Struct KnowledgeGraph 

pub struct KnowledgeGraph { /* private fields */ }

Knowledge Graph Manager - main entry point for the library.

Implementations

impl KnowledgeGraph

pub fn open<P: AsRef<Path>>(path: P) -> Result<Self>

Open a new knowledge graph database connection.

pub fn open_in_memory() -> Result<Self>

Open an in-memory knowledge graph (useful for testing).

pub fn connection(&self) -> &Connection

Get a reference to the underlying SQLite connection.

pub fn transaction(&self) -> Result<Transaction<'_>>

Begin a transaction for batch operations.

pub fn insert_entity(&self, entity: &Entity) -> Result<i64>

Insert an entity into the knowledge graph.

pub fn get_entity(&self, id: i64) -> Result<Entity>

Get an entity by ID.

pub fn list_entities( &self, entity_type: Option<&str>, limit: Option<i64>, ) -> Result<Vec<Entity>>

List entities with optional filtering.

pub fn update_entity(&self, entity: &Entity) -> Result<()>

Update an entity.

pub fn delete_entity(&self, id: i64) -> Result<()>

Delete an entity.

pub fn insert_relation(&self, relation: &Relation) -> Result<i64>

Insert a relation between entities.

pub fn get_neighbors(&self, entity_id: i64, depth: u32) -> Result<Vec<Neighbor>>

Get neighbors of an entity using BFS traversal.

pub fn insert_vector(&self, entity_id: i64, vector: Vec<f32>) -> Result<()>

Insert a vector embedding for an entity.

pub fn search_vectors( &self, query: Vec<f32>, k: usize, ) -> Result<Vec<SearchResult>>

Search for similar entities using vector embeddings.

pub fn create_turboquant_index( &self, config: Option<TurboQuantConfig>, ) -> Result<TurboQuantIndex>

Create a TurboQuant index for fast approximate nearest neighbor search.

TurboQuant provides:

• Instant indexing (no training required)
• 6x memory compression
• Near-zero accuracy loss

Arguments

• config - Optional configuration (uses defaults if None)

Example

let config = TurboQuantConfig {
    dimension: 384,
    bit_width: 3,
    seed: 42,
};
let mut index = kg.create_turboquant_index(Some(config))?;

// Add vectors to index
for (entity_id, vector) in all_vectors {
    index.add_vector(entity_id, &vector)?;
}

// Fast search
let results = index.search(&query_vector, 10)?;

pub fn build_turboquant_index( &self, config: Option<TurboQuantConfig>, ) -> Result<TurboQuantIndex>

Build a TurboQuant index from all existing vectors in the database. This is a convenience method that loads all vectors and indexes them.

pub fn insert_hyperedge(&self, hyperedge: &Hyperedge) -> Result<i64>

Insert a hyperedge (higher-order relation) into the knowledge graph.

pub fn get_hyperedge(&self, id: i64) -> Result<Hyperedge>

Get a hyperedge by ID.

pub fn list_hyperedges( &self, hyperedge_type: Option<&str>, min_arity: Option<usize>, max_arity: Option<usize>, limit: Option<i64>, ) -> Result<Vec<Hyperedge>>

List hyperedges with optional filtering.

pub fn update_hyperedge(&self, hyperedge: &Hyperedge) -> Result<()>

Update a hyperedge.

pub fn delete_hyperedge(&self, id: i64) -> Result<()>

Delete a hyperedge by ID.

pub fn get_higher_order_neighbors( &self, entity_id: i64, min_arity: Option<usize>, max_arity: Option<usize>, ) -> Result<Vec<HigherOrderNeighbor>>

Get higher-order neighbors of an entity (connected through hyperedges).

pub fn get_entity_hyperedges(&self, entity_id: i64) -> Result<Vec<Hyperedge>>

Get all hyperedges that an entity participates in.

pub fn kg_higher_order_bfs( &self, start_id: i64, max_depth: u32, min_arity: Option<usize>, ) -> Result<Vec<TraversalNode>>

Higher-order BFS traversal through hyperedges.

pub fn kg_higher_order_shortest_path( &self, from_id: i64, to_id: i64, max_depth: u32, ) -> Result<Option<HigherOrderPath>>

Find shortest path between two entities through hyperedges.

pub fn kg_hyperedge_degree(&self, entity_id: i64) -> Result<f64>

Compute hyperedge degree centrality for an entity.

pub fn kg_hypergraph_entity_pagerank( &self, damping: Option<f64>, max_iter: Option<usize>, tolerance: Option<f64>, ) -> Result<HashMap<i64, f64>>

Compute entity-level hypergraph PageRank using Zhou formula.

pub fn kg_semantic_search( &self, query_embedding: Vec<f32>, k: usize, ) -> Result<Vec<SearchResultWithEntity>>

Semantic search using vector embeddings. Returns entities sorted by similarity score.

pub fn kg_get_context(&self, entity_id: i64, depth: u32) -> Result<GraphContext>

Get context around an entity using graph traversal. Returns neighbors up to the specified depth.

pub fn kg_hybrid_search( &self, _query_text: &str, query_embedding: Vec<f32>, k: usize, ) -> Result<Vec<HybridSearchResult>>

Hybrid search combining semantic search and graph context. Performs semantic search first, then retrieves context for top-k results.

pub fn kg_bfs_traversal( &self, start_id: i64, direction: Direction, max_depth: u32, ) -> Result<Vec<TraversalNode>>

BFS traversal from a starting entity. Returns all reachable entities within max_depth with depth information.

pub fn kg_dfs_traversal( &self, start_id: i64, direction: Direction, max_depth: u32, ) -> Result<Vec<TraversalNode>>

DFS traversal from a starting entity. Returns all reachable entities within max_depth.

pub fn kg_shortest_path( &self, from_id: i64, to_id: i64, max_depth: u32, ) -> Result<Option<TraversalPath>>

Find shortest path between two entities using BFS. Returns the path with all intermediate steps (if exists).

pub fn kg_graph_stats(&self) -> Result<GraphStats>

Compute graph statistics.

pub fn kg_pagerank( &self, config: Option<PageRankConfig>, ) -> Result<Vec<(i64, f64)>>

Compute PageRank scores for all entities. Returns a vector of (entity_id, score) sorted by score descending.

pub fn kg_louvain(&self) -> Result<CommunityResult>

Detect communities using Louvain algorithm. Returns community memberships and modularity score.

pub fn kg_connected_components(&self) -> Result<Vec<Vec<i64>>>

Find connected components in the graph. Returns a list of components, each being a list of entity IDs.

pub fn kg_analyze(&self) -> Result<GraphAnalysis>

Run full graph analysis (PageRank + Louvain + Connected Components).

pub fn export_json(&self) -> Result<D3ExportGraph>

Export the knowledge graph in D3.js JSON format.

Returns a D3ExportGraph containing nodes, links, and metadata, ready for use with D3.js force-directed graph visualizations.

Example

let graph = kg.export_json()?;
let json = serde_json::to_string_pretty(&graph)?;
std::fs::write("graph.json", json)?;

pub fn export_dot(&self, config: &DotConfig) -> Result<String>

Export the knowledge graph in DOT (Graphviz) format.

Returns a DOT format string that can be rendered with Graphviz tools (dot, neato, fdp, etc.).

Example

let dot = kg.export_dot(&DotConfig::default())?;
std::fs::write("graph.dot", dot)?;
// Then: dot -Tpng graph.dot -o graph.png

Trait Implementations

impl Debug for KnowledgeGraph

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.