OPTICS<T>

Models & Types · AiDotNet.Clustering.Density

OPTICS (Ordering Points To Identify the Clustering Structure) implementation.

Inheritance: ClusteringBase<T> → OPTICS<T>

Implements: IGeneratedParameterRegistrar<T>

For Beginners

OPTICS creates a “profile” of your data’s cluster structure.

Imagine walking through a landscape:

  • In dense areas (clusters), reachability is low
  • At cluster boundaries, reachability spikes up
  • This creates a “reachability plot” showing cluster structure

Benefits:

  • No need to choose epsilon beforehand
  • Can find clusters at multiple scales
  • Better for clusters with varying densities

How It Works

OPTICS is a density-based clustering algorithm that creates an ordering of points based on their density-reachability. Unlike DBSCAN, it doesn’t require a fixed epsilon and can reveal hierarchical clustering structure.

Key concepts:

  • Core distance: Minimum radius to include MinSamples points
  • Reachability distance: max(core distance, distance to point)
  • Cluster ordering: Sequence of points with reachability distances

Example

using AiDotNet;
using AiDotNet.Data.Loaders;
using AiDotNet.Clustering.Density;
using AiDotNet.Tensors.LinearAlgebra;

var data = new Matrix<double>(6, 2);
double[][] rows = { new[] { 1.0, 1.0 }, new[] { 1.2, 0.9 }, new[] { 1.1, 1.1 },
                    new[] { 8.0, 8.0 }, new[] { 8.2, 7.9 }, new[] { 7.9, 8.1 } };
for (int i = 0; i < 6; i++) { data[i, 0] = rows[i][0]; data[i, 1] = rows[i][1]; }

var result = await new AiModelBuilder<double, Matrix<double>, Vector<double>>()
    .ConfigureModel(new OPTICS<double>())
    .ConfigureDataLoader(DataLoaders.FromMatrix(data))
    .BuildAsync();

var labels = result.Predict(data);
Console.WriteLine($"OPTICS: clustered {labels.Length} points.");

Constructors

Constructor Summary
OPTICS(OPTICSOptions<T> options) Initializes a new OPTICS instance.

Properties

Property Summary
Vector<T> CoreDistances { get; } Gets the core distances for each point.
Vector<T> ReachabilityDistances { get; } Gets the reachability distances in ordering.
int[] Ordering { get; } Gets the cluster ordering.

Methods

Method Summary
IFullModel<T, Matrix<T>, Vector<T>> WithParameters(Vector<T> parameters)
ModelOptions GetOptions()
ValueTuple<int, T>[] GetReachabilityPlot() Gets the reachability plot data.
Vector<T> ExtractClustersAtEpsilon(double epsilon) Extracts clusters at a different epsilon value.
Vector<T> FitPredict(Matrix<T> x)
Vector<T> Predict(Matrix<T> x)
void Train(Matrix<T> x, Vector<T> y)

Fields

Field Summary
int NoiseLabel Noise label for points not in any cluster.