clustering-tfjs/dist/utils/smallest_eigenvectors_with_values.js

High-performance TypeScript clustering algorithms (K-Means, Spectral, Agglomerative) with TensorFlow.js acceleration and scikit-learn compatibility

"use strict";
var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
    if (k2 === undefined) k2 = k;
    var desc = Object.getOwnPropertyDescriptor(m, k);
    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
      desc = { enumerable: true, get: function() { return m[k]; } };
    }
    Object.defineProperty(o, k2, desc);
}) : (function(o, m, k, k2) {
    if (k2 === undefined) k2 = k;
    o[k2] = m[k];
}));
var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
    Object.defineProperty(o, "default", { enumerable: true, value: v });
}) : function(o, v) {
    o["default"] = v;
});
var __importStar = (this && this.__importStar) || (function () {
    var ownKeys = function(o) {
        ownKeys = Object.getOwnPropertyNames || function (o) {
            var ar = [];
            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
            return ar;
        };
        return ownKeys(o);
    };
    return function (mod) {
        if (mod && mod.__esModule) return mod;
        var result = {};
        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
        __setModuleDefault(result, mod);
        return result;
    };
})();
Object.defineProperty(exports, "__esModule", { value: true });
exports.smallest_eigenvectors_with_values = smallest_eigenvectors_with_values;
const tf = __importStar(require("../tf-adapter"));
const eigen_post_1 = require("./eigen_post");
/**
 * Returns the `k` smallest eigenvectors AND eigenvalues of the provided symmetric matrix.
 * This is needed to apply diffusion map scaling like sklearn does.
 */
function smallest_eigenvectors_with_values(matrix, k) {
    if (!Number.isInteger(k) || k < 1) {
        throw new Error('k must be a positive integer.');
    }
    return tf.tidy(() => {
        // Import improved solver for better accuracy
        // eslint-disable-next-line @typescript-eslint/no-var-requires
        const { improved_jacobi_eigen } = require('./eigen_improved');
        // 1) Full eigendecomposition with improved solver
        // For normalized Laplacians, we know it's PSD
        const { eigenvalues, eigenvectors } = improved_jacobi_eigen(matrix, {
            isPSD: true,
            maxIterations: 3000,
            tolerance: 1e-14,
        });
        // 2) Deterministic ordering & sign fixing
        const processed = (0, eigen_post_1.deterministic_eigenpair_processing)({
            eigenvalues,
            eigenvectors,
        });
        // 3) Determine number of numerically-zero eigenvalues
        const TOL = 1e-2;
        let c = 0;
        for (const v of processed.eigenvalues) {
            if (v <= TOL)
                c += 1;
            else
                break;
        }
        const n = processed.eigenvectors.length;
        const sliceCols = Math.min(k + c, n);
        // Extract selected eigenvectors
        const selectedVecs = Array.from({ length: n }, () => new Array(sliceCols));
        const selectedVals = new Array(sliceCols);
        for (let col = 0; col < sliceCols; col++) {
            selectedVals[col] = processed.eigenvalues[col];
            for (let row = 0; row < n; row++) {
                selectedVecs[row][col] = processed.eigenvectors[row][col];
            }
        }
        return {
            eigenvectors: tf.tensor2d(selectedVecs, [n, sliceCols], 'float32'),
            eigenvalues: tf.tensor1d(selectedVals, 'float32'),
        };
    });
}