TF-IDF vectors will
// be normalized.
bowSpc.CutLowWeightsPerc = 0.2; // The terms with the lowest
// weights, summing up to 20% of the overall weight sum,
// will be removed from each TF-IDF vector.
bowSpc.Initialize(docs); // Initialize the BOW space.
// Compute 100 clusters of documents.
KMeansFast kMeans = new KMeansFast(100); // Set k to 100.
kMeans.Trials = 3; // Perform 3 repetitions. Take the best
// result.
kMeans.Eps = 0.001; // Stop iterating when the partition
// quality increases for less than 0.001.
ClusteringResult cr = kMeans.Cluster(bowSpc); // Execute.
// Extract the top 5 terms with the highest TF-IDF weights
// from each of the clusters' centroids and output the
// number of documents (companies) in each cluster.
foreach (Cluster cl in cr.Roots)
{
SparseVector <double> .ReadOnly centroid
= cl.ComputeCentroid(bowSpc, CentroidType.NrmL2);
Console.Write(bowSpc.GetKeywordsStr(centroid, 5));
Console.WriteLine(" ({0} companies)", cl.Items.Count);
}
// Output the documents that are contained in the first
// cluster.
foreach (int docIdx in cr.Roots[0].Items)
{
Console.WriteLine(docs[docIdx]);
}
}
public Vector2D[] ComputeLayout(LayoutSettings settings)
{
UnlabeledDataset <SparseVector <double> > dataset = new UnlabeledDataset <SparseVector <double> >(mDataset);
// clustering
mLogger.Info("ComputeLayout", "Clustering ...");
KMeansFast kMeans = new KMeansFast(mKClust);
kMeans.Eps = mKMeansEps;
kMeans.Random = mRandom;
kMeans.Trials = 1;
ClusteringResult clustering = kMeans.Cluster(mDataset); // throws ArgumentValueException
// determine reference instances
UnlabeledDataset <SparseVector <double> > dsRefInst = new UnlabeledDataset <SparseVector <double> >();
foreach (Cluster cluster in clustering.Roots)
{
SparseVector <double> centroid
= cluster.Items.Count > 0 ? cluster.ComputeCentroid(mDataset, CentroidType.NrmL2) : new SparseVector <double>();
dsRefInst.Add(centroid); // dataset of reference instances
dataset.Add(centroid); // add centroids to the main dataset
}
// position reference instances
mLogger.Info("ComputeLayout", "Positioning reference instances ...");
SparseMatrix <double> simMtx = ModelUtils.GetDotProductSimilarity(dsRefInst, mSimThresh, /*fullMatrix=*/ false);
StressMajorizationLayout sm = new StressMajorizationLayout(dsRefInst.Count, new DistFunc(simMtx));
sm.Random = mRandom;
Vector2D[] centrPos = sm.ComputeLayout();
// k-NN
mLogger.Info("ComputeLayout", "Computing similarities ...");
simMtx = ModelUtils.GetDotProductSimilarity(dataset, mSimThresh, /*fullMatrix=*/ true);
mLogger.Info("ComputeLayout", "Constructing system of linear equations ...");
LabeledDataset <double, SparseVector <double> > lsqrDs = new LabeledDataset <double, SparseVector <double> >();
foreach (IdxDat <SparseVector <double> > simMtxRow in simMtx)
{
if (simMtxRow.Dat.Count <= 1)
{
mLogger.Warn("ComputeLayout", "Instance #{0} has no neighborhood.", simMtxRow.Idx);
}
ArrayList <KeyDat <double, int> > knn = new ArrayList <KeyDat <double, int> >(simMtxRow.Dat.Count);
foreach (IdxDat <double> item in simMtxRow.Dat)
{
if (item.Idx != simMtxRow.Idx)
{
knn.Add(new KeyDat <double, int>(item.Dat, item.Idx));
}
}
knn.Sort(DescSort <KeyDat <double, int> > .Instance);
int count = Math.Min(knn.Count, mKNN);
SparseVector <double> eq = new SparseVector <double>();
double wgt = 1.0 / (double)count;
for (int i = 0; i < count; i++)
{
eq.InnerIdx.Add(knn[i].Dat);
eq.InnerDat.Add(-wgt);
}
eq.InnerIdx.Sort(); // *** sort only indices
eq[simMtxRow.Idx] = 1;
lsqrDs.Add(0, eq);
}
Vector2D[] layout = new Vector2D[dataset.Count - mKClust];
for (int i = dataset.Count - mKClust, j = 0; i < dataset.Count; i++, j++)
{
SparseVector <double> eq = new SparseVector <double>(new IdxDat <double>[] { new IdxDat <double>(i, 1) });
lsqrDs.Add(centrPos[j].X, eq);
}
LSqrModel lsqr = new LSqrModel();
lsqr.Train(lsqrDs);
for (int i = 0; i < layout.Length; i++)
{
layout[i].X = lsqr.Solution[i];
}
for (int i = lsqrDs.Count - mKClust, j = 0; i < lsqrDs.Count; i++, j++)
{
lsqrDs[i].Label = centrPos[j].Y;
}
lsqr.Train(lsqrDs);
for (int i = 0; i < layout.Length; i++)
{
layout[i].Y = lsqr.Solution[i];
}
return(settings == null ? layout : settings.AdjustLayout(layout));
}
internal void AddMathSet(IMathSet set)
{
this.m_mathSet = set;
this.m_cResult = null;
}
public void AddMathSet(IMathSet set)
{
this.m_mathSet = set;
this.m_cResult = null;
}
//private double GetQual()
//{
// double clustQual = 0;
// foreach (Centroid centroid in mCentroids)
// {
// foreach (int itemIdx in centroid.CurrentItems)
// {
// clustQual += centroid.GetDotProduct(mDataset[itemIdx]);
// }
// }
// clustQual /= (double)mDataset.Count;
// return clustQual;
//}
// TODO: exceptions
public ClusteringResult Update(int dequeueN, IEnumerable <SparseVector <double> > addList, ref int iter)
{
StopWatch stopWatch = new StopWatch();
// update centroid data (1)
foreach (CentroidData centroid in mCentroids)
{
foreach (int item in centroid.CurrentItems)
{
if (item >= dequeueN)
{
centroid.Items.Add(item);
}
}
centroid.Update(mDataset);
centroid.UpdateCentroidLen();
}
//Console.WriteLine(">>> {0} >>> update centroid data (1)", stopWatch.TotalMilliseconds);
stopWatch.Reset();
// update dataset
mDataset.RemoveRange(0, dequeueN);
int ofs = mDataset.Count;
mDataset.AddRange(addList);
//Console.WriteLine(">>> {0} >>> update dataset", stopWatch.TotalMilliseconds);
stopWatch.Reset();
// update centroid data (2)
foreach (CentroidData centroid in mCentroids)
{
Set <int> itemsOfs = new Set <int>();
foreach (int item in centroid.CurrentItems)
{
itemsOfs.Add(item - dequeueN);
}
centroid.CurrentItems.Inner.SetItems(itemsOfs);
centroid.Items.SetItems(itemsOfs);
}
//Console.WriteLine(">>> {0} >>> update centroid data (2)", stopWatch.TotalMilliseconds);
stopWatch.Reset();
// assign new instances
double bestClustQual = 0;
{
mLogger.Info("Update", "Initializing ...");
int i = 0;
foreach (SparseVector <double> example in addList)
{
double maxSim = double.MinValue;
ArrayList <int> candidates = new ArrayList <int>();
for (int j = 0; j < mK; j++)
{
double sim = mCentroids[j].GetDotProduct(example);
if (sim > maxSim)
{
maxSim = sim;
candidates.Clear();
candidates.Add(j);
}
else if (sim == maxSim)
{
candidates.Add(j);
}
}
if (candidates.Count > 1)
{
candidates.Shuffle(mRnd);
}
if (candidates.Count > 0) // *** is this always true?
{
mCentroids[candidates[0]].Items.Add(ofs + i);
}
i++;
}
// update centroids
foreach (CentroidData centroid in mCentroids)
{
centroid.Update(mDataset);
centroid.UpdateCentroidLen();
}
//Console.WriteLine(GetQual());
foreach (CentroidData centroid in mCentroids)
{
foreach (int itemIdx in centroid.CurrentItems)
{
bestClustQual += centroid.GetDotProduct(mDataset[itemIdx]);
}
}
bestClustQual /= (double)mDataset.Count;
mLogger.Info("Update", "Quality: {0:0.0000}", bestClustQual);
}
//Console.WriteLine(">>> {0} >>> assign new instances", stopWatch.TotalMilliseconds);
stopWatch.Reset();
// main k-means loop
iter = 0;
while (true)
{
iter++;
mLogger.Info("Update", "Iteration {0} ...", iter);
// assign items to clusters
for (int i = 0; i < mDataset.Count; i++)
{
SparseVector <double> example = mDataset[i];
double maxSim = double.MinValue;
ArrayList <int> candidates = new ArrayList <int>();
for (int j = 0; j < mK; j++)
{
double sim = mCentroids[j].GetDotProduct(example);
if (sim > maxSim)
{
maxSim = sim;
candidates.Clear();
candidates.Add(j);
}
else if (sim == maxSim)
{
candidates.Add(j);
}
}
if (candidates.Count > 1)
{
candidates.Shuffle(mRnd);
}
if (candidates.Count > 0) // *** is this always true?
{
mCentroids[candidates[0]].Items.Add(i);
}
}
//
// *** OPTIMIZE THIS with GetDotProductSimilarity (see this.Cluster) !!! ***
//
//Console.WriteLine(">>> {0} >>> loop: assign items to clusters", stopWatch.TotalMilliseconds);
stopWatch.Reset();
double clustQual = 0;
// update centroids
foreach (CentroidData centroid in mCentroids)
{
centroid.Update(mDataset);
centroid.UpdateCentroidLen();
}
//Console.WriteLine(GetQual());
foreach (CentroidData centroid in mCentroids)
{
foreach (int itemIdx in centroid.CurrentItems)
{
clustQual += centroid.GetDotProduct(mDataset[itemIdx]);
}
}
clustQual /= (double)mDataset.Count;
//Console.WriteLine(">>> {0} >>> loop: update centroids", stopWatch.TotalMilliseconds);
stopWatch.Reset();
mLogger.Info("Update", "Quality: {0:0.0000} Diff: {1:0.0000}", clustQual, clustQual - bestClustQual);
// check if done
if (clustQual - bestClustQual <= mEps)
{
break;
}
bestClustQual = clustQual;
}
// save the result
ClusteringResult clustering = new ClusteringResult();
for (int i = 0; i < mK; i++)
{
clustering.AddRoot(new Cluster());
clustering.Roots.Last.Items.AddRange(mCentroids[i].Items);
}
return(clustering);
}
public void AddClusteringResult(ClusteringResult set)
{
this.m_cResult = set;
}
public ClusteringResult Cluster(IUnlabeledExampleCollection <SparseVector <double> > dataset)
{
Utils.ThrowException(dataset == null ? new ArgumentNullException("dataset") : null);
Utils.ThrowException(dataset.Count < mK ? new ArgumentValueException("dataset") : null);
mDataset = new UnlabeledDataset <SparseVector <double> >(dataset);
ClusteringResult clustering = null;
double globalBestClustQual = 0;
for (int trial = 1; trial <= mTrials; trial++)
{
mLogger.Info("Cluster", "Clustering trial {0} of {1} ...", trial, mTrials);
ArrayList <CentroidData> centroids = new ArrayList <CentroidData>(mK);
ArrayList <int> bestSeeds = null;
for (int i = 0; i < mK; i++)
{
centroids.Add(new CentroidData());
}
// select seed items
double minSim = double.MaxValue;
ArrayList <int> tmp = new ArrayList <int>(mDataset.Count);
for (int i = 0; i < mDataset.Count; i++)
{
tmp.Add(i);
}
for (int k = 0; k < 3; k++)
{
ArrayList <SparseVector <double> > seeds = new ArrayList <SparseVector <double> >(mK);
tmp.Shuffle(mRnd);
for (int i = 0; i < mK; i++)
{
seeds.Add(mDataset[tmp[i]]);
}
// assess quality of seed items
double simAvg = 0;
foreach (SparseVector <double> seed1 in seeds)
{
foreach (SparseVector <double> seed2 in seeds)
{
if (seed1 != seed2)
{
simAvg += DotProductSimilarity.Instance.GetSimilarity(seed1, seed2);
}
}
}
simAvg /= (double)(mK * mK - mK);
//Console.WriteLine(simAvg);
if (simAvg < minSim)
{
minSim = simAvg;
bestSeeds = new ArrayList <int>(mK);
for (int i = 0; i < mK; i++)
{
bestSeeds.Add(tmp[i]);
}
}
}
ArrayList <KeyDat <double, int> > medoids = new ArrayList <KeyDat <double, int> >(mK);
for (int i = 0; i < mK; i++)
{
centroids[i].Items.Add(bestSeeds[i]);
centroids[i].Update(mDataset);
centroids[i].UpdateCentroidLen();
medoids.Add(new KeyDat <double, int>(-1, bestSeeds[i]));
}
double[,] dotProd = new double[mDataset.Count, mK];
SparseMatrix <double> dsMat = ModelUtils.GetTransposedMatrix(mDataset);
// main loop
int iter = 0;
double bestClustQual = 0;
double clustQual;
while (true)
{
iter++;
mLogger.Info("Cluster", "Iteration {0} ...", iter);
clustQual = 0;
// assign items to clusters
//StopWatch stopWatch = new StopWatch();
int j = 0;
foreach (CentroidData cen in centroids)
{
SparseVector <double> cenVec = cen.GetSparseVector();
double[] dotProdSimVec = ModelUtils.GetDotProductSimilarity(dsMat, mDataset.Count, cenVec);
for (int i = 0; i < dotProdSimVec.Length; i++)
{
if (dotProdSimVec[i] > 0)
{
dotProd[i, j] = dotProdSimVec[i];
}
}
j++;
}
for (int dsInstIdx = 0; dsInstIdx < mDataset.Count; dsInstIdx++)
{
double maxSim = double.MinValue;
ArrayList <int> candidates = new ArrayList <int>();
for (int cenIdx = 0; cenIdx < mK; cenIdx++)
{
double sim = dotProd[dsInstIdx, cenIdx];
if (sim > maxSim)
{
maxSim = sim;
candidates.Clear();
candidates.Add(cenIdx);
}
else if (sim == maxSim)
{
candidates.Add(cenIdx);
}
}
if (candidates.Count > 1)
{
candidates.Shuffle(mRnd);
}
if (candidates.Count > 0) // *** is this always true?
{
centroids[candidates[0]].Items.Add(dsInstIdx);
clustQual += maxSim;
if (medoids[candidates[0]].Key < maxSim)
{
medoids[candidates[0]] = new KeyDat <double, int>(maxSim, dsInstIdx);
}
}
}
//Console.WriteLine(stopWatch.TotalMilliseconds);
clustQual /= (double)mDataset.Count;
mLogger.Info("Cluster", "Quality: {0:0.0000}", clustQual);
// compute new centroids
for (int i = 0; i < mK; i++)
{
centroids[i].Update(mDataset);
centroids[i].UpdateCentroidLen();
}
// check if done
if (iter > 1 && clustQual - bestClustQual <= mEps)
{
break;
}
bestClustQual = clustQual;
for (int i = 0; i < medoids.Count; i++)
{
medoids[i] = new KeyDat <double, int>(-1, medoids[i].Dat);
}
}
if (trial == 1 || clustQual > globalBestClustQual)
{
globalBestClustQual = clustQual;
mCentroids = centroids;
mMedoids = medoids;
// save the result
clustering = new ClusteringResult();
for (int i = 0; i < mK; i++)
{
clustering.AddRoot(new Cluster());
clustering.Roots.Last.Items.AddRange(centroids[i].Items);
}
}
}
return(clustering);
}
public Vector2D[] ComputeLayout(LayoutSettings settings)
{
// clustering
mLogger.Info("ComputeLayout", "Clustering ...");
mKMeans = new IncrementalKMeans(mKClust);
mKMeans.Eps = mKMeansEps;
mKMeans.Random = mRandom;
mKMeans.Trials = 3;
ClusteringResult clustering = mKMeans.Cluster(mDataset); // throws ArgumentValueException
// determine reference instances
UnlabeledDataset <SparseVector <double> > dsRefInst = new UnlabeledDataset <SparseVector <double> >();
foreach (SparseVector <double> centroid in mKMeans.GetCentroids())
{
dsRefInst.Add(centroid); // dataset of reference instances
mDataset.Add(centroid); // add centroids to the main dataset
}
// position reference instances
mLogger.Info("ComputeLayout", "Positioning reference instances ...");
SparseMatrix <double> simMtx = ModelUtils.GetDotProductSimilarity(dsRefInst, mSimThresh, /*fullMatrix=*/ false);
StressMajorizationLayout sm = new StressMajorizationLayout(dsRefInst.Count, new DistFunc(simMtx));
sm.Random = mRandom;
sm.MaxSteps = int.MaxValue;
sm.MinDiff = 0.00001;
mRefPos = sm.ComputeLayout();
// k-NN
mLogger.Info("ComputeLayout", "Computing similarities ...");
simMtx = ModelUtils.GetDotProductSimilarity(mDataset, mSimThresh, /*fullMatrix=*/ true);
mLogger.Info("ComputeLayout", "Constructing system of linear equations ...");
LabeledDataset <double, SparseVector <double> > lsqrDs = new LabeledDataset <double, SparseVector <double> >();
mPatches = new ArrayList <Patch>(mDataset.Count);
for (int i = 0; i < mDataset.Count; i++)
{
mPatches.Add(new Patch(i));
}
foreach (IdxDat <SparseVector <double> > simMtxRow in simMtx)
{
if (simMtxRow.Dat.Count <= 1)
{
mLogger.Warn("ComputeLayout", "Instance #{0} has no neighborhood.", simMtxRow.Idx);
}
ArrayList <KeyDat <double, int> > knn = new ArrayList <KeyDat <double, int> >(simMtxRow.Dat.Count);
foreach (IdxDat <double> item in simMtxRow.Dat)
{
if (item.Idx != simMtxRow.Idx)
{
knn.Add(new KeyDat <double, int>(item.Dat, item.Idx));
}
}
knn.Sort(DescSort <KeyDat <double, int> > .Instance);
int count = Math.Min(knn.Count, mKNnExt);
for (int i = 0; i < count; i++)
{
mPatches[simMtxRow.Idx].List.Add(new KeyDat <double, Patch>(knn[i].Key, mPatches[knn[i].Dat]));
}
mPatches[simMtxRow.Idx].ProcessList();
count = Math.Min(knn.Count, mKNn);
SparseVector <double> eq = new SparseVector <double>();
double wgt = 1.0 / (double)count;
for (int i = 0; i < count; i++)
{
eq.InnerIdx.Add(knn[i].Dat);
eq.InnerDat.Add(-wgt);
}
eq.InnerIdx.Sort(); // *** sort only indices
eq[simMtxRow.Idx] = 1;
lsqrDs.Add(0, eq);
}
Vector2D[] layout = new Vector2D[mDataset.Count - mKClust];
for (int i = mDataset.Count - mKClust, j = 0; i < mDataset.Count; i++, j++)
{
SparseVector <double> eq = new SparseVector <double>(new IdxDat <double>[] { new IdxDat <double>(i, 1) });
lsqrDs.Add(mRefPos[j].X, eq);
}
LSqrModel lsqr = new LSqrModel();
lsqr.Train(lsqrDs);
mSolX = lsqr.Solution.GetWritableCopy();
for (int i = 0; i < layout.Length; i++)
{
layout[i].X = lsqr.Solution[i];
}
for (int i = lsqrDs.Count - mKClust, j = 0; i < lsqrDs.Count; i++, j++)
{
lsqrDs[i].Label = mRefPos[j].Y;
}
lsqr.Train(lsqrDs);
mSolY = lsqr.Solution.GetWritableCopy();
for (int i = 0; i < layout.Length; i++)
{
layout[i].Y = lsqr.Solution[i];
}
return(settings == null ? layout : settings.AdjustLayout(layout));
}
internal void AddClusteringResult(ClusteringResult set)
{
this.m_cResult = set;
}
public static void Main(string[] args)
{
Plot generatedDataPlot = new Plot();
Spawner spawner = new Spawner(STD_DEV);
List <PointF> allPoints = new List <PointF>();
for (int i = 0; i < CLUSTER_COUNT; ++i)
{
spawner.ResetCenter(MIN_CENTER_DISTANCE, MAX_CENTER_DISTANCE);
PointF[] points = spawner.Spawn(POINT_COUNT);
allPoints.AddRange(points);
Color color = generatedDataPlot.GetNextColor();
generatedDataPlot.AddScatterPoints(points, color, label: $"Points {i + 1}");
generatedDataPlot.AddPoint(spawner.Center.X, spawner.Center.Y, color, 25);
}
generatedDataPlot.Legend();
PlotForm generatedDataPlotForm = new PlotForm(generatedDataPlot, "source_data");
generatedDataPlotForm.ShowDialog();
Plot grayDataPlot = new Plot();
grayDataPlot.AddScatterPoints(allPoints.ToArray(), label: "Gray points");
grayDataPlot.Legend();
PlotForm grayDataPlotForm = new PlotForm(grayDataPlot, "gray_data");
grayDataPlotForm.ShowDialog();
KMeansClusterizer clusterizer = new KMeansClusterizer();
List <Dictionary <PointF, List <PointF> > > clusterizingHistory = clusterizer.Clusterize(allPoints, CLUSTER_COUNT);
PlotForm resultPlotForm = new PlotForm(CreateClusterizingPlot(clusterizingHistory.Last()), "crusterized");
resultPlotForm.ShowDialog();
PlotForm historyForm = new PlotForm(clusterizingHistory.Select(c => CreateClusterizingPlot(c)).ToList(), "history_");
historyForm.ShowDialog();
CentroidLinkage <DataPoint> linkage = new CentroidLinkage <DataPoint>(
new DissimilarityMetric(),
cluster => new DataPoint(
cluster.Average(p => p.X),
cluster.Average(p => p.Y)
)
);
AgglomerativeClusteringAlgorithm <DataPoint> algorithm = new AgglomerativeClusteringAlgorithm <DataPoint>(linkage);
HashSet <DataPoint> dataPoints = allPoints.Select(p => new DataPoint(p)).ToHashSet();
ClusteringResult <DataPoint> clusteringResult = algorithm.GetClustering(dataPoints);
ClusterSet <DataPoint> result = clusteringResult[clusteringResult.Count - 3];
Plot aglomeraPlot = new Plot();
foreach (Cluster <DataPoint> resultCluster in result)
{
Color color = aglomeraPlot.GetNextColor();
aglomeraPlot.AddScatterPoints(
resultCluster.Select(p => (double)p.X).ToArray(),
resultCluster.Select(p => (double)p.Y).ToArray(),
color
);
aglomeraPlot.AddPoint(
resultCluster.Select(p => p.X).Average(),
resultCluster.Select(p => p.Y).Average(),
color, 25
);
}
PlotForm aglomeraForm = new PlotForm(aglomeraPlot, "aglomera");
aglomeraForm.ShowDialog();
clusteringResult.SaveD3DendrogramFile(Environment.CurrentDirectory + "/dendro.json");
Console.ReadLine();
}
