Main MRPT website > C++ reference for MRPT 1.5.7
matrix_ops4_unittest.cpp
Go to the documentation of this file.
1 /* +---------------------------------------------------------------------------+
2  | Mobile Robot Programming Toolkit (MRPT) |
3  | http://www.mrpt.org/ |
4  | |
5  | Copyright (c) 2005-2017, Individual contributors, see AUTHORS file |
6  | See: http://www.mrpt.org/Authors - All rights reserved. |
7  | Released under BSD License. See details in http://www.mrpt.org/License |
8  +---------------------------------------------------------------------------+ */
9 
10 // Note: Matrices unit tests have been split in different files since
11 // building them with eigen3 eats a lot of RAM and may be a problem while
12 // compiling in small systems.
13 
14 #include <mrpt/math/CMatrixFixedNumeric.h>
15 #include <mrpt/utils/metaprogramming.h>
16 #include <mrpt/utils/CMemoryStream.h>
17 #include <mrpt/random.h>
18 #include <gtest/gtest.h>
19 #include <numeric> // std::accumulate()
20 
21 using namespace mrpt;
22 using namespace mrpt::utils;
23 using namespace mrpt::math;
24 using namespace mrpt::random;
25 using namespace mrpt::utils::metaprogramming;
26 using namespace std;
27 
28 
29 TEST(Matrices,meanAndStd)
30 {
31  /* meanAndStd: Computes a row with the mean values of each column in the matrix and
32  the associated vector with the standard deviation of each column. */
33 
34  const double dat_A[] = {2.8955668335,2.3041932983,1.9002381085,1.7993158652,1.8456197228,2.9632296740,1.9368565578,2.1988923358,2.0547605617,2.5655678993,2.3041932983,3.8406914364,2.1811218706,3.2312564555,2.4736403918,3.4703311380,1.4874417483,3.1073538218,2.1353324397,2.9541115932,1.9002381085,2.1811218706,2.4942067597,1.6851007198,1.4585872052,2.3015952197,1.0955231591,2.2979627790,1.3918738834,2.1854562572,1.7993158652,3.2312564555,1.6851007198,3.1226161015,1.6779632687,2.7195826381,1.2397348013,2.3757864319,1.6291224768,2.4463194915,1.8456197228,2.4736403918,1.4585872052,1.6779632687,2.8123267839,2.5860688816,1.4131630919,2.1914803135,1.5542420639,2.7170092067,2.9632296740,3.4703311380,2.3015952197,2.7195826381,2.5860688816,4.1669180394,2.1145239023,3.3214801332,2.6694845663,3.0742063088,1.9368565578,1.4874417483,1.0955231591,1.2397348013,1.4131630919,2.1145239023,1.8928811570,1.7097998455,1.7205860530,1.8710847505,2.1988923358,3.1073538218,2.2979627790,2.3757864319,2.1914803135,3.3214801332,1.7097998455,3.4592638415,2.1518695071,2.8907499694,2.0547605617,2.1353324397,1.3918738834,1.6291224768,1.5542420639,2.6694845663,1.7205860530,2.1518695071,2.1110960664,1.6731209980,2.5655678993,2.9541115932,2.1854562572,2.4463194915,2.7170092067,3.0742063088,1.8710847505,2.8907499694,1.6731209980,3.9093678727};
35  CMatrixFixedNumeric<double,10,10> A(dat_A);
36 
37  // Compute mean & std of each column:
38  CVectorDouble result_mean, result_std;
39  A.meanAndStd(result_mean, result_std);
40 
41  // Result from MATLAB:
42  const double dat_good_M[] = { 2.246424086, 2.718547419, 1.899166596, 2.192679825, 2.073010093, 2.938742050, 1.648159507, 2.570463898, 1.909148862, 2.628699435 };
43  const Eigen::Matrix<double,10,1> good_M(dat_good_M);
44  const double dat_good_S[] = { 0.428901371, 0.720352792, 0.468999497, 0.684910097, 0.546595053, 0.604303301, 0.328759015, 0.582584159, 0.382009344, 0.644788760 };
45  const Eigen::Matrix<double,10,1> good_S(dat_good_S);
46 
47  EXPECT_NEAR((result_mean-good_M).array().abs().sum(),0,1e-4);
48  EXPECT_NEAR((result_std-good_S).array().abs().sum(),0,1e-4);
49 }
50 
51 TEST(Matrices,meanAndStdAll)
52 {
53  /* meanAndStd: Computes a row with the mean values of each column in the matrix and
54  the associated vector with the standard deviation of each column. */
55 
56  const double dat_A[] = {2.8955668335,2.3041932983,1.9002381085,1.7993158652,1.8456197228,2.9632296740,1.9368565578,2.1988923358,2.0547605617,2.5655678993,2.3041932983,3.8406914364,2.1811218706,3.2312564555,2.4736403918,3.4703311380,1.4874417483,3.1073538218,2.1353324397,2.9541115932,1.9002381085,2.1811218706,2.4942067597,1.6851007198,1.4585872052,2.3015952197,1.0955231591,2.2979627790,1.3918738834,2.1854562572,1.7993158652,3.2312564555,1.6851007198,3.1226161015,1.6779632687,2.7195826381,1.2397348013,2.3757864319,1.6291224768,2.4463194915,1.8456197228,2.4736403918,1.4585872052,1.6779632687,2.8123267839,2.5860688816,1.4131630919,2.1914803135,1.5542420639,2.7170092067,2.9632296740,3.4703311380,2.3015952197,2.7195826381,2.5860688816,4.1669180394,2.1145239023,3.3214801332,2.6694845663,3.0742063088,1.9368565578,1.4874417483,1.0955231591,1.2397348013,1.4131630919,2.1145239023,1.8928811570,1.7097998455,1.7205860530,1.8710847505,2.1988923358,3.1073538218,2.2979627790,2.3757864319,2.1914803135,3.3214801332,1.7097998455,3.4592638415,2.1518695071,2.8907499694,2.0547605617,2.1353324397,1.3918738834,1.6291224768,1.5542420639,2.6694845663,1.7205860530,2.1518695071,2.1110960664,1.6731209980,2.5655678993,2.9541115932,2.1854562572,2.4463194915,2.7170092067,3.0742063088,1.8710847505,2.8907499694,1.6731209980,3.9093678727};
57  CMatrixFixedNumeric<double,10,10> A(dat_A);
58 
59  // Compute mean & std of each column:
60  double result_mean, result_std;
61  A.meanAndStdAll(result_mean, result_std);
62 
63  // Result from MATLAB:
64  const double good_M = 2.282504177034;
65  const double good_S = 0.660890754096;
66 
67  EXPECT_NEAR(std::abs(result_mean-good_M),0,1e-4);
68  EXPECT_NEAR(std::abs(result_std-good_S),0,1e-4);
69 }
70 
71 TEST(Matrices,laplacian)
72 {
73  // The laplacian matrix of W is L = D - W. (D:diagonals with degrees of nodes)
74  const double W_vals[6*6]={
75  0, 1, 0, 0, 1, 0,
76  1, 0, 1, 0, 1, 0,
77  0, 1, 0, 1, 0, 0,
78  0, 0, 1, 0, 1, 1,
79  1, 1, 0, 1, 0, 0,
80  0, 0, 0, 1, 0, 0 };
81  const CMatrixDouble W(6,6, W_vals);
82 
83  CMatrixDouble L;
84  W.laplacian(L);
85 
86  const double real_laplacian_vals[6*6]={
87  2, -1, 0, 0, -1, 0,
88  -1, 3, -1, 0, -1, 0,
89  0, -1, 2, -1, 0, 0,
90  0, 0, -1, 3, -1, -1,
91  -1, -1, 0, -1, 3, 0,
92  0, 0, 0, -1, 0, 1 };
93  const CMatrixDouble GT_L(6,6, real_laplacian_vals);
94 
95  EXPECT_NEAR( (GT_L-L).array().abs().sum(), 0, 1e-4);
96 }
97 
98 TEST(Matrices,largestEigenvector)
99 {
100  {
101  const double dat_C1[] = {
102  13.737245,10.248641,-5.839599,11.108320,
103  10.248641,14.966139,-5.259922,11.662222,
104  -5.839599,-5.259922,9.608822,-4.342505,
105  11.108320,11.662222,-4.342505,12.121940 };
106  const CMatrixDouble44 C1(dat_C1);
107 
108  const double dat_REAL_EIGVEC[] = { 0.54800 , 0.57167, -0.29604 , 0.53409 };
109  const Eigen::Matrix<double,4,1> REAL_EIGVEC(dat_REAL_EIGVEC);
110  //const double REAL_LARGEST_EIGENVALUE = 38.40966;
111 
112  mrpt::math::CVectorDouble lev;
113  C1.largestEigenvector(lev,1e-3, 20);
114  EXPECT_NEAR( (REAL_EIGVEC-lev).array().abs().sum(), 0, 1e-3);
115  }
116 }
117 
118 TEST(Matrices,loadFromTextFile)
119 {
120  {
121  const std::string s1 =
122  "1 2 3\n"
123  "4 5 6";
124  std::stringstream s(s1);
125  CMatrixDouble M;
126  bool retval = false;
127  try { M.loadFromTextFile(s); retval=true; } catch(std::exception &e) { std::cerr << e.what() << std::endl; }
128  EXPECT_TRUE(retval) << "string:\n" << s1 << endl;
129  EXPECT_EQ(M.rows(),2);
130  EXPECT_EQ(M.cols(),3);
131  }
132  {
133  const std::string s1 =
134  "1 \t 2\n"
135  " 4 \t\t 1 ";
136  std::stringstream s(s1);
137  CMatrixDouble M;
138  bool retval = false;
139  try { M.loadFromTextFile(s); retval=true; } catch(std::exception &e) { std::cerr << e.what() << std::endl; }
140  EXPECT_TRUE(retval) << "string:\n" << s1 << endl;
141  EXPECT_EQ(M.rows(),2);
142  EXPECT_EQ(M.cols(),2);
143  }
144  {
145  const std::string s1 =
146  "1 2";
147  std::stringstream s(s1);
148  CMatrixDouble M;
149  bool retval = false;
150  try { M.loadFromTextFile(s); retval=true; } catch(std::exception &e) { std::cerr << e.what() << std::endl; }
151  EXPECT_TRUE(retval) << "string:\n" << s1 << endl;
152  EXPECT_EQ(M.rows(),1);
153  EXPECT_EQ(M.cols(),2);
154  }
155  {
156  const std::string s1 =
157  "1 2 3\n"
158  "4 5 6\n";
159  std::stringstream s(s1);
160  CMatrixFixedNumeric<double,2,3> M;
161  bool retval = false;
162  try { M.loadFromTextFile(s); retval=true; } catch(std::exception &e) { std::cerr << e.what() << std::endl; }
163  EXPECT_TRUE(retval) << "string:\n" << s1 << endl;
164  EXPECT_EQ(M.rows(),2);
165  EXPECT_EQ(M.cols(),3);
166  }
167  {
168  const std::string s1 =
169  "1 2 3\n"
170  "4 5\n";
171  std::stringstream s(s1);
172  CMatrixFixedNumeric<double,2,3> M;
173  bool retval = false;
174  try { M.loadFromTextFile(s); retval=true; } catch(std::exception &) { }
175  EXPECT_FALSE(retval) << "string:\n" << s1 << endl;
176  }
177  {
178  const std::string s1 =
179  "1 2 3\n"
180  "4 5\n";
181  std::stringstream s(s1);
182  CMatrixDouble M;
183  bool retval = false;
184  try { M.loadFromTextFile(s); retval=true; } catch(std::exception &) { }
185  EXPECT_FALSE(retval) << "string:\n" << s1 << endl;
186  }
187  {
188  const std::string s1 =
189  " \n";
190  std::stringstream s(s1);
191  CMatrixFixedNumeric<double,2,3> M;
192  bool retval = false;
193  try { M.loadFromTextFile(s); retval=true; } catch(std::exception &) { }
194  EXPECT_FALSE(retval) << "string:\n" << s1 << endl;
195  }
196  {
197  const std::string s1 =
198  "1 2 3\n"
199  "1 2 3\n"
200  "1 2 3";
201  std::stringstream s(s1);
202  CMatrixFixedNumeric<double,2,3> M;
203  bool retval = false;
204  try { M.loadFromTextFile(s); retval=true; } catch(std::exception &) { }
205  EXPECT_FALSE(retval) << "string:\n" << s1 << endl;
206  }
207 }
208 
209 
mrpt::math::CMatrixFixedNumeric
A numeric matrix of compile-time fixed size.
Definition: CMatrixFixedNumeric.h:43
mrpt::math::dynamic_vector
Column vector, like Eigen::MatrixX*, but automatically initialized to zeros since construction.
Definition: types_math.h:65
largestEigenvector
void largestEigenvector(OUTVECT &x, Scalar resolution=Scalar(0.01), size_t maxIterations=6, int *out_Iterations=NULL, float *out_estimatedResolution=NULL) const
Efficiently computes only the biggest eigenvector of the matrix using the Power Method,...
Definition: eigen_plugins.h:301
meanAndStd
void meanAndStd(VEC &outMeanVector, VEC &outStdVector, const bool unbiased_variance=true) const
Computes a row with the mean values of each column in the matrix and the associated vector with the s...
Definition: eigen_plugins.h:366
meanAndStdAll
void meanAndStdAll(double &outMean, double &outStd, const bool unbiased_variance=true) const
Computes the mean and standard deviation of all the elements in the matrix as a whole.
Definition: eigen_plugins.h:388
loadFromTextFile
void loadFromTextFile(const std::string &file)
Load matrix from a text file, compatible with MATLAB text format.
laplacian
EIGEN_STRONG_INLINE void laplacian(Eigen::MatrixBase< OtherDerived > &ret) const
Computes the laplacian of this square graph weight matrix.
Definition: eigen_plugins.h:265
s
GLdouble s
Definition: glext.h:3602
string
GLsizei const GLchar ** string
Definition: glext.h:3919
dat_A
const double dat_A[]
Definition: matrix_ops1_unittest.cpp:31
TEST
TEST(Matrices, meanAndStd)
Definition: matrix_ops4_unittest.cpp:29
mrpt::math
This base provides a set of functions for maths stuff.
Definition: CArrayNumeric.h:20
mrpt::math::sum
CONTAINER::Scalar sum(const CONTAINER &v)
Computes the sum of all the elements.
Definition: ops_containers.h:167
mrpt::random
A namespace of pseudo-random numbers genrators of diferent distributions.
Definition: RandomGenerators.h:23
mrpt::utils::metaprogramming
A set of utility objects for metaprogramming with STL algorithms.
Definition: metaprogramming.h:25
mrpt::utils
Classes for serialization, sockets, ini-file manipulation, streams, list of properties-values,...
Definition: zip.h:16
mrpt
This is the global namespace for all Mobile Robot Programming Toolkit (MRPT) libraries.
Definition: CParticleFilter.h:17


Page generated by Doxygen 1.9.1 for MRPT 1.5.7 Git: 5902e14cc Wed Apr 24 15:04:01 2019 +0200 at mar 26 may 2026 13:12:03 CEST