def test_regularized_least_squares_err(self): x_train = TEST_DATA['rls']['x_train'] y_train = TEST_DATA['rls']['y_train'] M = TEST_DATA['rls']['M'] err = TEST_DATA['rls']['err'] regularization_lambda = TEST_DATA['rls']['lambda'] _, err_computed = regularized_least_squares(x_train, y_train, M, regularization_lambda) self.assertAlmostEqual(err, err_computed, 8)
def test_regularized_least_squares_w(self): x_train = TEST_DATA['rls']['x_train'] y_train = TEST_DATA['rls']['y_train'] M = TEST_DATA['rls']['M'] w = TEST_DATA['rls']['w'] regularization_lambda = TEST_DATA['rls']['lambda'] w_computed, _ = regularized_least_squares(x_train, y_train, M, regularization_lambda) max_diff = np.max(np.abs(w - w_computed)) self.assertAlmostEqual(max_diff, 0, 6)
def test_regularized_least_squares_w(self): x_train = TEST_DATA['rls']['x_train'] y_train = TEST_DATA['rls']['y_train'] M = TEST_DATA['rls']['M'] w_expected = TEST_DATA['rls']['w'] regularization_lambda = TEST_DATA['rls']['lambda'] w, _ = regularized_least_squares(x_train, y_train, M, regularization_lambda) self.assertEqual(np.shape(w), (5, 1)) np.testing.assert_almost_equal(w, w_expected)