Include Lapack svd in Solver interface (another model for SvdTraits)

Documentation/doc/Documentation/Installation.txt

@@ -508,6 +508,15 @@ and the \ref PkgRidges_3 packages.
 
 The \sc{Eigen} web site is <A HREF="http://eigen.tuxfamily.org">`http://eigen.tuxfamily.org`</A>.
 
+\subsection thirdpartyLapack Lapack
+
+\sc{Lapack} is a `Fortran` library for linear algebra.
+
+In \cgal, \sc{Lapack} is an alternative to \sc{Eigen} for singular value decomposition for the \ref
+kgJet_fitting_3 and the \ref PkgRidges_3 packages.
+
+The \sc{Lapack} web site is <A HREF="http://www.netlib.org/lapack/">`http://www.netlib.org/lapack/`</A>.
+
 \subsection thirdpartylibQGLViewer libQGLViewer 
 
 libQGLViewer is a 3D widget based on \sc{Qt} 4's `QGLWidget`. In case of \sc{Qt}5 used, libQGLViewer needs to be recompiled with the proper \sc{Qt}5 version.

Solver_interface/doc/Solver_interface/Solver_interface.txt

@@ -42,15 +42,15 @@ class:
 
 The concept `SvdTraits` defines an interface for solving in the least
 square sense a linear system with a singular value decomposition. The
-field type is `double`.
+field type is `double`. We provide two models for this concept:
 
-The class `Eigen_svd` provides an implementation of `SvdTraits` using
+- `Eigen_svd` uses the \ref thirdpartyEigen library.
-the \ref thirdpartyEigen library.
+- `Lapack_svd` uses the \ref thirdpartyLapack library.
 
 Here is a simple example that shows how to handle matrices, vectors
 and this solver:
 
-\cgalExample{Solver_interface/eigen_singular_value_decomposition.cpp}
+\cgalExample{Solver_interface/singular_value_decomposition.cpp}
 
 
 

Solver_interface/doc/Solver_interface/examples.txt

@@ -1,5 +1,5 @@
 /*!
 \example Solver_interface/diagonalize_matrix.cpp	
-\example Solver_interface/eigen_singular_value_decomposition.cpp
+\example Solver_interface/singular_value_decomposition.cpp
 \example Solver_interface/eigen_sparse_solvers.cpp
 */

Solver_interface/examples/Solver_interface/CMakeLists.txt

@@ -24,7 +24,7 @@ if ( CGAL_FOUND )
   
   include_directories (BEFORE "../include")
 
-  create_single_source_cgal_program( "eigen_singular_value_decomposition.cpp" )
+  create_single_source_cgal_program( "singular_value_decomposition.cpp" )
   create_single_source_cgal_program( "diagonalize_matrix.cpp" )
   create_single_source_cgal_program( "eigen_sparse_solvers.cpp" )
 

Solver_interface/examples/Solver_interface/singular_value_decomposition.cpp

@@ -1,11 +1,20 @@
 #include <CGAL/Simple_cartesian.h>
+
+#ifdef CGAL_EIGEN3_ENABLED
 #include <CGAL/Eigen_matrix.h>
 #include <CGAL/Eigen_vector.h>
 #include <CGAL/Eigen_svd.h>
+typedef CGAL::Eigen_svd Svd;
+#else
+#ifdef CGAL_LAPACK_ENABLED
+#include <CGAL/Lapack_svd.h>
+typedef CGAL::Lapack_svd Svd;
+#endif
+#endif
 
-typedef CGAL::Eigen_svd::FT     FT;
+typedef Svd::FT     FT;
-typedef CGAL::Eigen_svd::Vector Eigen_vector;
+typedef Svd::Vector Eigen_vector;
-typedef CGAL::Eigen_svd::Matrix Eigen_matrix;
+typedef Svd::Matrix Eigen_matrix;
 
 int main(void)
 {
@@ -14,7 +23,6 @@ int main(void)
   Eigen_vector B (degree);
   Eigen_matrix M (degree, degree);
 
-
   // Fill B and M with random numbers
   for (std::size_t i = 0; i < degree; ++ i)
     {
@@ -24,7 +32,7 @@ int main(void)
     }
 
   // Solve MX=B
-  std::cout << CGAL::Eigen_svd::solve(M, B) << std::endl;
+  std::cout << Svd::solve(M, B) << std::endl;
 
   // Print result
   std::cout << "Solution of SVD = [ ";