1 // This file is part of Eigen, a lightweight C++ template library 2 // for linear algebra. 3 // 4 // Copyright (C) 2013 Christoph Hertzberg <chtz (at) informatik.uni-bremen.de> 5 // 6 // This Source Code Form is subject to the terms of the Mozilla 7 // Public License v. 2.0. If a copy of the MPL was not distributed 8 // with this file, You can obtain one at http://mozilla.org/MPL/2.0/. 9 10 #include "main.h" 11 #include <unsupported/Eigen/AutoDiff> 12 13 /* 14 * In this file scalar derivations are tested for correctness. 15 * TODO add more tests! 16 */ 17 18 template<typename Scalar> void check_atan2() 19 { 20 typedef Matrix<Scalar, 1, 1> Deriv1; 21 typedef AutoDiffScalar<Deriv1> AD; 22 23 AD x(internal::random<Scalar>(-3.0, 3.0), Deriv1::UnitX()); 24 25 using std::exp; 26 Scalar r = exp(internal::random<Scalar>(-10, 10)); 27 28 AD s = sin(x), c = cos(x); 29 AD res = atan2(r*s, r*c); 30 31 VERIFY_IS_APPROX(res.value(), x.value()); 32 VERIFY_IS_APPROX(res.derivatives(), x.derivatives()); 33 34 res = atan2(r*s+0, r*c+0); 35 VERIFY_IS_APPROX(res.value(), x.value()); 36 VERIFY_IS_APPROX(res.derivatives(), x.derivatives()); 37 } 38 39 template<typename Scalar> void check_hyperbolic_functions() 40 { 41 using std::sinh; 42 using std::cosh; 43 using std::tanh; 44 typedef Matrix<Scalar, 1, 1> Deriv1; 45 typedef AutoDiffScalar<Deriv1> AD; 46 Deriv1 p = Deriv1::Random(); 47 AD val(p.x(),Deriv1::UnitX()); 48 49 Scalar cosh_px = std::cosh(p.x()); 50 AD res1 = tanh(val); 51 VERIFY_IS_APPROX(res1.value(), std::tanh(p.x())); 52 VERIFY_IS_APPROX(res1.derivatives().x(), Scalar(1.0) / (cosh_px * cosh_px)); 53 54 AD res2 = sinh(val); 55 VERIFY_IS_APPROX(res2.value(), std::sinh(p.x())); 56 VERIFY_IS_APPROX(res2.derivatives().x(), cosh_px); 57 58 AD res3 = cosh(val); 59 VERIFY_IS_APPROX(res3.value(), cosh_px); 60 VERIFY_IS_APPROX(res3.derivatives().x(), std::sinh(p.x())); 61 62 // Check constant values. 63 const Scalar sample_point = Scalar(1) / Scalar(3); 64 val = AD(sample_point,Deriv1::UnitX()); 65 res1 = tanh(val); 66 VERIFY_IS_APPROX(res1.derivatives().x(), Scalar(0.896629559604914)); 67 68 res2 = sinh(val); 69 VERIFY_IS_APPROX(res2.derivatives().x(), Scalar(1.056071867829939)); 70 71 res3 = cosh(val); 72 VERIFY_IS_APPROX(res3.derivatives().x(), Scalar(0.339540557256150)); 73 } 74 75 template <typename Scalar> 76 void check_limits_specialization() 77 { 78 typedef Eigen::Matrix<Scalar, 1, 1> Deriv; 79 typedef Eigen::AutoDiffScalar<Deriv> AD; 80 81 typedef std::numeric_limits<AD> A; 82 typedef std::numeric_limits<Scalar> B; 83 84 #if EIGEN_HAS_CXX11 85 VERIFY(bool(std::is_base_of<B, A>::value)); 86 #endif 87 } 88 89 void test_autodiff_scalar() 90 { 91 for(int i = 0; i < g_repeat; i++) { 92 CALL_SUBTEST_1( check_atan2<float>() ); 93 CALL_SUBTEST_2( check_atan2<double>() ); 94 CALL_SUBTEST_3( check_hyperbolic_functions<float>() ); 95 CALL_SUBTEST_4( check_hyperbolic_functions<double>() ); 96 CALL_SUBTEST_5( check_limits_specialization<double>()); 97 } 98 } 99
