Skip to content

graph_queries.md

Latest commit

 

History

History
114 lines (79 loc) · 2.2 KB

graph_queries.md

File metadata and controls

114 lines (79 loc) · 2.2 KB

SQLY Graph Queries

📖 Introduction

Graph queries in SQLY enable working with networked data, such as social connections, recommendation systems, and hierarchical relationships.

🔗 Basic Graph Querying

SQLY allows traversing relationships between entities using graph_traverse.

✅ Example 1: Find Direct Connections Between Users

query:
  select: [user_a, user_b]
  from: user_connections
  where:
    graph_traverse:
      start: 42
      depth: 1

This retrieves all users directly connected to user 42.

🔄 Multi-Hop Traversals

Graph traversal supports multi-hop queries to find indirect relationships.

✅ Example 2: Find Friends of Friends

query:
  select: [user_a, user_b]
  from: user_connections
  where:
    graph_traverse:
      start: 42
      depth: 2

This finds all friends and their friends (up to two levels away).

🔍 Shortest Path Queries

SQLY can compute the shortest path between nodes in a graph.

✅ Example 3: Find the Shortest Path Between Two Users

query:
  select: [user_a, user_b, path_length]
  from: user_connections
  where:
    shortest_path:
      start: 42
      end: 99

This retrieves the shortest connection path between users 42 and 99.

🎯 Weighted Graph Queries

Some relationships may have weights, such as distances or costs.

✅ Example 4: Find the Cheapest Delivery Route

query:
  select: [source, destination, total_cost]
  from: delivery_routes
  where:
    shortest_path:
      start: "Warehouse A"
      end: "Customer B"
      weight: shipping_cost

This retrieves the most cost-effective delivery route.

🏛️ Community Detection

Graph analytics can group entities based on relationships.

✅ Example 5: Detect Social Communities

query:
  select: [user_id, community_id]
  from: users
  where:
    community_detection:
      method: louvain

This clusters users into social communities using the Louvain method.

📌 Summary

  • graph_traverse finds direct and multi-hop connections.
  • shortest_path computes optimal routes between nodes.
  • Weighted graphs allow cost-based queries.
  • Community detection groups entities into clusters.