1// This file is part of Eigen, a lightweight C++ template library
2// for linear algebra.
3//
4// Copyright (C) 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
12template<typename MatrixType> void matrixRedux(const MatrixType& m)
13{
14  typedef typename MatrixType::Index Index;
15  typedef typename MatrixType::Scalar Scalar;
16  typedef typename MatrixType::RealScalar RealScalar;
17
18  Index rows = m.rows();
19  Index cols = m.cols();
20
21  MatrixType m1 = MatrixType::Random(rows, cols);
22
23  // The entries of m1 are uniformly distributed in [0,1], so m1.prod() is very small. This may lead to test
24  // failures if we underflow into denormals. Thus, we scale so that entires are close to 1.
25  MatrixType m1_for_prod = MatrixType::Ones(rows, cols) + Scalar(0.2) * m1;
26
27  VERIFY_IS_MUCH_SMALLER_THAN(MatrixType::Zero(rows, cols).sum(), Scalar(1));
28  VERIFY_IS_APPROX(MatrixType::Ones(rows, cols).sum(), Scalar(float(rows*cols))); // the float() here to shut up excessive MSVC warning about int->complex conversion being lossy
29  Scalar s(0), p(1), minc(internal::real(m1.coeff(0))), maxc(internal::real(m1.coeff(0)));
30  for(int j = 0; j < cols; j++)
31  for(int i = 0; i < rows; i++)
32  {
33    s += m1(i,j);
34    p *= m1_for_prod(i,j);
35    minc = (std::min)(internal::real(minc), internal::real(m1(i,j)));
36    maxc = (std::max)(internal::real(maxc), internal::real(m1(i,j)));
37  }
38  const Scalar mean = s/Scalar(RealScalar(rows*cols));
39
40  VERIFY_IS_APPROX(m1.sum(), s);
41  VERIFY_IS_APPROX(m1.mean(), mean);
42  VERIFY_IS_APPROX(m1_for_prod.prod(), p);
43  VERIFY_IS_APPROX(m1.real().minCoeff(), internal::real(minc));
44  VERIFY_IS_APPROX(m1.real().maxCoeff(), internal::real(maxc));
45
46  // test slice vectorization assuming assign is ok
47  Index r0 = internal::random<Index>(0,rows-1);
48  Index c0 = internal::random<Index>(0,cols-1);
49  Index r1 = internal::random<Index>(r0+1,rows)-r0;
50  Index c1 = internal::random<Index>(c0+1,cols)-c0;
51  VERIFY_IS_APPROX(m1.block(r0,c0,r1,c1).sum(), m1.block(r0,c0,r1,c1).eval().sum());
52  VERIFY_IS_APPROX(m1.block(r0,c0,r1,c1).mean(), m1.block(r0,c0,r1,c1).eval().mean());
53  VERIFY_IS_APPROX(m1_for_prod.block(r0,c0,r1,c1).prod(), m1_for_prod.block(r0,c0,r1,c1).eval().prod());
54  VERIFY_IS_APPROX(m1.block(r0,c0,r1,c1).real().minCoeff(), m1.block(r0,c0,r1,c1).real().eval().minCoeff());
55  VERIFY_IS_APPROX(m1.block(r0,c0,r1,c1).real().maxCoeff(), m1.block(r0,c0,r1,c1).real().eval().maxCoeff());
56
57  // test empty objects
58  VERIFY_IS_APPROX(m1.block(r0,c0,0,0).sum(),   Scalar(0));
59  VERIFY_IS_APPROX(m1.block(r0,c0,0,0).prod(),  Scalar(1));
60}
61
62template<typename VectorType> void vectorRedux(const VectorType& w)
63{
64  typedef typename VectorType::Index Index;
65  typedef typename VectorType::Scalar Scalar;
66  typedef typename NumTraits<Scalar>::Real RealScalar;
67  Index size = w.size();
68
69  VectorType v = VectorType::Random(size);
70  VectorType v_for_prod = VectorType::Ones(size) + Scalar(0.2) * v; // see comment above declaration of m1_for_prod
71
72  for(int i = 1; i < size; i++)
73  {
74    Scalar s(0), p(1);
75    RealScalar minc(internal::real(v.coeff(0))), maxc(internal::real(v.coeff(0)));
76    for(int j = 0; j < i; j++)
77    {
78      s += v[j];
79      p *= v_for_prod[j];
80      minc = (std::min)(minc, internal::real(v[j]));
81      maxc = (std::max)(maxc, internal::real(v[j]));
82    }
83    VERIFY_IS_MUCH_SMALLER_THAN(internal::abs(s - v.head(i).sum()), Scalar(1));
84    VERIFY_IS_APPROX(p, v_for_prod.head(i).prod());
85    VERIFY_IS_APPROX(minc, v.real().head(i).minCoeff());
86    VERIFY_IS_APPROX(maxc, v.real().head(i).maxCoeff());
87  }
88
89  for(int i = 0; i < size-1; i++)
90  {
91    Scalar s(0), p(1);
92    RealScalar minc(internal::real(v.coeff(i))), maxc(internal::real(v.coeff(i)));
93    for(int j = i; j < size; j++)
94    {
95      s += v[j];
96      p *= v_for_prod[j];
97      minc = (std::min)(minc, internal::real(v[j]));
98      maxc = (std::max)(maxc, internal::real(v[j]));
99    }
100    VERIFY_IS_MUCH_SMALLER_THAN(internal::abs(s - v.tail(size-i).sum()), Scalar(1));
101    VERIFY_IS_APPROX(p, v_for_prod.tail(size-i).prod());
102    VERIFY_IS_APPROX(minc, v.real().tail(size-i).minCoeff());
103    VERIFY_IS_APPROX(maxc, v.real().tail(size-i).maxCoeff());
104  }
105
106  for(int i = 0; i < size/2; i++)
107  {
108    Scalar s(0), p(1);
109    RealScalar minc(internal::real(v.coeff(i))), maxc(internal::real(v.coeff(i)));
110    for(int j = i; j < size-i; j++)
111    {
112      s += v[j];
113      p *= v_for_prod[j];
114      minc = (std::min)(minc, internal::real(v[j]));
115      maxc = (std::max)(maxc, internal::real(v[j]));
116    }
117    VERIFY_IS_MUCH_SMALLER_THAN(internal::abs(s - v.segment(i, size-2*i).sum()), Scalar(1));
118    VERIFY_IS_APPROX(p, v_for_prod.segment(i, size-2*i).prod());
119    VERIFY_IS_APPROX(minc, v.real().segment(i, size-2*i).minCoeff());
120    VERIFY_IS_APPROX(maxc, v.real().segment(i, size-2*i).maxCoeff());
121  }
122
123  // test empty objects
124  VERIFY_IS_APPROX(v.head(0).sum(),   Scalar(0));
125  VERIFY_IS_APPROX(v.tail(0).prod(),  Scalar(1));
126  VERIFY_RAISES_ASSERT(v.head(0).mean());
127  VERIFY_RAISES_ASSERT(v.head(0).minCoeff());
128  VERIFY_RAISES_ASSERT(v.head(0).maxCoeff());
129}
130
131void test_redux()
132{
133  // the max size cannot be too large, otherwise reduxion operations obviously generate large errors.
134  int maxsize = (std::min)(100,EIGEN_TEST_MAX_SIZE);
135  EIGEN_UNUSED_VARIABLE(maxsize);
136  for(int i = 0; i < g_repeat; i++) {
137    CALL_SUBTEST_1( matrixRedux(Matrix<float, 1, 1>()) );
138    CALL_SUBTEST_1( matrixRedux(Array<float, 1, 1>()) );
139    CALL_SUBTEST_2( matrixRedux(Matrix2f()) );
140    CALL_SUBTEST_2( matrixRedux(Array2f()) );
141    CALL_SUBTEST_3( matrixRedux(Matrix4d()) );
142    CALL_SUBTEST_3( matrixRedux(Array4d()) );
143    CALL_SUBTEST_4( matrixRedux(MatrixXcf(internal::random<int>(1,maxsize), internal::random<int>(1,maxsize))) );
144    CALL_SUBTEST_4( matrixRedux(ArrayXXcf(internal::random<int>(1,maxsize), internal::random<int>(1,maxsize))) );
145    CALL_SUBTEST_5( matrixRedux(MatrixXd (internal::random<int>(1,maxsize), internal::random<int>(1,maxsize))) );
146    CALL_SUBTEST_5( matrixRedux(ArrayXXd (internal::random<int>(1,maxsize), internal::random<int>(1,maxsize))) );
147    CALL_SUBTEST_6( matrixRedux(MatrixXi (internal::random<int>(1,maxsize), internal::random<int>(1,maxsize))) );
148    CALL_SUBTEST_6( matrixRedux(ArrayXXi (internal::random<int>(1,maxsize), internal::random<int>(1,maxsize))) );
149  }
150  for(int i = 0; i < g_repeat; i++) {
151    CALL_SUBTEST_7( vectorRedux(Vector4f()) );
152    CALL_SUBTEST_7( vectorRedux(Array4f()) );
153    CALL_SUBTEST_5( vectorRedux(VectorXd(internal::random<int>(1,maxsize))) );
154    CALL_SUBTEST_5( vectorRedux(ArrayXd(internal::random<int>(1,maxsize))) );
155    CALL_SUBTEST_8( vectorRedux(VectorXf(internal::random<int>(1,maxsize))) );
156    CALL_SUBTEST_8( vectorRedux(ArrayXf(internal::random<int>(1,maxsize))) );
157  }
158}
159