/**
 * K-means clustering for semantic code grouping.
 * Groups code chunks by similarity to discover concept areas in the codebase.
 */
/**
 * Represents a discovered concept cluster
 */
export interface ConceptCluster {
    /** Unique cluster identifier (0-indexed) */
    id: number;
    /** Human-readable label derived from representative chunks */
    label: string;
    /** Number of code chunks in this cluster */
    size: number;
    /** Representative chunk IDs for this cluster */
    representativeChunks: string[];
    /** Centroid vector for this cluster */
    centroid: number[];
    /** Keywords extracted from representative chunks */
    keywords: string[];
}
/**
 * Result of clustering operation
 */
export interface ClusteringResult {
    /** Total number of clusters created */
    clusterCount: number;
    /** All discovered concept clusters */
    clusters: ConceptCluster[];
    /** Mapping from chunk ID to cluster ID */
    assignments: Map<string, number>;
}
/**
 * Options for clustering operation
 */
export interface ClusteringOptions {
    /** Target number of clusters (default: auto-determined based on chunk count) */
    numClusters?: number;
    /** Maximum iterations for k-means (default: 100) */
    maxIterations?: number;
    /** Convergence threshold (default: 0.001) */
    convergenceThreshold?: number;
    /** Number of representative chunks to store per cluster (default: 3) */
    numRepresentatives?: number;
}
/**
 * Chunk data needed for clustering
 */
export interface ChunkForClustering {
    id: string;
    content: string;
    filepath: string;
    embedding: number[];
    symbolName?: string;
    symbolType?: string;
}
/**
 * Calculate cosine similarity between two vectors
 */
export declare function cosineSimilarity(a: number[], b: number[]): number;
/**
 * Calculate Euclidean distance between two vectors
 */
export declare function euclideanDistance(a: number[], b: number[]): number;
/**
 * Perform k-means clustering on embeddings
 */
export declare function kMeansClustering(chunks: ChunkForClustering[], options?: ClusteringOptions): ClusteringResult;
/**
 * Find the cluster ID for a given embedding (nearest centroid)
 */
export declare function assignToCluster(embedding: number[], clusters: ConceptCluster[]): number;
/**
 * Calculate silhouette score for clustering quality assessment
 */
export declare function calculateSilhouetteScore(chunks: ChunkForClustering[], assignments: Map<string, number>, clusters: ConceptCluster[]): number;
//# sourceMappingURL=clustering.d.ts.map