spiral-radial-clustering-display

What problem does it solve?

This Skill visualizes hierarchical clustering structure by combining macro-clusters, micro-clusters, decorrelated ordering, and centrality into a single interactive 2D spiral-radial display.

Core Features & Use Cases

• Hierarchical layer encoding: maps macro clustering to color, micro clustering agreement to opacity, decorrelated ordering to angle, and centrality/hubness to size.
• UMAP-based 3D→2D projection: constructs a 3D feature space (macro ID, distance-to-macro centroid, ordering-derived angle) and projects to 2D while preserving neighborhood/topological structure.
• Interactive Gestalt visualization: produces a Plotly scatter where zoom/pan and hover metadata help you inspect layer membership at point level.
• Use Case: analyze a dataset (e.g., songs, embeddings, customer segments) where you want to see whether micro-clusters organize into spiral/radial patterns within macro regions.

Quick Start

Use the spiral-radial-clustering-display skill to produce an interactive HTML visualization from your features, GMM macro labels, HDBSCAN micro labels, and decorrelated ordering indices.