some more documentation
diff --git a/unsupported/Eigen/NonLinearOptimization b/unsupported/Eigen/NonLinearOptimization index 774e37a..80a11f1 100644 --- a/unsupported/Eigen/NonLinearOptimization +++ b/unsupported/Eigen/NonLinearOptimization
@@ -33,12 +33,41 @@ /** \ingroup Unsupported_modules * \defgroup NonLinearOptimization_Module Non linear optimization module * + * This module provides implementation of two important algorithms in non linear + * optimization. In both cases, we consider a system of non linear functions. Of + * course, this should work, and even work very well if those functions are + * actually linear. But if this is so, you should probably better use other + * methods more fitted to this special case. + * + * One algorithm allows to find the extremum of such a system (Levenberg + * Marquardt algorithm) and the second one is used to find + * a zero for the system (Powell hybrid "dogleg" method). + * + * This code is a port of a reknown implementation for both algorithms, + * called minpack (http://en.wikipedia.org/wiki/MINPACK). Those + * implementations have been carefully tuned, tested, and used for several + * decades. + * The original fortran code was automatically translated in C and then c++, + * and then cleaned by several authors + * (check http://devernay.free.fr/hacks/cminpack.html). + * + * Finally, we ported this code to Eigen, creating classes and API + * coherent with Eigen. When possible, we switched to Eigen + * implementation, such as most linear algebra (vectors, matrices, "good" norms). + * + * Doing so, we were very careful to check the tests we setup at the very + * beginning, which ensure that the same results are found, with the same + * number of iterations. + * * \code * #include <unsupported/Eigen/NonLinearOptimization> * \endcode */ + //@{ +#ifndef EIGEN_PARSED_BY_DOXYGEN + #include "src/NonLinearOptimization/qrsolv.h" #include "src/NonLinearOptimization/r1updt.h" #include "src/NonLinearOptimization/r1mpyq.h" @@ -52,9 +81,10 @@ #include "src/NonLinearOptimization/chkder.h" +#endif + #include "src/NonLinearOptimization/HybridNonLinearSolver.h" #include "src/NonLinearOptimization/LevenbergMarquardt.h" - //@} }
diff --git a/unsupported/Eigen/NumericalDiff b/unsupported/Eigen/NumericalDiff index 16e9f62..3a8bc32 100644 --- a/unsupported/Eigen/NumericalDiff +++ b/unsupported/Eigen/NumericalDiff
@@ -36,6 +36,22 @@ * Warning : this should NOT be confused with automatic differentiation, which * is a different method and has its own module in Eigen. * + * Currently only "Forward" and "Central" scheme are implemented. Those + * are basic methods, and there exist some more elaborated way of + * computing such approximates. They are implemented using both + * proprietary and free software, and usually requires linking to an + * external library. It is very easy for you to write a functor + * using such software, and the purpose is quite orthogonal to what we + * want to achieve with Eigen. + * + * This is why we will not provide wrappers for every great numerical + * differenciation software that exist, but should rather stick with those + * basic ones, that still are useful for testing. + * + * Also, the module "Non linear optimization" needs this in order to + * provide full features compatibility with the original (c)minpack + * package. + * * \code * #include <unsupported/Eigen/NumericalDiff> * \endcode
diff --git a/unsupported/Eigen/src/NonLinearOptimization/HybridNonLinearSolver.h b/unsupported/Eigen/src/NonLinearOptimization/HybridNonLinearSolver.h index e86401a..e944f0e 100644 --- a/unsupported/Eigen/src/NonLinearOptimization/HybridNonLinearSolver.h +++ b/unsupported/Eigen/src/NonLinearOptimization/HybridNonLinearSolver.h
@@ -28,6 +28,16 @@ #ifndef EIGEN_HYBRIDNONLINEARSOLVER_H #define EIGEN_HYBRIDNONLINEARSOLVER_H +/** + * \brief Finds a zero of a system of n + * nonlinear functions in n variables by a modification of the Powell + * hybrid method ("dogleg"). + * + * The user must provide a subroutine which calculates the + * functions. The Jacobian is either provided by the user, or approximated + * using a forward-difference method. + * + */ template<typename FunctorType, typename Scalar=double> class HybridNonLinearSolver {
diff --git a/unsupported/Eigen/src/NumericalDiff/NumericalDiff.h b/unsupported/Eigen/src/NumericalDiff/NumericalDiff.h index 4a84802..98872e0 100644 --- a/unsupported/Eigen/src/NumericalDiff/NumericalDiff.h +++ b/unsupported/Eigen/src/NumericalDiff/NumericalDiff.h
@@ -35,32 +35,15 @@ /** - * \brief asdf - * * This class allows you to add a method df() to your functor, which will * use numerical differentiation to compute an approximate of the * derivative for the functor. Of course, if you have an analytical form - * for the derivative, you should rather implement df() using it. + * for the derivative, you should rather implement df() by yourself. * * More information on * http://en.wikipedia.org/wiki/Numerical_differentiation * - * Currently only "Forward" and "Central" scheme are implemented. Those - * are basic methods, and there exist some more elaborated way of - * computing such approximates. They are implemented using both - * proprietary and free software, and usually requires linking to an - * external library. It is very easy for you to write a functor - * using such software, and the purpose is quite orthogonal to what we - * want to achieve with Eigen. - * - * This is why we will not provide wrappers for every great numerical - * differenciation software that exist, but should rather stick with those - * basic ones, that still are useful for testing. - * - * Also, the module "Non linear optimization" needs this in order to - * provide full features compatibility with the original (c)minpack - * package. - * + * Currently only "Forward" and "Central" scheme are implemented. */ template<typename Functor, NumericalDiffMode mode=Forward> class NumericalDiff : public Functor