Tutorial 1 — A combinatorial graph from a file

Import a network from an edge-list file, lay it out in 3D, compute node importance and community structure, encode those metrics visually, and render a publication-quality image.

What you need: a CSV edge list (two columns of node identifiers, optionally a weight and extra attribute columns) or a GEXF/GraphML file. Any small-to-medium graph works.

1. Import the graph

  1. Open the SciGraphs tab and expand Data.
  2. In Step 1, keep the source on File and set the Filepath. For a CSV, set the Delimiter; native GEXF files are detected automatically. Press Load File.
  3. In Step 2, map Source Nodes and Target Nodes to the right columns. Leave Directed Graph off for an undirected network. Tick any extra columns under Import as Attributes to bring them in.
  4. In Step 3, press Create Graph.
  5. In Step 4, press Setup Visual so the graph becomes renderable geometry (instanced node spheres and edge tubes).

2. Lay out the nodes

  1. Expand Layout & Positioning.
  2. In Algorithm Selection, choose an algorithm. For most graphs Yifan Hu (sfdp) gives high-quality 3D results; for very large graphs prefer igraph DrL.
  3. Optionally tune values in Algorithm Parameters.
  4. Press Apply Layout Now in Quick Settings for a one-shot layout — or open Interactive Mode, press Start Execution, watch convergence, and Bake to Animation.

3. Analyze

  1. Expand Analysis.
  2. Under Centrality Metrics, pick a Method (e.g. betweenness) and press Calculate. The result is stored as a node attribute.
  3. Under Community Detection, choose an algorithm (e.g. Infomap) and press Detect Communities. The Surprise quality metric is computed automatically.
  4. Optionally run Statistical Analysis for global descriptors (density, diameter, clustering, degree distribution).

4. Encode the metrics visually

  1. Expand Visualization.
  2. In Appearance, set base node and edge sizes.
  3. Drive node colour from the community id or centrality, and node size from degree or centrality, using a scientific colormap.
  4. In Scene Setup, apply a Material Preset and Setup Lighting.
  5. In Edge Style, pick a preset such as Gephi or Bundled depending on edge density.
  6. Optionally enable Text Labels and use the Attribute Filter to label only the most central nodes.

5. Render and export

  1. Set the camera and choose Cycles (photorealistic) or EEVEE (fast) in Blender’s render properties, then render (F12).
  2. Expand Export & Tools to write the graph to GEXF/GraphML/JSON/Pajek, export node positions, or generate a statistics report.
Tip

Because metrics are stored as named attributes, you can re-colour or re-size the same layout on a different metric instantly — no need to recompute the layout.

Back to top