src/lemon/graph_wrapper.h
changeset 1262 61f989e3e525
parent 1242 e48c4fe47aaf
child 1269 4c63ff4e16fa
equal deleted inserted replaced
20:1260f7ef15dd 21:c6281cef5ead
   432   problem of searching a maximum number of edge-disjoint shortest paths 
   432   problem of searching a maximum number of edge-disjoint shortest paths 
   433   between 
   433   between 
   434   two nodes \c s and \c t. Shortest here means being shortest w.r.t. 
   434   two nodes \c s and \c t. Shortest here means being shortest w.r.t. 
   435   non-negative edge-lengths. Note that 
   435   non-negative edge-lengths. Note that 
   436   the comprehension of the presented solution 
   436   the comprehension of the presented solution 
   437   need's some knowledge from elementary combinatorial optimization. 
   437   need's some elementary knowledge from combinatorial optimization. 
   438 
   438 
   439   If a single shortest path is to be 
   439   If a single shortest path is to be 
   440   searched between two nodes \c s and \c t, then this can be done easily by 
   440   searched between \c s and \c t, then this can be done easily by 
   441   applying the Dijkstra algorithm class. What happens, if a maximum number of 
   441   applying the Dijkstra algorithm. What happens, if a maximum number of 
   442   edge-disjoint shortest paths is to be computed. It can be proved that an 
   442   edge-disjoint shortest paths is to be computed. It can be proved that an 
   443   edge can be in a shortest path if and only if it is tight with respect to 
   443   edge can be in a shortest path if and only if it is tight with respect to 
   444   the potential function computed by Dijkstra. Moreover, any path containing 
   444   the potential function computed by Dijkstra. Moreover, any path containing 
   445   only such edges is a shortest one. Thus we have to compute a maximum number 
   445   only such edges is a shortest one. Thus we have to compute a maximum number 
   446   of edge-disjoint paths between \c s and \c t in the graph which has edge-set 
   446   of edge-disjoint paths between \c s and \c t in the graph which has edge-set 
   921   /// everywhere true. We note that for sake of efficiency, 
   921   /// everywhere true. We note that for sake of efficiency, 
   922   /// \c RevGraphWrapper is implemented in a different way. 
   922   /// \c RevGraphWrapper is implemented in a different way. 
   923   /// But BidirGraphWrapper is obtained from 
   923   /// But BidirGraphWrapper is obtained from 
   924   /// SubBidirGraphWrapper by considering everywhere true 
   924   /// SubBidirGraphWrapper by considering everywhere true 
   925   /// valued maps both for forward_filter and backward_filter. 
   925   /// valued maps both for forward_filter and backward_filter. 
   926   /// Finally, one of the most important applications of SubBidirGraphWrapper 
   926   ///
       
   927   /// The most important application of SubBidirGraphWrapper 
   927   /// is ResGraphWrapper, which stands for the residual graph in directed 
   928   /// is ResGraphWrapper, which stands for the residual graph in directed 
   928   /// flow and circulation problems. 
   929   /// flow and circulation problems. 
   929   /// As wrappers usually, the SubBidirGraphWrapper implements the 
   930   /// As wrappers usually, the SubBidirGraphWrapper implements the 
   930   /// above mentioned graph structure without its physical storage, 
   931   /// above mentioned graph structure without its physical storage, 
   931   /// that is the whole stuff is stored in constant memory. 
   932   /// that is the whole stuff is stored in constant memory.