Module hierarchy 

Hierarchical clustering functions

This module provides hierarchical clustering algorithms for agglomerative clustering, linkage methods, and dendrogram visualization.

Examples

use scirs2_core::ndarray::{Array2, ArrayView2};
use scirs2_cluster::hierarchy::{linkage, fcluster, LinkageMethod, Metric};

// Example data
let data = Array2::from_shape_vec((6, 2), vec![
    1.0, 2.0,
    1.2, 1.8,
    0.8, 1.9,
    3.7, 4.2,
    3.9, 3.9,
    4.2, 4.1,
]).unwrap();

// Calculate linkage matrix using Ward's method
let linkage_matrix = linkage(data.view(), LinkageMethod::Ward, Metric::Euclidean).unwrap();

// Form flat clusters by cutting the dendrogram at a height that gives 2 clusters
let num_clusters = 2;
let labels = fcluster(&linkage_matrix, num_clusters, None).unwrap();

// Print the results
println!("Cluster assignments: {:?}", labels);

Re-exports

pub use self::agglomerative::cut_tree_by_distance;

pub use self::agglomerative::cut_tree_by_inconsistency;

pub use self::cluster_extraction::estimate_optimal_clusters;

pub use self::cluster_extraction::extract_clusters_multi_criteria;

pub use self::cluster_extraction::prune_clusters;

pub use self::condensed_matrix::condensed_size;

pub use self::condensed_matrix::condensed_to_square;

pub use self::condensed_matrix::get_distance;

pub use self::condensed_matrix::points_from_condensed_size;

pub use self::condensed_matrix::square_to_condensed;

pub use self::condensed_matrix::validate_condensed_matrix;

pub use self::dendrogram::cophenet;

pub use self::dendrogram::dendrogram;

pub use self::dendrogram::inconsistent;

pub use self::dendrogram::optimal_leaf_ordering;

pub use self::disjoint_set::DisjointSet;

pub use self::leaf_ordering::apply_leaf_ordering;

pub use self::leaf_ordering::optimal_leaf_ordering_exact;

pub use self::leaf_ordering::optimal_leaf_ordering_heuristic;

pub use self::optimized_ward::lance_williams_ward_update;

pub use self::optimized_ward::memory_efficient_ward_linkage;

pub use self::optimized_ward::optimized_ward_linkage;

pub use self::validation::validate_cluster_consistency;

pub use self::validation::validate_cluster_extraction_params;

pub use self::validation::validate_distance_matrix;

pub use self::validation::validate_linkage_matrix;

pub use self::validation::validate_monotonic_distances;

pub use self::validation::validate_square_distance_matrix;

pub use self::visualization::create_dendrogramplot;

pub use self::visualization::get_color_palette;

pub use self::visualization::Branch;

pub use self::visualization::ColorScheme;

pub use self::visualization::ColorThreshold;

pub use self::visualization::DendrogramConfig;

pub use self::visualization::DendrogramOrientation;

pub use self::visualization::DendrogramPlot;

pub use self::visualization::Leaf;

pub use self::visualization::LegendEntry;

pub use self::visualization::TruncateMode;

Modules

agglomerative

Agglomerative clustering methods

cluster_extraction

Enhanced cluster extraction utilities

condensed_matrix

Condensed distance matrix utilities

dendrogram

Dendrogram visualization and analysis tools

disjoint_set

Disjoint Set (Union-Find) data structure for connectivity queries

leaf_ordering

Optimal leaf ordering algorithms for dendrograms

linkage

Linkage methods for hierarchical clustering

optimized_ward

Optimized Ward’s linkage implementation for hierarchical clustering

parallel_linkage

Parallel implementation of linkage methods for hierarchical clustering

validation

Validation utilities for hierarchical clustering

visualization

Enhanced visualization utilities for dendrograms

Enums

ClusterCriterion

Criterion for forming flat clusters from a hierarchical clustering

LinkageMethod

Linkage methods for hierarchical clustering

Metric

Distance metrics for hierarchical clustering

Functions

condensed_index_to_coords

Converts a condensed distance matrix index to (i, j) coordinates

coords_to_condensed_index

Converts (i, j) coordinates to a condensed distance matrix index

fcluster

Forms flat clusters from a hierarchical clustering result

fcluster_generic

Forms flat clusters from a hierarchical clustering result with generic threshold type

linkage

Performs hierarchical clustering using the specified linkage method

parallel_linkage

Performs parallel hierarchical clustering using the specified linkage method