Austin Schuh | 70cc955 | 2019-01-21 19:46:48 -0800 | [diff] [blame] | 1 | // Ceres Solver - A fast non-linear least squares minimizer |
| 2 | // Copyright 2015 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: vitus@google.com (Michael Vitus) |
| 30 | |
Austin Schuh | 1d1e6ea | 2020-12-23 21:56:30 -0800 | [diff] [blame^] | 31 | #ifndef CERES_PUBLIC_INTERNAL_HOUSEHOLDER_VECTOR_H_ |
| 32 | #define CERES_PUBLIC_INTERNAL_HOUSEHOLDER_VECTOR_H_ |
Austin Schuh | 70cc955 | 2019-01-21 19:46:48 -0800 | [diff] [blame] | 33 | |
| 34 | #include "Eigen/Core" |
| 35 | #include "glog/logging.h" |
| 36 | |
| 37 | namespace ceres { |
| 38 | namespace internal { |
| 39 | |
| 40 | // Algorithm 5.1.1 from 'Matrix Computations' by Golub et al. (Johns Hopkins |
| 41 | // Studies in Mathematical Sciences) but using the nth element of the input |
| 42 | // vector as pivot instead of first. This computes the vector v with v(n) = 1 |
| 43 | // and beta such that H = I - beta * v * v^T is orthogonal and |
| 44 | // H * x = ||x||_2 * e_n. |
Austin Schuh | 1d1e6ea | 2020-12-23 21:56:30 -0800 | [diff] [blame^] | 45 | // |
| 46 | // NOTE: Some versions of MSVC have trouble deducing the type of v if |
| 47 | // you do not specify all the template arguments explicitly. |
| 48 | template <typename XVectorType, typename Scalar, int N> |
| 49 | void ComputeHouseholderVector(const XVectorType& x, |
| 50 | Eigen::Matrix<Scalar, N, 1>* v, |
Austin Schuh | 70cc955 | 2019-01-21 19:46:48 -0800 | [diff] [blame] | 51 | Scalar* beta) { |
| 52 | CHECK(beta != nullptr); |
| 53 | CHECK(v != nullptr); |
| 54 | CHECK_GT(x.rows(), 1); |
| 55 | CHECK_EQ(x.rows(), v->rows()); |
| 56 | |
| 57 | Scalar sigma = x.head(x.rows() - 1).squaredNorm(); |
| 58 | *v = x; |
| 59 | (*v)(v->rows() - 1) = Scalar(1.0); |
| 60 | |
| 61 | *beta = Scalar(0.0); |
| 62 | const Scalar& x_pivot = x(x.rows() - 1); |
| 63 | |
| 64 | if (sigma <= Scalar(std::numeric_limits<double>::epsilon())) { |
| 65 | if (x_pivot < Scalar(0.0)) { |
| 66 | *beta = Scalar(2.0); |
| 67 | } |
| 68 | return; |
| 69 | } |
| 70 | |
| 71 | const Scalar mu = sqrt(x_pivot * x_pivot + sigma); |
| 72 | Scalar v_pivot = Scalar(1.0); |
| 73 | |
| 74 | if (x_pivot <= Scalar(0.0)) { |
| 75 | v_pivot = x_pivot - mu; |
| 76 | } else { |
| 77 | v_pivot = -sigma / (x_pivot + mu); |
| 78 | } |
| 79 | |
| 80 | *beta = Scalar(2.0) * v_pivot * v_pivot / (sigma + v_pivot * v_pivot); |
| 81 | |
| 82 | v->head(v->rows() - 1) /= v_pivot; |
| 83 | } |
| 84 | |
| 85 | } // namespace internal |
| 86 | } // namespace ceres |
| 87 | |
Austin Schuh | 1d1e6ea | 2020-12-23 21:56:30 -0800 | [diff] [blame^] | 88 | #endif // CERES_PUBLIC_INTERNAL_HOUSEHOLDER_VECTOR_H_ |