[heatmap-adjacency] Network Adjacency Matrix Heatmap

[MarkusNeusinger]

Description

A matrix-based representation of a network/graph where rows and columns represent nodes and cell color indicates the presence or weight of edges between them. This complements existing node-link network visualizations by excelling at revealing clusters, structural patterns, and density in large or dense networks where node-link diagrams become hairball-like.

Applications

• Visualizing social network connections in community detection
• Displaying brain connectivity matrices in neuroscience
• Analyzing trade relationships between countries
• Showing co-occurrence patterns in text analysis

Data

• source (categorical) — source node name
• target (categorical) — target node name
• weight (float, optional) — edge weight or connection strength
• Size: 20–200 nodes

Notes

• Nodes should be reorderable (e.g., by cluster, degree, or community)
• Symmetric matrix for undirected graphs
• Color intensity maps to edge weight
• Optional dendrogram on axes for clustered ordering

[claude]

✅ Spec ID: heatmap-adjacency

Rationale: Combines the heatmap visualization form with the specific adjacency-matrix concept from graph theory — distinct from existing network node-link diagrams and generic clustered/correlation heatmaps.

Valid and useful plot type that fills a gap between our network visualizations (node-link) and matrix heatmaps. No duplicate found among existing specs.

[github-actions]

✅ Specification Ready: heatmap-adjacency

PR: #5258

---

specification.md

heatmap-adjacency: Network Adjacency Matrix Heatmap

Description

A matrix-based representation of a network or graph where rows and columns represent nodes and cell color indicates the presence or weight of edges between them. This visualization complements node-link diagrams by excelling at revealing clusters, structural patterns, and density in large or dense networks where node-link layouts become cluttered. Reordering nodes by cluster, degree, or community membership exposes block-diagonal structure and makes group boundaries immediately visible.

Applications

• Visualizing social network connections to detect community structure and cliques
• Displaying brain connectivity matrices in neuroscience to identify functionally linked regions
• Analyzing bilateral trade relationships between countries to reveal trading blocs
• Showing co-occurrence patterns in text analysis to find related terms or topics

Data

• source (categorical) — source node identifier
• target (categorical) — target node identifier
• weight (float, optional) — edge weight or connection strength; defaults to 1 (binary presence) when absent
• Size: 20–200 nodes (adjacency matrix will be nodes × nodes)
• Example: A social network of 50 people with friendship strength scores

Notes

• Nodes should be reorderable by cluster, degree, or community assignment to expose block-diagonal structure
• For undirected graphs the matrix is symmetric; implementations should fill both triangles
• Color intensity maps to edge weight; absent edges should use a distinct background (e.g., white or near-white)
• Optional dendrogram along axes to show hierarchical clustering of node ordering
• Include a colorbar legend showing the weight scale
• Axis tick labels should display node names; for large networks consider showing only group boundaries

---

Next: Add approved label to merge this specification to main.

---

🤖 spec-create workflow

[github-actions]

🚀 Specification Merged!

Specification: heatmap-adjacency
PR: #5258 (merged)

The specification is now in main and synced to PostgreSQL.

---

Next: Add generate:{library} labels to start generating implementations.

Available libraries: matplotlib, seaborn, plotly, bokeh, altair, plotnine, pygal, highcharts, letsplot

Or use workflow dispatch for bulk operations.

---

🤖 spec-create workflow