1 // This file is part of Eigen, a lightweight C++ template library
2 // for linear algebra.
3 //
4 // Copyright (C) 2006-2008 Benoit Jacob <[email protected]>
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 "product.h"
11 #include <Eigen/LU>
12
13 template<typename T>
test_aliasing()14 void test_aliasing()
15 {
16 int rows = internal::random<int>(1,12);
17 int cols = internal::random<int>(1,12);
18 typedef Matrix<T,Dynamic,Dynamic> MatrixType;
19 typedef Matrix<T,Dynamic,1> VectorType;
20 VectorType x(cols); x.setRandom();
21 VectorType z(x);
22 VectorType y(rows); y.setZero();
23 MatrixType A(rows,cols); A.setRandom();
24 // CwiseBinaryOp
25 VERIFY_IS_APPROX(x = y + A*x, A*z); // OK because "y + A*x" is marked as "assume-aliasing"
26 x = z;
27 // CwiseUnaryOp
28 VERIFY_IS_APPROX(x = T(1.)*(A*x), A*z); // OK because 1*(A*x) is replaced by (1*A*x) which is a Product<> expression
29 x = z;
30 // VERIFY_IS_APPROX(x = y-A*x, -A*z); // Not OK in 3.3 because x is resized before A*x gets evaluated
31 x = z;
32 }
33
34 template<int>
product_large_regressions()35 void product_large_regressions()
36 {
37 {
38 // test a specific issue in DiagonalProduct
39 int N = 1000000;
40 VectorXf v = VectorXf::Ones(N);
41 MatrixXf m = MatrixXf::Ones(N,3);
42 m = (v+v).asDiagonal() * m;
43 VERIFY_IS_APPROX(m, MatrixXf::Constant(N,3,2));
44 }
45
46 {
47 // test deferred resizing in Matrix::operator=
48 MatrixXf a = MatrixXf::Random(10,4), b = MatrixXf::Random(4,10), c = a;
49 VERIFY_IS_APPROX((a = a * b), (c * b).eval());
50 }
51
52 {
53 // check the functions to setup blocking sizes compile and do not segfault
54 // FIXME check they do what they are supposed to do !!
55 std::ptrdiff_t l1 = internal::random<int>(10000,20000);
56 std::ptrdiff_t l2 = internal::random<int>(100000,200000);
57 std::ptrdiff_t l3 = internal::random<int>(1000000,2000000);
58 setCpuCacheSizes(l1,l2,l3);
59 VERIFY(l1==l1CacheSize());
60 VERIFY(l2==l2CacheSize());
61 std::ptrdiff_t k1 = internal::random<int>(10,100)*16;
62 std::ptrdiff_t m1 = internal::random<int>(10,100)*16;
63 std::ptrdiff_t n1 = internal::random<int>(10,100)*16;
64 // only makes sure it compiles fine
65 internal::computeProductBlockingSizes<float,float,std::ptrdiff_t>(k1,m1,n1,1);
66 }
67
68 {
69 // test regression in row-vector by matrix (bad Map type)
70 MatrixXf mat1(10,32); mat1.setRandom();
71 MatrixXf mat2(32,32); mat2.setRandom();
72 MatrixXf r1 = mat1.row(2)*mat2.transpose();
73 VERIFY_IS_APPROX(r1, (mat1.row(2)*mat2.transpose()).eval());
74
75 MatrixXf r2 = mat1.row(2)*mat2;
76 VERIFY_IS_APPROX(r2, (mat1.row(2)*mat2).eval());
77 }
78
79 {
80 Eigen::MatrixXd A(10,10), B, C;
81 A.setRandom();
82 C = A;
83 for(int k=0; k<79; ++k)
84 C = C * A;
85 B.noalias() = (((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A)) * ((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A)))
86 * (((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A)) * ((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A))*((A*A)*(A*A)));
87 VERIFY_IS_APPROX(B,C);
88 }
89 }
90
91 template<int>
bug_1622()92 void bug_1622() {
93 typedef Matrix<double, 2, -1, 0, 2, -1> Mat2X;
94 Mat2X x(2,2); x.setRandom();
95 MatrixXd y(2,2); y.setRandom();
96 const Mat2X K1 = x * y.inverse();
97 const Matrix2d K2 = x * y.inverse();
98 VERIFY_IS_APPROX(K1,K2);
99 }
100
EIGEN_DECLARE_TEST(product_large)101 EIGEN_DECLARE_TEST(product_large)
102 {
103 for(int i = 0; i < g_repeat; i++) {
104 CALL_SUBTEST_1( product(MatrixXf(internal::random<int>(1,EIGEN_TEST_MAX_SIZE), internal::random<int>(1,EIGEN_TEST_MAX_SIZE))) );
105 CALL_SUBTEST_2( product(MatrixXd(internal::random<int>(1,EIGEN_TEST_MAX_SIZE), internal::random<int>(1,EIGEN_TEST_MAX_SIZE))) );
106 CALL_SUBTEST_2( product(MatrixXd(internal::random<int>(1,10), internal::random<int>(1,10))) );
107
108 CALL_SUBTEST_3( product(MatrixXi(internal::random<int>(1,EIGEN_TEST_MAX_SIZE), internal::random<int>(1,EIGEN_TEST_MAX_SIZE))) );
109 CALL_SUBTEST_4( product(MatrixXcf(internal::random<int>(1,EIGEN_TEST_MAX_SIZE/2), internal::random<int>(1,EIGEN_TEST_MAX_SIZE/2))) );
110 CALL_SUBTEST_5( product(Matrix<float,Dynamic,Dynamic,RowMajor>(internal::random<int>(1,EIGEN_TEST_MAX_SIZE), internal::random<int>(1,EIGEN_TEST_MAX_SIZE))) );
111
112 CALL_SUBTEST_1( test_aliasing<float>() );
113
114 CALL_SUBTEST_6( bug_1622<1>() );
115
116 CALL_SUBTEST_7( product(MatrixXcd(internal::random<int>(1,EIGEN_TEST_MAX_SIZE/2), internal::random<int>(1,EIGEN_TEST_MAX_SIZE/2))) );
117 CALL_SUBTEST_8( product(Matrix<double,Dynamic,Dynamic,RowMajor>(internal::random<int>(1,EIGEN_TEST_MAX_SIZE), internal::random<int>(1,EIGEN_TEST_MAX_SIZE))) );
118 CALL_SUBTEST_9( product(Matrix<std::complex<float>,Dynamic,Dynamic,RowMajor>(internal::random<int>(1,EIGEN_TEST_MAX_SIZE), internal::random<int>(1,EIGEN_TEST_MAX_SIZE))) );
119 CALL_SUBTEST_10( product(Matrix<std::complex<double>,Dynamic,Dynamic,RowMajor>(internal::random<int>(1,EIGEN_TEST_MAX_SIZE), internal::random<int>(1,EIGEN_TEST_MAX_SIZE))) );
120 }
121
122 CALL_SUBTEST_6( product_large_regressions<0>() );
123
124 // Regression test for bug 714:
125 #if defined EIGEN_HAS_OPENMP
126 omp_set_dynamic(1);
127 for(int i = 0; i < g_repeat; i++) {
128 CALL_SUBTEST_6( product(Matrix<float,Dynamic,Dynamic>(internal::random<int>(1,EIGEN_TEST_MAX_SIZE), internal::random<int>(1,EIGEN_TEST_MAX_SIZE))) );
129 }
130 #endif
131 }
132