/**
 * Dynamic Time Warping (legacy signature).
 *
 * @param {Array} X - First sequence (2D: features x time)
 * @param {Array} Y - Second sequence (2D: features x time)
 * @param {string} metric - 'euclidean' | 'cosine' | 'manhattan' | 'sqeuclidean'
 * @param {Array} step_sizes_sigma - custom steps (appended to the defaults)
 * @param {Array} weights_add - additive step weights (honored)
 * @param {Array} weights_mul - multiplicative step weights (honored)
 * @param {boolean} subseq - subsequence matching (honored)
 * @param {boolean} backtrack - whether to return the optimal path
 * @param {boolean} global_constraints - Sakoe-Chiba band
 * @param {number} band_rad - band radius as a fraction of min(N, M)
 * @returns {Object} { distance, cost_matrix, path, normalized_distance }
 *   `cost_matrix` is the (N, M) accumulated cost matrix (no longer the padded
 *   (N+1, M+1) legacy matrix). `path` is start-to-end ascending, as before.
 */
export function dtw(X: any[], Y: any[], metric?: string, step_sizes_sigma?: any[], weights_add?: any[], weights_mul?: any[], subseq?: boolean, backtrack?: boolean, global_constraints?: boolean, band_rad?: number): any;
/**
 * Compute cost matrix between two sequences
 * @param {Array} X - First sequence
 * @param {Array} Y - Second sequence
 * @param {string} metric - Distance metric
 * @returns {Array} Cost matrix
 */
export function computeCostMatrix(X: any[], Y: any[], metric?: string): any[];
/**
 * Distance metrics
 */
export function euclideanDistance(a: any, b: any): number;
export function manhattanDistance(a: any, b: any): number;
export function cosineSimilarity(a: any, b: any): number;
/**
 * Compute DTW distance matrix for multiple sequences
 * @param {Array} sequences - Array of sequences
 * @param {string} metric - Distance metric
 * @returns {Array} Distance matrix
 */
export function dtwDistanceMatrix(sequences: any[], metric?: string): any[];
/**
 * Simple k-means clustering using DTW distances
 * @param {Array} sequences - Array of sequences to cluster
 * @param {number} k - Number of clusters
 * @param {number} maxIterations - Maximum iterations
 * @returns {Array} Cluster assignments
 */
export function dtwKMeans(sequences: any[], k?: number, maxIterations?: number): any[];