1 // Ceres Solver - A fast non-linear least squares minimizer
2 // Copyright 2010, 2011, 2012 Google Inc. All rights reserved.
3 // http://code.google.com/p/ceres-solver/
4 //
5 // Redistribution and use in source and binary forms, with or without
6 // modification, are permitted provided that the following conditions are met:
7 //
8 // * Redistributions of source code must retain the above copyright notice,
9 // this list of conditions and the following disclaimer.
10 // * Redistributions in binary form must reproduce the above copyright notice,
11 // this list of conditions and the following disclaimer in the documentation
12 // and/or other materials provided with the distribution.
13 // * Neither the name of Google Inc. nor the names of its contributors may be
14 // used to endorse or promote products derived from this software without
15 // specific prior written permission.
16 //
17 // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
18 // AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 // IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 // ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
21 // LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
22 // CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
23 // SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
24 // INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
25 // CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
26 // ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
27 // POSSIBILITY OF SUCH DAMAGE.
28 //
29 // Author: sameeragarwal@google.com (Sameer Agarwal)
30
31 #include "ceres/autodiff_cost_function.h"
32
33 #include <cstddef>
34
35 #include "gtest/gtest.h"
36 #include "ceres/cost_function.h"
37
38 namespace ceres {
39 namespace internal {
40
41 class BinaryScalarCost {
42 public:
BinaryScalarCost(double a)43 explicit BinaryScalarCost(double a): a_(a) {}
44 template <typename T>
operator ()(const T * const x,const T * const y,T * cost) const45 bool operator()(const T* const x, const T* const y,
46 T* cost) const {
47 cost[0] = x[0] * y[0] + x[1] * y[1] - T(a_);
48 return true;
49 }
50 private:
51 double a_;
52 };
53
TEST(AutodiffCostFunction,BilinearDifferentiationTest)54 TEST(AutodiffCostFunction, BilinearDifferentiationTest) {
55 CostFunction* cost_function =
56 new AutoDiffCostFunction<BinaryScalarCost, 1, 2, 2>(
57 new BinaryScalarCost(1.0));
58
59 double** parameters = new double*[2];
60 parameters[0] = new double[2];
61 parameters[1] = new double[2];
62
63 parameters[0][0] = 1;
64 parameters[0][1] = 2;
65
66 parameters[1][0] = 3;
67 parameters[1][1] = 4;
68
69 double** jacobians = new double*[2];
70 jacobians[0] = new double[2];
71 jacobians[1] = new double[2];
72
73 double residuals = 0.0;
74
75 cost_function->Evaluate(parameters, &residuals, NULL);
76 EXPECT_EQ(10.0, residuals);
77 cost_function->Evaluate(parameters, &residuals, jacobians);
78
79 EXPECT_EQ(3, jacobians[0][0]);
80 EXPECT_EQ(4, jacobians[0][1]);
81 EXPECT_EQ(1, jacobians[1][0]);
82 EXPECT_EQ(2, jacobians[1][1]);
83
84 delete[] jacobians[0];
85 delete[] jacobians[1];
86 delete[] parameters[0];
87 delete[] parameters[1];
88 delete[] jacobians;
89 delete[] parameters;
90 delete cost_function;
91 }
92
93 struct TenParameterCost {
94 template <typename T>
operator ()ceres::internal::TenParameterCost95 bool operator()(const T* const x0,
96 const T* const x1,
97 const T* const x2,
98 const T* const x3,
99 const T* const x4,
100 const T* const x5,
101 const T* const x6,
102 const T* const x7,
103 const T* const x8,
104 const T* const x9,
105 T* cost) const {
106 cost[0] = *x0 + *x1 + *x2 + *x3 + *x4 + *x5 + *x6 + *x7 + *x8 + *x9;
107 return true;
108 }
109 };
110
TEST(AutodiffCostFunction,ManyParameterAutodiffInstantiates)111 TEST(AutodiffCostFunction, ManyParameterAutodiffInstantiates) {
112 CostFunction* cost_function =
113 new AutoDiffCostFunction<
114 TenParameterCost, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1>(
115 new TenParameterCost);
116
117 double** parameters = new double*[10];
118 double** jacobians = new double*[10];
119 for (int i = 0; i < 10; ++i) {
120 parameters[i] = new double[1];
121 parameters[i][0] = i;
122 jacobians[i] = new double[1];
123 }
124
125 double residuals = 0.0;
126
127 cost_function->Evaluate(parameters, &residuals, NULL);
128 EXPECT_EQ(45.0, residuals);
129
130 cost_function->Evaluate(parameters, &residuals, jacobians);
131 EXPECT_EQ(residuals, 45.0);
132 for (int i = 0; i < 10; ++i) {
133 EXPECT_EQ(1.0, jacobians[i][0]);
134 }
135
136 for (int i = 0; i < 10; ++i) {
137 delete[] jacobians[i];
138 delete[] parameters[i];
139 }
140 delete[] jacobians;
141 delete[] parameters;
142 delete cost_function;
143 }
144
145 } // namespace internal
146 } // namespace ceres
147