src/work/marci/leda/bipartite_matching_leda_gen.cc
changeset 483 ce29ae5b2e1b
parent 459 68e6873f421a
child 496 7c463a7635d4
equal deleted inserted replaced
2:1b2cd21ef127 3:e167b817e49f
    13 #include <list_graph.h>
    13 #include <list_graph.h>
    14 //#include <smart_graph.h>
    14 //#include <smart_graph.h>
    15 //#include <dimacs.h>
    15 //#include <dimacs.h>
    16 #include <time_measure.h>
    16 #include <time_measure.h>
    17 #include <for_each_macros.h>
    17 #include <for_each_macros.h>
    18 //#include <bfs_iterator.h>
       
    19 #include <graph_wrapper.h>
    18 #include <graph_wrapper.h>
    20 #include <maps.h>
    19 #include <maps.h>
    21 #include <edmonds_karp.h>
    20 #include <max_flow.h>
    22 #include <preflow.h>
       
    23 
    21 
    24 /**
    22 /**
    25  * Inicializalja a veletlenszamgeneratort.
    23  * Inicializalja a veletlenszamgeneratort.
    26  * Figyelem, ez nem jo igazi random szamokhoz,
    24  * Figyelem, ez nem jo igazi random szamokhoz,
    27  * erre ne bizzad a titkaidat!
    25  * erre ne bizzad a titkaidat!
    76   std::cin >> m; 
    74   std::cin >> m; 
    77   int k;
    75   int k;
    78   std::cout << "A bipartite graph is a random group graph if the color classes \nA and B are partitiones to A_0, A_1, ..., A_{k-1} and B_0, B_1, ..., B_{k-1} \nas equally as possible \nand the edges from A_i goes to A_{i-1 mod k} and A_{i+1 mod k}.\n";
    76   std::cout << "A bipartite graph is a random group graph if the color classes \nA and B are partitiones to A_0, A_1, ..., A_{k-1} and B_0, B_1, ..., B_{k-1} \nas equally as possible \nand the edges from A_i goes to A_{i-1 mod k} and A_{i+1 mod k}.\n";
    79   std::cout << "number of groups in LEDA random group graph=";
    77   std::cout << "number of groups in LEDA random group graph=";
    80   std::cin >> k; 
    78   std::cin >> k; 
    81 
    79   std::cout << std::endl;
       
    80   
    82   leda_list<leda_node> lS;
    81   leda_list<leda_node> lS;
    83   leda_list<leda_node> lT;
    82   leda_list<leda_node> lT;
    84   random_bigraph(lg, a, b, m, lS, lT, k);
    83   random_bigraph(lg, a, b, m, lS, lT, k);
    85 
    84 
    86 //   for (int i=0; i<a; ++i) s_nodes.push_back(g.addNode());
    85   Graph::NodeMap<int> ref_map(g, -1);
    87 //   for (int i=0; i<b; ++i) t_nodes.push_back(g.addNode());
    86   IterableBoolMap< Graph::NodeMap<int> > bipartite_map(ref_map);
    88 
    87 
    89 //   random_init();
    88   //generating leda random group graph
    90 //   for(int i=0; i<m; ++i) {
       
    91 //     g.addEdge(s_nodes[random(a)], t_nodes[random(b)]);
       
    92 //   }
       
    93 
       
    94   Graph::NodeMap<int> ref_map(g, -1);
       
    95 
       
    96   IterableBoolMap< Graph::NodeMap<int> > bipartite_map(ref_map);
       
    97 //   for (int i=0; i<a; ++i) bipartite_map.insert(s_nodes[i], false);
       
    98 //   for (int i=0; i<b; ++i) bipartite_map.insert(t_nodes[i], true);
       
    99   leda_node ln;
    89   leda_node ln;
   100   forall(ln, lS) bipartite_map.insert(ln, false);
    90   forall(ln, lS) bipartite_map.insert(ln, false);
   101   forall(ln, lT) bipartite_map.insert(ln, true);
    91   forall(ln, lT) bipartite_map.insert(ln, true);
   102 
    92 
       
    93   //making bipartite graph
   103   typedef BipartiteGraphWrapper<Graph> BGW;
    94   typedef BipartiteGraphWrapper<Graph> BGW;
   104   BGW bgw(g, bipartite_map);
    95   BGW bgw(g, bipartite_map);
   105 
    96 
   106   //  BGW::NodeMap<int> dbyj(bgw);
       
   107   //  BGW::EdgeMap<int> dbyxcj(bgw);
       
   108 
    97 
       
    98   //st-wrapper
   109   typedef stGraphWrapper<BGW> stGW;
    99   typedef stGraphWrapper<BGW> stGW;
   110   stGW stgw(bgw);
   100   stGW stgw(bgw);
   111   ConstMap<stGW::Edge, int> const1map(1);
   101   ConstMap<stGW::Edge, int> const1map(1);
   112   stGW::EdgeMap<int> flow(stgw);
   102   stGW::EdgeMap<int> flow(stgw);
   113 
   103 
   114   Timer ts;
   104   Timer ts;
       
   105 
       
   106   ts.reset();
   115   FOR_EACH_LOC(stGW::EdgeIt, e, stgw) flow.set(e, 0);
   107   FOR_EACH_LOC(stGW::EdgeIt, e, stgw) flow.set(e, 0);
   116   ts.reset();
   108   MaxFlow<stGW, int, ConstMap<stGW::Edge, int>, stGW::EdgeMap<int> > 
   117   //  stGW::EdgeMap<int> pre_flow(stgw);
   109     max_flow_test(stgw, stgw.S_NODE, stgw.T_NODE, const1map, flow/*, true*/);
   118   Preflow<stGW, int, ConstMap<stGW::Edge, int>, stGW::EdgeMap<int> > 
   110   max_flow_test.run();
   119     pre_flow_test(stgw, stgw.S_NODE, stgw.T_NODE, const1map, flow/*, true*/);
   111   std::cout << "HUGO max matching algorithm based on preflow." << std::endl 
   120   pre_flow_test.run();
   112 	    << "Size of matching: " 
   121   std::cout << "HUGO pre flow value: " << pre_flow_test.flowValue() << std::endl;
   113 	    << max_flow_test.flowValue() << std::endl;
   122   std::cout << "elapsed time: " << ts << std::endl;
   114   std::cout << "elapsed time: " << ts << std::endl << std::endl;
   123 //   FOR_EACH_LOC(stGW::EdgeIt, e, stgw) { 
       
   124 //     std::cout << e << ": " << pre_flow[e] << "\n"; 
       
   125 //   }
       
   126   std::cout << "\n";
       
   127 
   115 
   128   ts.reset();  
   116   ts.reset();  
   129   leda_list<leda_edge> ml=MAX_CARD_BIPARTITE_MATCHING(lg);
   117   leda_list<leda_edge> ml=MAX_CARD_BIPARTITE_MATCHING(lg);
   130   //  stGW::EdgeMap<int> pre_flow(stgw);
   118   std::cout << "LEDA max matching algorithm." << std::endl 
   131   //Preflow<stGW, int, ConstMap<stGW::Edge, int>, stGW::EdgeMap<int> > 
   119 	    << "Size of matching: " 
   132   //  pre_flow_test(stgw, stgw.S_NODE, stgw.T_NODE, const1map, pre_flow, true);
   120 	    << ml.size() << std::endl;
   133   //pre_flow_test.run();
       
   134   std::cout << "LEDA matching value: " << ml.size() << std::endl;
       
   135   std::cout << "elapsed time: " << ts << std::endl;
   121   std::cout << "elapsed time: " << ts << std::endl;
   136 //   FOR_EACH_LOC(stGW::EdgeIt, e, stgw) { 
       
   137 //     std::cout << e << ": " << pre_flow[e] << "\n"; 
       
   138 //   }
       
   139   std::cout << "\n";
   122   std::cout << "\n";
   140 
   123 
       
   124   ts.reset();
   141   FOR_EACH_LOC(stGW::EdgeIt, e, stgw) flow.set(e, 0);
   125   FOR_EACH_LOC(stGW::EdgeIt, e, stgw) flow.set(e, 0);
   142   ts.reset();
       
   143   MaxFlow<stGW, int, ConstMap<stGW::Edge, int>, stGW::EdgeMap<int> > 
       
   144     max_flow_test(stgw, stgw.S_NODE, stgw.T_NODE, const1map, flow);
       
   145 //  while (max_flow_test.augmentOnShortestPath()) { }
       
   146   typedef ListGraph MutableGraph;
   126   typedef ListGraph MutableGraph;
   147 //  while (max_flow_test.augmentOnBlockingFlow1<MutableGraph>()) {
   127   while (max_flow_test.augmentOnBlockingFlow<MutableGraph>()) { }
   148   while (max_flow_test.augmentOnBlockingFlow2()) {
   128   std::cout << "HUGO max matching algorithm based on blocking flow augmentation." 
   149    std::cout << max_flow_test.flowValue() << std::endl;
   129 	    << std::endl << "Matching size: " 
   150   }
   130 	    << max_flow_test.flowValue() << std::endl;
   151   std::cout << "HUGO blocking flow value: " << max_flow_test.flowValue() << std::endl;
       
   152   std::cout << "elapsed time: " << ts << std::endl;
   131   std::cout << "elapsed time: " << ts << std::endl;
   153 //   FOR_EACH_LOC(stGW::EdgeIt, e, stgw) { 
       
   154 //     std::cout << e << ": " << max_flow[e] << "\n"; 
       
   155 //   }
       
   156 //   std::cout << "\n";
       
   157 
   132 
   158   return 0;
   133   return 0;
   159 }
   134 }