About this visualization
The data. This is the public SNAP Facebook ego-networks dataset from Stanford — 4,039 users and 88,234 friendship links. It is a fully anonymized research dataset: every node is an integer ID with all profile data removed. It contains no personal information.
What you see. Each dot is a person, each line a friendship. Positions come from a precomputed force-directed layout, so tightly connected friend circles cluster together. Colors mark communities found with the Louvain algorithm (modularity —) — these emerge purely from link structure, yet correspond to real-world social circles like schools, workplaces and families.
How it's built. Layout and community detection run once in a Python (networkx) build step; the browser only renders precomputed positions on an HTML5 canvas for smooth zoom/pan across all 88k edges.