@ruvector/graph Data Generator

@ruvector/graph-data-generator

AI-powered synthetic graph data generation with configurable topology models, density controls, and optional LLM-powered property generation via OpenRouter/Kimi K2 for realistic test data.

Quick Command Reference

Installation

Hub install (recommended): npx ruvector@latest includes this package. Standalone: npx @ruvector/graph-data-generator@latest See Installation Guide for the full ecosystem.

Core API

GraphGenerator Constructor

import { GraphGenerator } from '@ruvector/graph-data-generator';

const gen = new GraphGenerator({
  nodes: 1000,
  density: 0.1,
  topology: 'scale-free',
  labels: ['Person', 'Company', 'Product'],
  edgeLabels: ['KNOWS', 'WORKS_AT', 'PURCHASED'],
  propertyGenerator: 'random',   // 'random' | 'ai'
  seed: 42,
});

Constructor Options:

Topology Models:

Generation

// Generate graph
const graph = await gen.generate();

// Generate with AI properties (requires OPENROUTER_API_KEY)
const gen = new GraphGenerator({
  nodes: 100,
  propertyGenerator: 'ai',
  aiModel: 'moonshotai/kimi-k2',
});
const graph = await gen.generate();

Templates

// Social network
const gen = GraphGenerator.template('social-network', {
  users: 500,
  avgFriends: 10,
  communities: 5,
});

// Knowledge graph
const gen = GraphGenerator.template('knowledge-graph', {
  topics: 100,
  conceptsPerTopic: 20,
  crossLinks: 0.05,
});

// E-commerce
const gen = GraphGenerator.template('ecommerce', {
  users: 1000,
  products: 500,
  purchases: 5000,
});

// Dependency graph
const gen = GraphGenerator.template('dependency-graph', {
  packages: 200,
  avgDependencies: 5,
});

Export

// Export to various formats
await gen.export('json', './output.json');
await gen.export('graphml', './output.xml');
await gen.export('csv', './output/');     // vertices.csv + edges.csv
await gen.export('cypher', './output.cypher');  // Cypher CREATE statements
await gen.export('dot', './output.dot');        // Graphviz DOT
await gen.export('adjacency', './output.json'); // Adjacency list

Statistics

const stats = gen.stats();
// { nodes: 1000, edges: 4500, avgDegree: 9, density: 0.009, components: 1, diameter: 8 }

Common Patterns

Generate Test Data for Graph DB

const gen = new GraphGenerator({
  nodes: 10000,
  density: 0.01,
  topology: 'scale-free',
  labels: ['User', 'Post', 'Tag'],
  edgeLabels: ['AUTHORED', 'TAGGED', 'FOLLOWS'],
});
const graph = await gen.generate();
await gen.export('cypher', './seed.cypher');
// Load into database

Benchmark Data Generation

for (const size of [1000, 10000, 100000]) {
  const gen = new GraphGenerator({ nodes: size, density: 0.01, seed: 42 });
  await gen.generate();
  await gen.export('json', `./bench-${size}.json`);
}

AI-Powered Realistic Properties

const gen = new GraphGenerator({
  nodes: 100,
  propertyGenerator: 'ai',
  aiModel: 'moonshotai/kimi-k2',
  labels: ['Person'],
  propertySchema: {
    Person: { name: 'string', bio: 'text', age: 'int:18-80' }
  },
});
const graph = await gen.generate();

Key Options

RAN DDD Context

Bounded Context: Data Infrastructure

References

• API reference: See references/commands.md
• Full README
• npm