clusterkw/dist/algorithms/simple-clustering.js

A package for clustering keywords using OpenAI embeddings

"use strict";
Object.defineProperty(exports, "__esModule", { value: true });
exports.simpleClustering = simpleClustering;
/**
 * Simple clustering algorithm based on distance threshold
 * @param keywords Array of keywords to cluster
 * @param distances Matrix of distances between keywords
 * @param options Configuration options
 * @returns Array of clusters
 */
function simpleClustering(keywords, distances, options) {
    const { minClusterSize, distanceThreshold } = options;
    const n = keywords.length;
    const visited = new Set();
    const clusters = [];
    for (let i = 0; i < n; i++) {
        if (visited.has(i))
            continue;
        const cluster = [keywords[i]];
        visited.add(i);
        for (let j = 0; j < n; j++) {
            if (i === j || visited.has(j))
                continue;
            if (distances[i][j] <= distanceThreshold) {
                cluster.push(keywords[j]);
                visited.add(j);
            }
        }
        if (cluster.length >= minClusterSize) {
            clusters.push({ items: cluster });
        }
    }
    // Handle unclustered items
    const unclustered = [];
    for (let i = 0; i < n; i++) {
        if (!visited.has(i)) {
            unclustered.push(keywords[i]);
        }
    }
    if (unclustered.length > 0 && unclustered.length >= minClusterSize) {
        clusters.push({ items: unclustered });
    }
    return clusters;
}