diff --git a/matrix_sparse.hpp b/matrix_sparse.hpp
index baa3b5a8..9543eec9 100644
--- a/matrix_sparse.hpp
+++ b/matrix_sparse.hpp
@@ -246,8 +246,30 @@ namespace boost { namespace numeric { namespace ublas {
 
 #endif
 
-
+    // -----------------------------------
     // Index map based sparse matrix class
+    // -----------------------------------
+   /// \brief Index map based sparse matrix of values of type \c T. Orientation and storage can also be specified, otherwise a row major orientation is used.
+   /// Index map based sparse matrix of values of type \c T. Orientation and storage can also be specified, 
+   /// otherwise a row major orientation is used. It is \b not required by the storage to initialize elements 
+   /// of the matrix. By default, the orientation is \c row_major. 
+   /// This class represents a matrix by using a map \c key \f$\mapsto\f$ \c value. The default type is
+   /// \code
+   ///    template<class T, class L = row_major, class A =
+   ///    map_std<std::size_t, T> > class mapped_matrix;
+   /// \endcode
+   /// So, by default a STL map container is used to associate keys and values. The key is computed depending on 
+   /// the layout type \c L as
+   /// \code
+   ///    key = layout_type::element(i, size1_, j, size2_);
+   /// \endcode
+   /// which means \code key = (i*size2+j) \endcode for a row major matrix.
+   /// Limitations: The matrix size must not exceed \f$(size1*size2) < \f$ \code std::limits<std::size_t>\endcode. 
+   /// The \ref find1() and \ref find2() operations have a complexity of at least \f$\mathcal{O}(log(nnz))\f$, depending
+   /// on the efficiency of \c std::lower_bound on the key set of the map.
+   /// \sa fwd.hpp, storage_sparse.hpp
+   /// \tparam T the type of object stored in the matrix (like double, float, complex, etc...)
+   /// \tparam L the storage organization. It can be either \c row_major or \c column_major. By default it is \c row_major
     template<class T, class L, class A>
     class mapped_matrix:
         public matrix_container<mapped_matrix<T, L, A> > {