• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 // This file is part of Eigen, a lightweight C++ template library
2 // for linear algebra.
3 //
4 // Copyright (C) 2006-2008 Benoit Jacob <jacob.benoit.1@gmail.com>
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 <Eigen/QR>
12 
13 template<typename Derived1, typename Derived2>
14 bool areNotApprox(const MatrixBase<Derived1>& m1, const MatrixBase<Derived2>& m2, typename Derived1::RealScalar epsilon = NumTraits<typename Derived1::RealScalar>::dummy_precision())
15 {
16   return !((m1-m2).cwiseAbs2().maxCoeff() < epsilon * epsilon
17                           * (std::max)(m1.cwiseAbs2().maxCoeff(), m2.cwiseAbs2().maxCoeff()));
18 }
19 
product(const MatrixType & m)20 template<typename MatrixType> void product(const MatrixType& m)
21 {
22   /* this test covers the following files:
23      Identity.h Product.h
24   */
25   typedef typename MatrixType::Index Index;
26   typedef typename MatrixType::Scalar Scalar;
27   typedef typename NumTraits<Scalar>::NonInteger NonInteger;
28   typedef Matrix<Scalar, MatrixType::RowsAtCompileTime, 1> RowVectorType;
29   typedef Matrix<Scalar, MatrixType::ColsAtCompileTime, 1> ColVectorType;
30   typedef Matrix<Scalar, MatrixType::RowsAtCompileTime, MatrixType::RowsAtCompileTime> RowSquareMatrixType;
31   typedef Matrix<Scalar, MatrixType::ColsAtCompileTime, MatrixType::ColsAtCompileTime> ColSquareMatrixType;
32   typedef Matrix<Scalar, MatrixType::RowsAtCompileTime, MatrixType::ColsAtCompileTime,
33                          MatrixType::Flags&RowMajorBit?ColMajor:RowMajor> OtherMajorMatrixType;
34 
35   Index rows = m.rows();
36   Index cols = m.cols();
37 
38   // this test relies a lot on Random.h, and there's not much more that we can do
39   // to test it, hence I consider that we will have tested Random.h
40   MatrixType m1 = MatrixType::Random(rows, cols),
41              m2 = MatrixType::Random(rows, cols),
42              m3(rows, cols);
43   RowSquareMatrixType
44              identity = RowSquareMatrixType::Identity(rows, rows),
45              square = RowSquareMatrixType::Random(rows, rows),
46              res = RowSquareMatrixType::Random(rows, rows);
47   ColSquareMatrixType
48              square2 = ColSquareMatrixType::Random(cols, cols),
49              res2 = ColSquareMatrixType::Random(cols, cols);
50   RowVectorType v1 = RowVectorType::Random(rows);
51   ColVectorType vc2 = ColVectorType::Random(cols), vcres(cols);
52   OtherMajorMatrixType tm1 = m1;
53 
54   Scalar s1 = internal::random<Scalar>();
55 
56   Index r  = internal::random<Index>(0, rows-1),
57         c  = internal::random<Index>(0, cols-1),
58         c2 = internal::random<Index>(0, cols-1);
59 
60   // begin testing Product.h: only associativity for now
61   // (we use Transpose.h but this doesn't count as a test for it)
62   VERIFY_IS_APPROX((m1*m1.transpose())*m2,  m1*(m1.transpose()*m2));
63   m3 = m1;
64   m3 *= m1.transpose() * m2;
65   VERIFY_IS_APPROX(m3,                      m1 * (m1.transpose()*m2));
66   VERIFY_IS_APPROX(m3,                      m1 * (m1.transpose()*m2));
67 
68   // continue testing Product.h: distributivity
69   VERIFY_IS_APPROX(square*(m1 + m2),        square*m1+square*m2);
70   VERIFY_IS_APPROX(square*(m1 - m2),        square*m1-square*m2);
71 
72   // continue testing Product.h: compatibility with ScalarMultiple.h
73   VERIFY_IS_APPROX(s1*(square*m1),          (s1*square)*m1);
74   VERIFY_IS_APPROX(s1*(square*m1),          square*(m1*s1));
75 
76   // test Product.h together with Identity.h
77   VERIFY_IS_APPROX(v1,                      identity*v1);
78   VERIFY_IS_APPROX(v1.transpose(),          v1.transpose() * identity);
79   // again, test operator() to check const-qualification
80   VERIFY_IS_APPROX(MatrixType::Identity(rows, cols)(r,c), static_cast<Scalar>(r==c));
81 
82   if (rows!=cols)
83      VERIFY_RAISES_ASSERT(m3 = m1*m1);
84 
85   // test the previous tests were not screwed up because operator* returns 0
86   // (we use the more accurate default epsilon)
87   if (!NumTraits<Scalar>::IsInteger && (std::min)(rows,cols)>1)
88   {
89     VERIFY(areNotApprox(m1.transpose()*m2,m2.transpose()*m1));
90   }
91 
92   // test optimized operator+= path
93   res = square;
94   res.noalias() += m1 * m2.transpose();
95   VERIFY_IS_APPROX(res, square + m1 * m2.transpose());
96   if (!NumTraits<Scalar>::IsInteger && (std::min)(rows,cols)>1)
97   {
98     VERIFY(areNotApprox(res,square + m2 * m1.transpose()));
99   }
100   vcres = vc2;
101   vcres.noalias() += m1.transpose() * v1;
102   VERIFY_IS_APPROX(vcres, vc2 + m1.transpose() * v1);
103 
104   // test optimized operator-= path
105   res = square;
106   res.noalias() -= m1 * m2.transpose();
107   VERIFY_IS_APPROX(res, square - (m1 * m2.transpose()));
108   if (!NumTraits<Scalar>::IsInteger && (std::min)(rows,cols)>1)
109   {
110     VERIFY(areNotApprox(res,square - m2 * m1.transpose()));
111   }
112   vcres = vc2;
113   vcres.noalias() -= m1.transpose() * v1;
114   VERIFY_IS_APPROX(vcres, vc2 - m1.transpose() * v1);
115 
116   tm1 = m1;
117   VERIFY_IS_APPROX(tm1.transpose() * v1, m1.transpose() * v1);
118   VERIFY_IS_APPROX(v1.transpose() * tm1, v1.transpose() * m1);
119 
120   // test submatrix and matrix/vector product
121   for (int i=0; i<rows; ++i)
122     res.row(i) = m1.row(i) * m2.transpose();
123   VERIFY_IS_APPROX(res, m1 * m2.transpose());
124   // the other way round:
125   for (int i=0; i<rows; ++i)
126     res.col(i) = m1 * m2.transpose().col(i);
127   VERIFY_IS_APPROX(res, m1 * m2.transpose());
128 
129   res2 = square2;
130   res2.noalias() += m1.transpose() * m2;
131   VERIFY_IS_APPROX(res2, square2 + m1.transpose() * m2);
132   if (!NumTraits<Scalar>::IsInteger && (std::min)(rows,cols)>1)
133   {
134     VERIFY(areNotApprox(res2,square2 + m2.transpose() * m1));
135   }
136 
137   VERIFY_IS_APPROX(res.col(r).noalias() = square.adjoint() * square.col(r), (square.adjoint() * square.col(r)).eval());
138   VERIFY_IS_APPROX(res.col(r).noalias() = square * square.col(r), (square * square.col(r)).eval());
139 
140   // inner product
141   Scalar x = square2.row(c) * square2.col(c2);
142   VERIFY_IS_APPROX(x, square2.row(c).transpose().cwiseProduct(square2.col(c2)).sum());
143 }
144