Home | History | Annotate | Download | only in ceres 
      1 // Ceres Solver - A fast non-linear least squares minimizer
      2 // Copyright 2010, 2011, 2012 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: sameeragarwal (at) google.com (Sameer Agarwal)
     30 
     31 #include "ceres/block_sparse_matrix.h"
     32 
     33 #include <string>
     34 #include "ceres/casts.h"
     35 #include "ceres/internal/eigen.h"
     36 #include "ceres/internal/scoped_ptr.h"
     37 #include "ceres/linear_least_squares_problems.h"
     38 #include "ceres/matrix_proto.h"
     39 #include "ceres/triplet_sparse_matrix.h"
     40 #include "glog/logging.h"
     41 #include "gtest/gtest.h"
     42 
     43 namespace ceres {
     44 namespace internal {
     45 
     46 class BlockSparseMatrixTest : public ::testing::Test {
     47  protected :
     48   virtual void SetUp() {
     49     scoped_ptr<LinearLeastSquaresProblem> problem(
     50         CreateLinearLeastSquaresProblemFromId(2));
     51     CHECK_NOTNULL(problem.get());
     52     A_.reset(down_cast<BlockSparseMatrix*>(problem->A.release()));
     53 
     54     problem.reset(CreateLinearLeastSquaresProblemFromId(1));
     55     CHECK_NOTNULL(problem.get());
     56     B_.reset(down_cast<TripletSparseMatrix*>(problem->A.release()));
     57 
     58     CHECK_EQ(A_->num_rows(), B_->num_rows());
     59     CHECK_EQ(A_->num_cols(), B_->num_cols());
     60     CHECK_EQ(A_->num_nonzeros(), B_->num_nonzeros());
     61   }
     62 
     63   scoped_ptr<BlockSparseMatrix> A_;
     64   scoped_ptr<TripletSparseMatrix> B_;
     65 };
     66 
     67 TEST_F(BlockSparseMatrixTest, SetZeroTest) {
     68   A_->SetZero();
     69   EXPECT_EQ(13, A_->num_nonzeros());
     70 }
     71 
     72 TEST_F(BlockSparseMatrixTest, RightMultiplyTest) {
     73   Vector y_a = Vector::Zero(A_->num_rows());
     74   Vector y_b = Vector::Zero(A_->num_rows());
     75   for (int i = 0; i < A_->num_cols(); ++i) {
     76     Vector x = Vector::Zero(A_->num_cols());
     77     x[i] = 1.0;
     78     A_->RightMultiply(x.data(), y_a.data());
     79     B_->RightMultiply(x.data(), y_b.data());
     80     EXPECT_LT((y_a - y_b).norm(), 1e-12);
     81   }
     82 }
     83 
     84 TEST_F(BlockSparseMatrixTest, LeftMultiplyTest) {
     85   Vector y_a = Vector::Zero(A_->num_cols());
     86   Vector y_b = Vector::Zero(A_->num_cols());
     87   for (int i = 0; i < A_->num_rows(); ++i) {
     88     Vector x = Vector::Zero(A_->num_rows());
     89     x[i] = 1.0;
     90     A_->LeftMultiply(x.data(), y_a.data());
     91     B_->LeftMultiply(x.data(), y_b.data());
     92     EXPECT_LT((y_a - y_b).norm(), 1e-12);
     93   }
     94 }
     95 
     96 TEST_F(BlockSparseMatrixTest, SquaredColumnNormTest) {
     97   Vector y_a = Vector::Zero(A_->num_cols());
     98   Vector y_b = Vector::Zero(A_->num_cols());
     99   A_->SquaredColumnNorm(y_a.data());
    100   B_->SquaredColumnNorm(y_b.data());
    101   EXPECT_LT((y_a - y_b).norm(), 1e-12);
    102 }
    103 
    104 TEST_F(BlockSparseMatrixTest, ToDenseMatrixTest) {
    105   Matrix m_a;
    106   Matrix m_b;
    107   A_->ToDenseMatrix(&m_a);
    108   B_->ToDenseMatrix(&m_b);
    109   EXPECT_LT((m_a - m_b).norm(), 1e-12);
    110 }
    111 
    112 #ifndef CERES_NO_PROTOCOL_BUFFERS
    113 TEST_F(BlockSparseMatrixTest, Serialization) {
    114   // Roundtrip through serialization and check for equality.
    115   SparseMatrixProto proto;
    116   A_->ToProto(&proto);
    117 
    118   LOG(INFO) << proto.DebugString();
    119 
    120   BlockSparseMatrix A2(proto);
    121 
    122   Matrix m_a;
    123   Matrix m_b;
    124   A_->ToDenseMatrix(&m_a);
    125   A2.ToDenseMatrix(&m_b);
    126 
    127   LOG(INFO) << "\n" << m_a;
    128   LOG(INFO) << "\n" << m_b;
    129 
    130   EXPECT_LT((m_a - m_b).norm(), 1e-12);
    131 }
    132 #endif
    133 
    134 }  // namespace internal
    135 }  // namespace ceres
    136