def gs_iter(x_0, epsilon, max_iter): # 2a A = np.matrix([[1.0, 1.0/2, 1.0/3], [1.0/2, 1.0, 1.0/4], [1.0/3, 1.0/4, 1.0]]) B = np.array([.1, .1, .1]).reshape(3, 1) S_inv = np.zeros((3, 3)) a, b, c, d, e, f = A[0, 0], A[1, 0], A[1, 1], A[2, 0], A[2, 1], A[2, 2] S_inv[0, 0] = 1.0/a S_inv[1, 0] = -b/(a*c) S_inv[1, 1] = 1.0/c S_inv[2, 0] = (-c*d+b*e)/(a*c*f) S_inv[2, 1] = -e/(c*f) S_inv[2, 2] = 1.0/f S = np.tril(A) U = A - S iteration = 0 error = 1000 x_n = x_0 while iteration < max_iter: x_1 = pascal.mult(pascal.mult(S_inv, -U), x_n) x_1 += pascal.mult(S_inv, B) error = pascal.norm_inf(x_1 - x_n) x_n = x_1 if epsilon > error: return x_n, iteration iteration += 1 else: return None, None
def jacobi_iter(x_0, epsilon, max_iter): # 2a A = np.matrix([[1.0, 1.0/2, 1.0/3], [1.0/2, 1.0, 1.0/4], [1.0/3, 1.0/4, 1.0]]) b = np.array([.1, .1, .1]).reshape(3, 1) iteration = 0 error = 1000 x_n = x_0 while iteration < max_iter: d = np.diag(1 / np.diag(A)) alu = A - np.diag(np.diag(A)) x_1 = pascal.mult(d, b) + pascal.mult(pascal.mult(d, -alu), x_n) error = pascal.norm_inf(x_1 - x_n) x_n = x_1 if epsilon > error: return x_n, iteration iteration += 1 else: return None, None