/****************************************************************************** * Compilation: javac Cholesky.java * Execution: java Cholesky * * Compute Cholesky decomposition of symmetric positive definite * matrix A = LL^T. * * % java Cholesky * 2.00000 0.00000 0.00000 * 0.50000 2.17945 0.00000 * 0.50000 1.26179 3.62738 * ******************************************************************************/ public class Cholesky { private static final double EPSILON = 1e-10; // is symmetric public static boolean isSymmetric(double[][] A) { int N = A.length; for (int i = 0; i < N; i++) { for (int j = 0; j < i; j++) { if (A[i][j] != A[j][i]) return false; } } return true; } // is symmetric public static boolean isSquare(double[][] A) { int N = A.length; for (int i = 0; i < N; i++) { if (A[i].length != N) return false; } return true; } // return Cholesky factor L of psd matrix A = L L^T public static double[][] cholesky(double[][] A) { if (!isSquare(A)) { throw new RuntimeException("Matrix is not square"); } if (!isSymmetric(A)) { throw new RuntimeException("Matrix is not symmetric"); } int N = A.length; double[][] L = new double[N][N]; for (int i = 0; i < N; i++) { for (int j = 0; j <= i; j++) { double sum = 0.0; for (int k = 0; k < j; k++) { sum += L[i][k] * L[j][k]; } if (i == j) L[i][i] = Math.sqrt(A[i][i] - sum); else L[i][j] = 1.0 / L[j][j] * (A[i][j] - sum); } if (L[i][i] <= 0) { throw new RuntimeException("Matrix not positive definite"); } } return L; } // sample client public static void main(String[] args) { int N = 3; double[][] A = { { 4, 1, 1 }, { 1, 5, 3 }, { 1, 3, 15 } }; double[][] L = cholesky(A); for (int i = 0; i < N; i++) { for (int j = 0; j < N; j++) { StdOut.printf("%8.5f ", L[i][j]); } StdOut.println(); } } }