Kolja Brix | 58e086b | 2021-08-23 16:00:05 +0000 | [diff] [blame] | 1 | // This file is part of Eigen, a lightweight C++ template library |
| 2 | // for linear algebra. |
| 3 | // |
| 4 | // Copyright (C) 2021 Kolja Brix <kolja.brix@rwth-aachen.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 <Eigen/SVD> |
| 12 | |
| 13 | |
| 14 | template<typename MatrixType> |
| 15 | void check_generateRandomUnitaryMatrix(const Index dim) |
| 16 | { |
| 17 | const MatrixType Q = generateRandomUnitaryMatrix<MatrixType>(dim); |
| 18 | |
| 19 | // validate dimensions |
| 20 | VERIFY_IS_EQUAL(Q.rows(), dim); |
| 21 | VERIFY_IS_EQUAL(Q.cols(), dim); |
| 22 | |
| 23 | VERIFY_IS_UNITARY(Q); |
| 24 | } |
| 25 | |
| 26 | template<typename VectorType, typename RealScalarType> |
| 27 | void check_setupRandomSvs(const Index dim, const RealScalarType max) |
| 28 | { |
| 29 | const VectorType v = setupRandomSvs<VectorType, RealScalarType>(dim, max); |
| 30 | |
| 31 | // validate dimensions |
| 32 | VERIFY_IS_EQUAL(v.size(), dim); |
| 33 | |
| 34 | // check entries |
| 35 | for(Index i = 0; i < v.size(); ++i) |
| 36 | VERIFY_GE(v(i), 0); |
| 37 | for(Index i = 0; i < v.size()-1; ++i) |
| 38 | VERIFY_GE(v(i), v(i+1)); |
| 39 | } |
| 40 | |
| 41 | template<typename VectorType, typename RealScalarType> |
| 42 | void check_setupRangeSvs(const Index dim, const RealScalarType min, const RealScalarType max) |
| 43 | { |
| 44 | const VectorType v = setupRangeSvs<VectorType, RealScalarType>(dim, min, max); |
| 45 | |
| 46 | // validate dimensions |
| 47 | VERIFY_IS_EQUAL(v.size(), dim); |
| 48 | |
| 49 | // check entries |
| 50 | if(dim == 1) { |
| 51 | VERIFY_IS_APPROX(v(0), min); |
| 52 | } else { |
| 53 | VERIFY_IS_APPROX(v(0), max); |
| 54 | VERIFY_IS_APPROX(v(dim-1), min); |
| 55 | } |
| 56 | for(Index i = 0; i < v.size()-1; ++i) |
| 57 | VERIFY_GE(v(i), v(i+1)); |
| 58 | } |
| 59 | |
| 60 | template<typename MatrixType, typename RealScalar, typename RealVectorType> |
| 61 | void check_generateRandomMatrixSvs(const Index rows, const Index cols, const Index diag_size, |
| 62 | const RealScalar min_svs, const RealScalar max_svs) |
| 63 | { |
| 64 | RealVectorType svs = setupRangeSvs<RealVectorType, RealScalar>(diag_size, min_svs, max_svs); |
| 65 | |
| 66 | MatrixType M; |
| 67 | generateRandomMatrixSvs(svs, rows, cols, M); |
| 68 | |
| 69 | // validate dimensions |
| 70 | VERIFY_IS_EQUAL(M.rows(), rows); |
| 71 | VERIFY_IS_EQUAL(M.cols(), cols); |
| 72 | VERIFY_IS_EQUAL(svs.size(), diag_size); |
| 73 | |
| 74 | // validate singular values |
| 75 | Eigen::JacobiSVD<MatrixType> SVD(M); |
| 76 | VERIFY_IS_APPROX(svs, SVD.singularValues()); |
| 77 | } |
| 78 | |
| 79 | template<typename MatrixType> |
| 80 | void check_random_matrix(const MatrixType &m) |
| 81 | { |
| 82 | enum { |
| 83 | Rows = MatrixType::RowsAtCompileTime, |
| 84 | Cols = MatrixType::ColsAtCompileTime, |
Erik Schultheis | c20e908 | 2021-12-10 19:27:01 +0000 | [diff] [blame] | 85 | DiagSize = internal::min_size_prefer_dynamic(Rows, Cols) |
Kolja Brix | 58e086b | 2021-08-23 16:00:05 +0000 | [diff] [blame] | 86 | }; |
| 87 | typedef typename MatrixType::Scalar Scalar; |
| 88 | typedef typename NumTraits<Scalar>::Real RealScalar; |
| 89 | typedef Matrix<RealScalar, DiagSize, 1> RealVectorType; |
| 90 | |
| 91 | const Index rows = m.rows(), cols = m.cols(); |
| 92 | const Index diag_size = (std::min)(rows, cols); |
| 93 | const RealScalar min_svs = 1.0, max_svs = 1000.0; |
| 94 | |
| 95 | // check generation of unitary random matrices |
| 96 | typedef Matrix<Scalar, Rows, Rows> MatrixAType; |
| 97 | typedef Matrix<Scalar, Cols, Cols> MatrixBType; |
| 98 | check_generateRandomUnitaryMatrix<MatrixAType>(rows); |
| 99 | check_generateRandomUnitaryMatrix<MatrixBType>(cols); |
| 100 | |
| 101 | // test generators for singular values |
| 102 | check_setupRandomSvs<RealVectorType, RealScalar>(diag_size, max_svs); |
| 103 | check_setupRangeSvs<RealVectorType, RealScalar>(diag_size, min_svs, max_svs); |
| 104 | |
| 105 | // check generation of random matrices |
| 106 | check_generateRandomMatrixSvs<MatrixType, RealScalar, RealVectorType>(rows, cols, diag_size, min_svs, max_svs); |
| 107 | } |
| 108 | |
| 109 | EIGEN_DECLARE_TEST(random_matrix) |
| 110 | { |
| 111 | for(int i = 0; i < g_repeat; i++) { |
| 112 | CALL_SUBTEST_1(check_random_matrix(Matrix<float, 1, 1>())); |
| 113 | CALL_SUBTEST_2(check_random_matrix(Matrix<float, 4, 4>())); |
| 114 | CALL_SUBTEST_3(check_random_matrix(Matrix<float, 2, 3>())); |
| 115 | CALL_SUBTEST_4(check_random_matrix(Matrix<float, 7, 4>())); |
| 116 | |
| 117 | CALL_SUBTEST_5(check_random_matrix(Matrix<double, 1, 1>())); |
| 118 | CALL_SUBTEST_6(check_random_matrix(Matrix<double, 6, 6>())); |
| 119 | CALL_SUBTEST_7(check_random_matrix(Matrix<double, 5, 3>())); |
| 120 | CALL_SUBTEST_8(check_random_matrix(Matrix<double, 4, 9>())); |
| 121 | |
| 122 | CALL_SUBTEST_9(check_random_matrix(Matrix<std::complex<float>, 12, 12>())); |
| 123 | CALL_SUBTEST_10(check_random_matrix(Matrix<std::complex<float>, 7, 14>())); |
| 124 | CALL_SUBTEST_11(check_random_matrix(Matrix<std::complex<double>, 15, 11>())); |
| 125 | CALL_SUBTEST_12(check_random_matrix(Matrix<std::complex<double>, 6, 9>())); |
| 126 | |
| 127 | CALL_SUBTEST_13(check_random_matrix( |
| 128 | MatrixXf(internal::random<int>(1, EIGEN_TEST_MAX_SIZE), internal::random<int>(1, EIGEN_TEST_MAX_SIZE)))); |
| 129 | CALL_SUBTEST_14(check_random_matrix( |
| 130 | MatrixXd(internal::random<int>(1, EIGEN_TEST_MAX_SIZE), internal::random<int>(1, EIGEN_TEST_MAX_SIZE)))); |
| 131 | CALL_SUBTEST_15(check_random_matrix( |
| 132 | MatrixXcf(internal::random<int>(1, EIGEN_TEST_MAX_SIZE), internal::random<int>(1, EIGEN_TEST_MAX_SIZE)))); |
| 133 | CALL_SUBTEST_16(check_random_matrix( |
| 134 | MatrixXcd(internal::random<int>(1, EIGEN_TEST_MAX_SIZE), internal::random<int>(1, EIGEN_TEST_MAX_SIZE)))); |
| 135 | } |
| 136 | } |