• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 // This file is part of Eigen, a lightweight C++ template library
2 // for linear algebra.
3 //
4 // Copyright (C) 2011-2014 Gael Guennebaud <gael.guennebaud@inria.fr>
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 #ifndef EIGEN_CONJUGATE_GRADIENT_H
11 #define EIGEN_CONJUGATE_GRADIENT_H
12 
13 namespace Eigen {
14 
15 namespace internal {
16 
17 /** \internal Low-level conjugate gradient algorithm
18   * \param mat The matrix A
19   * \param rhs The right hand side vector b
20   * \param x On input and initial solution, on output the computed solution.
21   * \param precond A preconditioner being able to efficiently solve for an
22   *                approximation of Ax=b (regardless of b)
23   * \param iters On input the max number of iteration, on output the number of performed iterations.
24   * \param tol_error On input the tolerance error, on output an estimation of the relative error.
25   */
26 template<typename MatrixType, typename Rhs, typename Dest, typename Preconditioner>
27 EIGEN_DONT_INLINE
conjugate_gradient(const MatrixType & mat,const Rhs & rhs,Dest & x,const Preconditioner & precond,Index & iters,typename Dest::RealScalar & tol_error)28 void conjugate_gradient(const MatrixType& mat, const Rhs& rhs, Dest& x,
29                         const Preconditioner& precond, Index& iters,
30                         typename Dest::RealScalar& tol_error)
31 {
32   using std::sqrt;
33   using std::abs;
34   typedef typename Dest::RealScalar RealScalar;
35   typedef typename Dest::Scalar Scalar;
36   typedef Matrix<Scalar,Dynamic,1> VectorType;
37 
38   RealScalar tol = tol_error;
39   Index maxIters = iters;
40 
41   Index n = mat.cols();
42 
43   VectorType residual = rhs - mat * x; //initial residual
44 
45   RealScalar rhsNorm2 = rhs.squaredNorm();
46   if(rhsNorm2 == 0)
47   {
48     x.setZero();
49     iters = 0;
50     tol_error = 0;
51     return;
52   }
53   const RealScalar considerAsZero = (std::numeric_limits<RealScalar>::min)();
54   RealScalar threshold = numext::maxi(RealScalar(tol*tol*rhsNorm2),considerAsZero);
55   RealScalar residualNorm2 = residual.squaredNorm();
56   if (residualNorm2 < threshold)
57   {
58     iters = 0;
59     tol_error = sqrt(residualNorm2 / rhsNorm2);
60     return;
61   }
62 
63   VectorType p(n);
64   p = precond.solve(residual);      // initial search direction
65 
66   VectorType z(n), tmp(n);
67   RealScalar absNew = numext::real(residual.dot(p));  // the square of the absolute value of r scaled by invM
68   Index i = 0;
69   while(i < maxIters)
70   {
71     tmp.noalias() = mat * p;                    // the bottleneck of the algorithm
72 
73     Scalar alpha = absNew / p.dot(tmp);         // the amount we travel on dir
74     x += alpha * p;                             // update solution
75     residual -= alpha * tmp;                    // update residual
76 
77     residualNorm2 = residual.squaredNorm();
78     if(residualNorm2 < threshold)
79       break;
80 
81     z = precond.solve(residual);                // approximately solve for "A z = residual"
82 
83     RealScalar absOld = absNew;
84     absNew = numext::real(residual.dot(z));     // update the absolute value of r
85     RealScalar beta = absNew / absOld;          // calculate the Gram-Schmidt value used to create the new search direction
86     p = z + beta * p;                           // update search direction
87     i++;
88   }
89   tol_error = sqrt(residualNorm2 / rhsNorm2);
90   iters = i;
91 }
92 
93 }
94 
95 template< typename _MatrixType, int _UpLo=Lower,
96           typename _Preconditioner = DiagonalPreconditioner<typename _MatrixType::Scalar> >
97 class ConjugateGradient;
98 
99 namespace internal {
100 
101 template< typename _MatrixType, int _UpLo, typename _Preconditioner>
102 struct traits<ConjugateGradient<_MatrixType,_UpLo,_Preconditioner> >
103 {
104   typedef _MatrixType MatrixType;
105   typedef _Preconditioner Preconditioner;
106 };
107 
108 }
109 
110 /** \ingroup IterativeLinearSolvers_Module
111   * \brief A conjugate gradient solver for sparse (or dense) self-adjoint problems
112   *
113   * This class allows to solve for A.x = b linear problems using an iterative conjugate gradient algorithm.
114   * The matrix A must be selfadjoint. The matrix A and the vectors x and b can be either dense or sparse.
115   *
116   * \tparam _MatrixType the type of the matrix A, can be a dense or a sparse matrix.
117   * \tparam _UpLo the triangular part that will be used for the computations. It can be Lower,
118   *               \c Upper, or \c Lower|Upper in which the full matrix entries will be considered.
119   *               Default is \c Lower, best performance is \c Lower|Upper.
120   * \tparam _Preconditioner the type of the preconditioner. Default is DiagonalPreconditioner
121   *
122   * \implsparsesolverconcept
123   *
124   * The maximal number of iterations and tolerance value can be controlled via the setMaxIterations()
125   * and setTolerance() methods. The defaults are the size of the problem for the maximal number of iterations
126   * and NumTraits<Scalar>::epsilon() for the tolerance.
127   *
128   * The tolerance corresponds to the relative residual error: |Ax-b|/|b|
129   *
130   * \b Performance: Even though the default value of \c _UpLo is \c Lower, significantly higher performance is
131   * achieved when using a complete matrix and \b Lower|Upper as the \a _UpLo template parameter. Moreover, in this
132   * case multi-threading can be exploited if the user code is compiled with OpenMP enabled.
133   * See \ref TopicMultiThreading for details.
134   *
135   * This class can be used as the direct solver classes. Here is a typical usage example:
136     \code
137     int n = 10000;
138     VectorXd x(n), b(n);
139     SparseMatrix<double> A(n,n);
140     // fill A and b
141     ConjugateGradient<SparseMatrix<double>, Lower|Upper> cg;
142     cg.compute(A);
143     x = cg.solve(b);
144     std::cout << "#iterations:     " << cg.iterations() << std::endl;
145     std::cout << "estimated error: " << cg.error()      << std::endl;
146     // update b, and solve again
147     x = cg.solve(b);
148     \endcode
149   *
150   * By default the iterations start with x=0 as an initial guess of the solution.
151   * One can control the start using the solveWithGuess() method.
152   *
153   * ConjugateGradient can also be used in a matrix-free context, see the following \link MatrixfreeSolverExample example \endlink.
154   *
155   * \sa class LeastSquaresConjugateGradient, class SimplicialCholesky, DiagonalPreconditioner, IdentityPreconditioner
156   */
157 template< typename _MatrixType, int _UpLo, typename _Preconditioner>
158 class ConjugateGradient : public IterativeSolverBase<ConjugateGradient<_MatrixType,_UpLo,_Preconditioner> >
159 {
160   typedef IterativeSolverBase<ConjugateGradient> Base;
161   using Base::matrix;
162   using Base::m_error;
163   using Base::m_iterations;
164   using Base::m_info;
165   using Base::m_isInitialized;
166 public:
167   typedef _MatrixType MatrixType;
168   typedef typename MatrixType::Scalar Scalar;
169   typedef typename MatrixType::RealScalar RealScalar;
170   typedef _Preconditioner Preconditioner;
171 
172   enum {
173     UpLo = _UpLo
174   };
175 
176 public:
177 
178   /** Default constructor. */
179   ConjugateGradient() : Base() {}
180 
181   /** Initialize the solver with matrix \a A for further \c Ax=b solving.
182     *
183     * This constructor is a shortcut for the default constructor followed
184     * by a call to compute().
185     *
186     * \warning this class stores a reference to the matrix A as well as some
187     * precomputed values that depend on it. Therefore, if \a A is changed
188     * this class becomes invalid. Call compute() to update it with the new
189     * matrix A, or modify a copy of A.
190     */
191   template<typename MatrixDerived>
192   explicit ConjugateGradient(const EigenBase<MatrixDerived>& A) : Base(A.derived()) {}
193 
194   ~ConjugateGradient() {}
195 
196   /** \internal */
197   template<typename Rhs,typename Dest>
198   void _solve_vector_with_guess_impl(const Rhs& b, Dest& x) const
199   {
200     typedef typename Base::MatrixWrapper MatrixWrapper;
201     typedef typename Base::ActualMatrixType ActualMatrixType;
202     enum {
203       TransposeInput  =   (!MatrixWrapper::MatrixFree)
204                       &&  (UpLo==(Lower|Upper))
205                       &&  (!MatrixType::IsRowMajor)
206                       &&  (!NumTraits<Scalar>::IsComplex)
207     };
208     typedef typename internal::conditional<TransposeInput,Transpose<const ActualMatrixType>, ActualMatrixType const&>::type RowMajorWrapper;
209     EIGEN_STATIC_ASSERT(EIGEN_IMPLIES(MatrixWrapper::MatrixFree,UpLo==(Lower|Upper)),MATRIX_FREE_CONJUGATE_GRADIENT_IS_COMPATIBLE_WITH_UPPER_UNION_LOWER_MODE_ONLY);
210     typedef typename internal::conditional<UpLo==(Lower|Upper),
211                                            RowMajorWrapper,
212                                            typename MatrixWrapper::template ConstSelfAdjointViewReturnType<UpLo>::Type
213                                           >::type SelfAdjointWrapper;
214 
215     m_iterations = Base::maxIterations();
216     m_error = Base::m_tolerance;
217 
218     RowMajorWrapper row_mat(matrix());
219     internal::conjugate_gradient(SelfAdjointWrapper(row_mat), b, x, Base::m_preconditioner, m_iterations, m_error);
220     m_info = m_error <= Base::m_tolerance ? Success : NoConvergence;
221   }
222 
223 protected:
224 
225 };
226 
227 } // end namespace Eigen
228 
229 #endif // EIGEN_CONJUGATE_GRADIENT_H
230