OPTICSOptions<T>

Options & Configuration · AiDotNet.Clustering.Options

Configuration options for OPTICS clustering.

Inheritance: ModelOptions → ClusteringOptions<T> → OPTICSOptions<T>

For Beginners

OPTICS is like DBSCAN but smarter about finding clusters.

The problem with DBSCAN:

  • You must choose epsilon carefully
  • One epsilon might not work for all clusters

OPTICS solves this by:

  • Computing “reachability distances” for all points
  • Creating an ordering that reveals cluster structure
  • Allowing you to extract clusters at different density levels

Think of it like hiking through mountains:

  • Valleys are dense clusters
  • Peaks are boundaries between clusters
  • You can see the structure at different altitudes

How It Works

OPTICS (Ordering Points To Identify the Clustering Structure) is a density-based algorithm similar to DBSCAN but doesn’t require a specific epsilon value upfront. Instead, it creates an ordering of points and their reachability distances.

Constructors

Constructor Summary
OPTICSOptions()

Properties

Property Summary
NeighborAlgorithm Algorithm { get; set; } Gets or sets the neighbor finding algorithm.
OPTICSExtractionMethod ExtractionMethod { get; set; } Gets or sets the cluster extraction method.
bool PredecessorCorrection { get; set; } Gets or sets the predecessor correction setting.
double MaxEpsilon { get; set; } Gets or sets the maximum epsilon distance.
double MinClusterSizeFraction { get; set; } Gets or sets the minimum cluster size as fraction of total points.
double Xi { get; set; } Gets or sets the xi parameter for Xi extraction method.
double? ClusterEpsilon { get; set; } Gets or sets the epsilon for DBSCAN-style extraction.
int LeafSize { get; set; } Gets or sets the leaf size for tree algorithms.
int MinSamples { get; set; } Gets or sets the minimum number of samples in a neighborhood.