gravatar
kpeter (Peter Kovacs)
kpeter@inf.elte.hu
Support >= and <= constraints in NetworkSimplex (#219, #234) By default the same inequality constraints are supported as by Circulation (the GEQ form), but the LEQ form can also be selected using the problemType() function. The documentation of the min. cost flow module is reworked and extended with important notes and explanations about the different variants of the problem and about the dual solution and optimality conditions.
0 3 0
default
3 files changed with 377 insertions and 93 deletions:
↑ Collapse diff ↑
Ignore white space 12 line context
... ...
@@ -314,21 +314,21 @@
314 314

	
315 315
This group describes the algorithms for finding maximum flows and
316 316
feasible circulations.
317 317

	
318 318
The \e maximum \e flow \e problem is to find a flow of maximum value between
319 319
a single source and a single target. Formally, there is a \f$G=(V,A)\f$
320
digraph, a \f$cap:A\rightarrow\mathbf{R}^+_0\f$ capacity function and
320
digraph, a \f$cap: A\rightarrow\mathbf{R}^+_0\f$ capacity function and
321 321
\f$s, t \in V\f$ source and target nodes.
322
A maximum flow is an \f$f:A\rightarrow\mathbf{R}^+_0\f$ solution of the
322
A maximum flow is an \f$f: A\rightarrow\mathbf{R}^+_0\f$ solution of the
323 323
following optimization problem.
324 324

	
325
\f[ \max\sum_{a\in\delta_{out}(s)}f(a) - \sum_{a\in\delta_{in}(s)}f(a) \f]
326
\f[ \sum_{a\in\delta_{out}(v)} f(a) = \sum_{a\in\delta_{in}(v)} f(a)
327
    \qquad \forall v\in V\setminus\{s,t\} \f]
328
\f[ 0 \leq f(a) \leq cap(a) \qquad \forall a\in A \f]
325
\f[ \max\sum_{sv\in A} f(sv) - \sum_{vs\in A} f(vs) \f]
326
\f[ \sum_{uv\in A} f(uv) = \sum_{vu\in A} f(vu)
327
    \quad \forall u\in V\setminus\{s,t\} \f]
328
\f[ 0 \leq f(uv) \leq cap(uv) \quad \forall uv\in A \f]
329 329

	
330 330
LEMON contains several algorithms for solving maximum flow problems:
331 331
- \ref EdmondsKarp Edmonds-Karp algorithm.
332 332
- \ref Preflow Goldberg-Tarjan's preflow push-relabel algorithm.
333 333
- \ref DinitzSleatorTarjan Dinitz's blocking flow algorithm with dynamic trees.
334 334
- \ref GoldbergTarjan Preflow push-relabel algorithm with dynamic trees.
... ...
@@ -342,41 +342,109 @@
342 342
/**
343 343
@defgroup min_cost_flow Minimum Cost Flow Algorithms
344 344
@ingroup algs
345 345

	
346 346
\brief Algorithms for finding minimum cost flows and circulations.
347 347

	
348
This group describes the algorithms for finding minimum cost flows and
348
This group contains the algorithms for finding minimum cost flows and
349 349
circulations.
350 350

	
351 351
The \e minimum \e cost \e flow \e problem is to find a feasible flow of
352 352
minimum total cost from a set of supply nodes to a set of demand nodes
353
in a network with capacity constraints and arc costs.
353
in a network with capacity constraints (lower and upper bounds)
354
and arc costs.
354 355
Formally, let \f$G=(V,A)\f$ be a digraph,
355 356
\f$lower, upper: A\rightarrow\mathbf{Z}^+_0\f$ denote the lower and
356
upper bounds for the flow values on the arcs,
357
upper bounds for the flow values on the arcs, for which
358
\f$0 \leq lower(uv) \leq upper(uv)\f$ holds for all \f$uv\in A\f$.
357 359
\f$cost: A\rightarrow\mathbf{Z}^+_0\f$ denotes the cost per unit flow
358
on the arcs, and
359
\f$supply: V\rightarrow\mathbf{Z}\f$ denotes the supply/demand values
360
of the nodes.
361
A minimum cost flow is an \f$f:A\rightarrow\mathbf{R}^+_0\f$ solution of
362
the following optimization problem.
360
on the arcs, and \f$sup: V\rightarrow\mathbf{Z}\f$ denotes the
361
signed supply values of the nodes.
362
If \f$sup(u)>0\f$, then \f$u\f$ is a supply node with \f$sup(u)\f$
363
supply, if \f$sup(u)<0\f$, then \f$u\f$ is a demand node with
364
\f$-sup(u)\f$ demand.
365
A minimum cost flow is an \f$f: A\rightarrow\mathbf{Z}^+_0\f$ solution
366
of the following optimization problem.
363 367

	
364
\f[ \min\sum_{a\in A} f(a) cost(a) \f]
365
\f[ \sum_{a\in\delta_{out}(v)} f(a) - \sum_{a\in\delta_{in}(v)} f(a) =
366
    supply(v) \qquad \forall v\in V \f]
367
\f[ lower(a) \leq f(a) \leq upper(a) \qquad \forall a\in A \f]
368
\f[ \min\sum_{uv\in A} f(uv) \cdot cost(uv) \f]
369
\f[ \sum_{uv\in A} f(uv) - \sum_{vu\in A} f(vu) \geq
370
    sup(u) \quad \forall u\in V \f]
371
\f[ lower(uv) \leq f(uv) \leq upper(uv) \quad \forall uv\in A \f]
368 372

	
369
LEMON contains several algorithms for solving minimum cost flow problems:
370
 - \ref CycleCanceling Cycle-canceling algorithms.
371
 - \ref CapacityScaling Successive shortest path algorithm with optional
373
The sum of the supply values, i.e. \f$\sum_{u\in V} sup(u)\f$ must be
374
zero or negative in order to have a feasible solution (since the sum
375
of the expressions on the left-hand side of the inequalities is zero).
376
It means that the total demand must be greater or equal to the total
377
supply and all the supplies have to be carried out from the supply nodes,
378
but there could be demands that are not satisfied.
379
If \f$\sum_{u\in V} sup(u)\f$ is zero, then all the supply/demand
380
constraints have to be satisfied with equality, i.e. all demands
381
have to be satisfied and all supplies have to be used.
382

	
383
If you need the opposite inequalities in the supply/demand constraints
384
(i.e. the total demand is less than the total supply and all the demands
385
have to be satisfied while there could be supplies that are not used),
386
then you could easily transform the problem to the above form by reversing
387
the direction of the arcs and taking the negative of the supply values
388
(e.g. using \ref ReverseDigraph and \ref NegMap adaptors).
389
However \ref NetworkSimplex algorithm also supports this form directly
390
for the sake of convenience.
391

	
392
A feasible solution for this problem can be found using \ref Circulation.
393

	
394
Note that the above formulation is actually more general than the usual
395
definition of the minimum cost flow problem, in which strict equalities
396
are required in the supply/demand contraints, i.e.
397

	
398
\f[ \sum_{uv\in A} f(uv) - \sum_{vu\in A} f(vu) =
399
    sup(u) \quad \forall u\in V. \f]
400

	
401
However if the sum of the supply values is zero, then these two problems
402
are equivalent. So if you need the equality form, you have to ensure this
403
additional contraint for the algorithms.
404

	
405
The dual solution of the minimum cost flow problem is represented by node 
406
potentials \f$\pi: V\rightarrow\mathbf{Z}\f$.
407
An \f$f: A\rightarrow\mathbf{Z}^+_0\f$ feasible solution of the problem
408
is optimal if and only if for some \f$\pi: V\rightarrow\mathbf{Z}\f$
409
node potentials the following \e complementary \e slackness optimality
410
conditions hold.
411

	
412
 - For all \f$uv\in A\f$ arcs:
413
   - if \f$cost^\pi(uv)>0\f$, then \f$f(uv)=lower(uv)\f$;
414
   - if \f$lower(uv)<f(uv)<upper(uv)\f$, then \f$cost^\pi(uv)=0\f$;
415
   - if \f$cost^\pi(uv)<0\f$, then \f$f(uv)=upper(uv)\f$.
416
 - For all \f$u\in V\f$:
417
   - if \f$\sum_{uv\in A} f(uv) - \sum_{vu\in A} f(vu) \neq sup(u)\f$,
418
     then \f$\pi(u)=0\f$.
419
 
420
Here \f$cost^\pi(uv)\f$ denotes the \e reduced \e cost of the arc
421
\f$uv\in A\f$ with respect to the node potentials \f$\pi\f$, i.e.
422
\f[ cost^\pi(uv) = cost(uv) + \pi(u) - \pi(v).\f]
423

	
424
All algorithms provide dual solution (node potentials) as well
425
if an optimal flow is found.
426

	
427
LEMON contains several algorithms for solving minimum cost flow problems.
428
 - \ref NetworkSimplex Primal Network Simplex algorithm with various
429
   pivot strategies.
430
 - \ref CostScaling Push-Relabel and Augment-Relabel algorithms based on
431
   cost scaling.
432
 - \ref CapacityScaling Successive Shortest %Path algorithm with optional
372 433
   capacity scaling.
373
 - \ref CostScaling Push-relabel and augment-relabel algorithms based on
374
   cost scaling.
375
 - \ref NetworkSimplex Primal network simplex algorithm with various
376
   pivot strategies.
434
 - \ref CancelAndTighten The Cancel and Tighten algorithm.
435
 - \ref CycleCanceling Cycle-Canceling algorithms.
436

	
437
Most of these implementations support the general inequality form of the
438
minimum cost flow problem, but CancelAndTighten and CycleCanceling
439
only support the equality form due to the primal method they use.
440

	
441
In general NetworkSimplex is the most efficient implementation,
442
but in special cases other algorithms could be faster.
443
For example, if the total supply and/or capacities are rather small,
444
CapacityScaling is usually the fastest algorithm (without effective scaling).
377 445
*/
378 446

	
379 447
/**
380 448
@defgroup min_cut Minimum Cut Algorithms
381 449
@ingroup algs
382 450

	
Ignore white space 6 line context
... ...
@@ -27,12 +27,15 @@
27 27
#include <vector>
28 28
#include <limits>
29 29
#include <algorithm>
30 30

	
31 31
#include <lemon/core.h>
32 32
#include <lemon/math.h>
33
#include <lemon/maps.h>
34
#include <lemon/circulation.h>
35
#include <lemon/adaptors.h>
33 36

	
34 37
namespace lemon {
35 38

	
36 39
  /// \addtogroup min_cost_flow
37 40
  /// @{
38 41

	
... ...
@@ -44,37 +47,47 @@
44 47
  /// This algorithm is a specialized version of the linear programming
45 48
  /// simplex method directly for the minimum cost flow problem.
46 49
  /// It is one of the most efficient solution methods.
47 50
  ///
48 51
  /// In general this class is the fastest implementation available
49 52
  /// in LEMON for the minimum cost flow problem.
53
  /// Moreover it supports both direction of the supply/demand inequality
54
  /// constraints. For more information see \ref ProblemType.
50 55
  ///
51 56
  /// \tparam GR The digraph type the algorithm runs on.
52 57
  /// \tparam F The value type used for flow amounts, capacity bounds
53 58
  /// and supply values in the algorithm. By default it is \c int.
54 59
  /// \tparam C The value type used for costs and potentials in the
55 60
  /// algorithm. By default it is the same as \c F.
56 61
  ///
57 62
  /// \warning Both value types must be signed and all input data must
58 63
  /// be integer.
59 64
  ///
60 65
  /// \note %NetworkSimplex provides five different pivot rule
61
  /// implementations. For more information see \ref PivotRule.
66
  /// implementations, from which the most efficient one is used
67
  /// by default. For more information see \ref PivotRule.
62 68
  template <typename GR, typename F = int, typename C = F>
63 69
  class NetworkSimplex
64 70
  {
65 71
  public:
66 72

	
67 73
    /// The flow type of the algorithm
68 74
    typedef F Flow;
69 75
    /// The cost type of the algorithm
70 76
    typedef C Cost;
77
#ifdef DOXYGEN
78
    /// The type of the flow map
79
    typedef GR::ArcMap<Flow> FlowMap;
80
    /// The type of the potential map
81
    typedef GR::NodeMap<Cost> PotentialMap;
82
#else
71 83
    /// The type of the flow map
72 84
    typedef typename GR::template ArcMap<Flow> FlowMap;
73 85
    /// The type of the potential map
74 86
    typedef typename GR::template NodeMap<Cost> PotentialMap;
87
#endif
75 88

	
76 89
  public:
77 90

	
78 91
    /// \brief Enum type for selecting the pivot rule.
79 92
    ///
80 93
    /// Enum type for selecting the pivot rule for the \ref run()
... ...
@@ -114,12 +127,64 @@
114 127
      /// The Altering Candidate List pivot rule.
115 128
      /// It is a modified version of the Candidate List method.
116 129
      /// It keeps only the several best eligible arcs from the former
117 130
      /// candidate list and extends this list in every iteration.
118 131
      ALTERING_LIST
119 132
    };
133
    
134
    /// \brief Enum type for selecting the problem type.
135
    ///
136
    /// Enum type for selecting the problem type, i.e. the direction of
137
    /// the inequalities in the supply/demand constraints of the
138
    /// \ref min_cost_flow "minimum cost flow problem".
139
    ///
140
    /// The default problem type is \c GEQ, since this form is supported
141
    /// by other minimum cost flow algorithms and the \ref Circulation
142
    /// algorithm as well.
143
    /// The \c LEQ problem type can be selected using the \ref problemType()
144
    /// function.
145
    ///
146
    /// Note that the equality form is a special case of both problem type.
147
    enum ProblemType {
148

	
149
      /// This option means that there are "<em>greater or equal</em>"
150
      /// constraints in the defintion, i.e. the exact formulation of the
151
      /// problem is the following.
152
      /**
153
          \f[ \min\sum_{uv\in A} f(uv) \cdot cost(uv) \f]
154
          \f[ \sum_{uv\in A} f(uv) - \sum_{vu\in A} f(vu) \geq
155
              sup(u) \quad \forall u\in V \f]
156
          \f[ lower(uv) \leq f(uv) \leq upper(uv) \quad \forall uv\in A \f]
157
      */
158
      /// It means that the total demand must be greater or equal to the 
159
      /// total supply (i.e. \f$\sum_{u\in V} sup(u)\f$ must be zero or
160
      /// negative) and all the supplies have to be carried out from 
161
      /// the supply nodes, but there could be demands that are not 
162
      /// satisfied.
163
      GEQ,
164
      /// It is just an alias for the \c GEQ option.
165
      CARRY_SUPPLIES = GEQ,
166

	
167
      /// This option means that there are "<em>less or equal</em>"
168
      /// constraints in the defintion, i.e. the exact formulation of the
169
      /// problem is the following.
170
      /**
171
          \f[ \min\sum_{uv\in A} f(uv) \cdot cost(uv) \f]
172
          \f[ \sum_{uv\in A} f(uv) - \sum_{vu\in A} f(vu) \leq
173
              sup(u) \quad \forall u\in V \f]
174
          \f[ lower(uv) \leq f(uv) \leq upper(uv) \quad \forall uv\in A \f]
175
      */
176
      /// It means that the total demand must be less or equal to the 
177
      /// total supply (i.e. \f$\sum_{u\in V} sup(u)\f$ must be zero or
178
      /// positive) and all the demands have to be satisfied, but there
179
      /// could be supplies that are not carried out from the supply
180
      /// nodes.
181
      LEQ,
182
      /// It is just an alias for the \c LEQ option.
183
      SATISFY_DEMANDS = LEQ
184
    };
120 185

	
121 186
  private:
122 187

	
123 188
    TEMPLATE_DIGRAPH_TYPEDEFS(GR);
124 189

	
125 190
    typedef typename GR::template ArcMap<Flow> FlowArcMap;
... ...
@@ -152,12 +217,13 @@
152 217
    FlowArcMap *_pupper;
153 218
    CostArcMap *_pcost;
154 219
    FlowNodeMap *_psupply;
155 220
    bool _pstsup;
156 221
    Node _psource, _ptarget;
157 222
    Flow _pstflow;
223
    ProblemType _ptype;
158 224

	
159 225
    // Result maps
160 226
    FlowMap *_flow_map;
161 227
    PotentialMap *_potential_map;
162 228
    bool _local_flow;
163 229
    bool _local_potential;
... ...
@@ -583,19 +649,19 @@
583 649
    }; //class AlteringListPivotRule
584 650

	
585 651
  public:
586 652

	
587 653
    /// \brief Constructor.
588 654
    ///
589
    /// Constructor.
655
    /// The constructor of the class.
590 656
    ///
591 657
    /// \param graph The digraph the algorithm runs on.
592 658
    NetworkSimplex(const GR& graph) :
593 659
      _graph(graph),
594 660
      _plower(NULL), _pupper(NULL), _pcost(NULL),
595
      _psupply(NULL), _pstsup(false),
661
      _psupply(NULL), _pstsup(false), _ptype(GEQ),
596 662
      _flow_map(NULL), _potential_map(NULL),
597 663
      _local_flow(false), _local_potential(false),
598 664
      _node_id(graph)
599 665
    {
600 666
      LEMON_ASSERT(std::numeric_limits<Flow>::is_integer &&
601 667
                   std::numeric_limits<Flow>::is_signed,
... ...
@@ -608,12 +674,18 @@
608 674
    /// Destructor.
609 675
    ~NetworkSimplex() {
610 676
      if (_local_flow) delete _flow_map;
611 677
      if (_local_potential) delete _potential_map;
612 678
    }
613 679

	
680
    /// \name Parameters
681
    /// The parameters of the algorithm can be specified using these
682
    /// functions.
683

	
684
    /// @{
685

	
614 686
    /// \brief Set the lower bounds on the arcs.
615 687
    ///
616 688
    /// This function sets the lower bounds on the arcs.
617 689
    /// If neither this function nor \ref boundMaps() is used before
618 690
    /// calling \ref run(), the lower bounds will be set to zero
619 691
    /// on all arcs.
... ...
@@ -757,12 +829,26 @@
757 829
      _pstsup = true;
758 830
      _psource = s;
759 831
      _ptarget = t;
760 832
      _pstflow = k;
761 833
      return *this;
762 834
    }
835
    
836
    /// \brief Set the problem type.
837
    ///
838
    /// This function sets the problem type for the algorithm.
839
    /// If it is not used before calling \ref run(), the \ref GEQ problem
840
    /// type will be used.
841
    ///
842
    /// For more information see \ref ProblemType.
843
    ///
844
    /// \return <tt>(*this)</tt>
845
    NetworkSimplex& problemType(ProblemType problem_type) {
846
      _ptype = problem_type;
847
      return *this;
848
    }
763 849

	
764 850
    /// \brief Set the flow map.
765 851
    ///
766 852
    /// This function sets the flow map.
767 853
    /// If it is not used before calling \ref run(), an instance will
768 854
    /// be allocated automatically. The destructor deallocates this
... ...
@@ -792,25 +878,28 @@
792 878
        delete _potential_map;
793 879
        _local_potential = false;
794 880
      }
795 881
      _potential_map = &map;
796 882
      return *this;
797 883
    }
884
    
885
    /// @}
798 886

	
799 887
    /// \name Execution Control
800 888
    /// The algorithm can be executed using \ref run().
801 889

	
802 890
    /// @{
803 891

	
804 892
    /// \brief Run the algorithm.
805 893
    ///
806 894
    /// This function runs the algorithm.
807
    /// The paramters can be specified using \ref lowerMap(),
895
    /// The paramters can be specified using functions \ref lowerMap(),
808 896
    /// \ref upperMap(), \ref capacityMap(), \ref boundMaps(),
809
    /// \ref costMap(), \ref supplyMap() and \ref stSupply()
810
    /// functions. For example,
897
    /// \ref costMap(), \ref supplyMap(), \ref stSupply(), 
898
    /// \ref problemType(), \ref flowMap() and \ref potentialMap().
899
    /// For example,
811 900
    /// \code
812 901
    ///   NetworkSimplex<ListDigraph> ns(graph);
813 902
    ///   ns.boundMaps(lower, upper).costMap(cost)
814 903
    ///     .supplyMap(sup).run();
815 904
    /// \endcode
816 905
    ///
... ...
@@ -827,15 +916,16 @@
827 916
      return init() && start(pivot_rule);
828 917
    }
829 918

	
830 919
    /// \brief Reset all the parameters that have been given before.
831 920
    ///
832 921
    /// This function resets all the paramaters that have been given
833
    /// using \ref lowerMap(), \ref upperMap(), \ref capacityMap(),
834
    /// \ref boundMaps(), \ref costMap(), \ref supplyMap() and
835
    /// \ref stSupply() functions before.
922
    /// before using functions \ref lowerMap(), \ref upperMap(),
923
    /// \ref capacityMap(), \ref boundMaps(), \ref costMap(),
924
    /// \ref supplyMap(), \ref stSupply(), \ref problemType(), 
925
    /// \ref flowMap() and \ref potentialMap().
836 926
    ///
837 927
    /// It is useful for multiple run() calls. If this function is not
838 928
    /// used, all the parameters given before are kept for the next
839 929
    /// \ref run() call.
840 930
    ///
841 931
    /// For example,
... ...
@@ -866,12 +956,20 @@
866 956
      delete _psupply;
867 957
      _plower = NULL;
868 958
      _pupper = NULL;
869 959
      _pcost = NULL;
870 960
      _psupply = NULL;
871 961
      _pstsup = false;
962
      _ptype = GEQ;
963
      if (_local_flow) delete _flow_map;
964
      if (_local_potential) delete _potential_map;
965
      _flow_map = NULL;
966
      _potential_map = NULL;
967
      _local_flow = false;
968
      _local_potential = false;
969

	
872 970
      return *this;
873 971
    }
874 972

	
875 973
    /// @}
876 974

	
877 975
    /// \name Query Functions
... ...
@@ -997,61 +1095,147 @@
997 1095
      _succ_num.resize(all_node_num);
998 1096
      _last_succ.resize(all_node_num);
999 1097
      _state.resize(all_arc_num);
1000 1098

	
1001 1099
      // Initialize node related data
1002 1100
      bool valid_supply = true;
1101
      Flow sum_supply = 0;
1003 1102
      if (!_pstsup && !_psupply) {
1004 1103
        _pstsup = true;
1005 1104
        _psource = _ptarget = NodeIt(_graph);
1006 1105
        _pstflow = 0;
1007 1106
      }
1008 1107
      if (_psupply) {
1009
        Flow sum = 0;
1010 1108
        int i = 0;
1011 1109
        for (NodeIt n(_graph); n != INVALID; ++n, ++i) {
1012 1110
          _node_id[n] = i;
1013 1111
          _supply[i] = (*_psupply)[n];
1014
          sum += _supply[i];
1112
          sum_supply += _supply[i];
1015 1113
        }
1016
        valid_supply = (sum == 0);
1114
        valid_supply = (_ptype == GEQ && sum_supply <= 0) ||
1115
                       (_ptype == LEQ && sum_supply >= 0);
1017 1116
      } else {
1018 1117
        int i = 0;
1019 1118
        for (NodeIt n(_graph); n != INVALID; ++n, ++i) {
1020 1119
          _node_id[n] = i;
1021 1120
          _supply[i] = 0;
1022 1121
        }
1023 1122
        _supply[_node_id[_psource]] =  _pstflow;
1024
        _supply[_node_id[_ptarget]]   = -_pstflow;
1123
        _supply[_node_id[_ptarget]] = -_pstflow;
1025 1124
      }
1026 1125
      if (!valid_supply) return false;
1027 1126

	
1127
      // Infinite capacity value
1128
      Flow inf_cap =
1129
        std::numeric_limits<Flow>::has_infinity ?
1130
        std::numeric_limits<Flow>::infinity() :
1131
        std::numeric_limits<Flow>::max();
1132

	
1133
      // Initialize artifical cost
1134
      Cost art_cost;
1135
      if (std::numeric_limits<Cost>::is_exact) {
1136
        art_cost = std::numeric_limits<Cost>::max() / 4 + 1;
1137
      } else {
1138
        art_cost = std::numeric_limits<Cost>::min();
1139
        for (int i = 0; i != _arc_num; ++i) {
1140
          if (_cost[i] > art_cost) art_cost = _cost[i];
1141
        }
1142
        art_cost = (art_cost + 1) * _node_num;
1143
      }
1144

	
1145
      // Run Circulation to check if a feasible solution exists
1146
      typedef ConstMap<Arc, Flow> ConstArcMap;
1147
      FlowNodeMap *csup = NULL;
1148
      bool local_csup = false;
1149
      if (_psupply) {
1150
        csup = _psupply;
1151
      } else {
1152
        csup = new FlowNodeMap(_graph, 0);
1153
        (*csup)[_psource] =  _pstflow;
1154
        (*csup)[_ptarget] = -_pstflow;
1155
        local_csup = true;
1156
      }
1157
      bool circ_result = false;
1158
      if (_ptype == GEQ || (_ptype == LEQ && sum_supply == 0)) {
1159
        // GEQ problem type
1160
        if (_plower) {
1161
          if (_pupper) {
1162
            Circulation<GR, FlowArcMap, FlowArcMap, FlowNodeMap>
1163
              circ(_graph, *_plower, *_pupper, *csup);
1164
            circ_result = circ.run();
1165
          } else {
1166
            Circulation<GR, FlowArcMap, ConstArcMap, FlowNodeMap>
1167
              circ(_graph, *_plower, ConstArcMap(inf_cap), *csup);
1168
            circ_result = circ.run();
1169
          }
1170
        } else {
1171
          if (_pupper) {
1172
            Circulation<GR, ConstArcMap, FlowArcMap, FlowNodeMap>
1173
              circ(_graph, ConstArcMap(0), *_pupper, *csup);
1174
            circ_result = circ.run();
1175
          } else {
1176
            Circulation<GR, ConstArcMap, ConstArcMap, FlowNodeMap>
1177
              circ(_graph, ConstArcMap(0), ConstArcMap(inf_cap), *csup);
1178
            circ_result = circ.run();
1179
          }
1180
        }
1181
      } else {
1182
        // LEQ problem type
1183
        typedef ReverseDigraph<const GR> RevGraph;
1184
        typedef NegMap<FlowNodeMap> NegNodeMap;
1185
        RevGraph rgraph(_graph);
1186
        NegNodeMap neg_csup(*csup);
1187
        if (_plower) {
1188
          if (_pupper) {
1189
            Circulation<RevGraph, FlowArcMap, FlowArcMap, NegNodeMap>
1190
              circ(rgraph, *_plower, *_pupper, neg_csup);
1191
            circ_result = circ.run();
1192
          } else {
1193
            Circulation<RevGraph, FlowArcMap, ConstArcMap, NegNodeMap>
1194
              circ(rgraph, *_plower, ConstArcMap(inf_cap), neg_csup);
1195
            circ_result = circ.run();
1196
          }
1197
        } else {
1198
          if (_pupper) {
1199
            Circulation<RevGraph, ConstArcMap, FlowArcMap, NegNodeMap>
1200
              circ(rgraph, ConstArcMap(0), *_pupper, neg_csup);
1201
            circ_result = circ.run();
1202
          } else {
1203
            Circulation<RevGraph, ConstArcMap, ConstArcMap, NegNodeMap>
1204
              circ(rgraph, ConstArcMap(0), ConstArcMap(inf_cap), neg_csup);
1205
            circ_result = circ.run();
1206
          }
1207
        }
1208
      }
1209
      if (local_csup) delete csup;
1210
      if (!circ_result) return false;
1211

	
1028 1212
      // Set data for the artificial root node
1029 1213
      _root = _node_num;
1030 1214
      _parent[_root] = -1;
1031 1215
      _pred[_root] = -1;
1032 1216
      _thread[_root] = 0;
1033 1217
      _rev_thread[0] = _root;
1034 1218
      _succ_num[_root] = all_node_num;
1035 1219
      _last_succ[_root] = _root - 1;
1036
      _supply[_root] = 0;
1037
      _pi[_root] = 0;
1220
      _supply[_root] = -sum_supply;
1221
      if (sum_supply < 0) {
1222
        _pi[_root] = -art_cost;
1223
      } else {
1224
        _pi[_root] = art_cost;
1225
      }
1038 1226

	
1039 1227
      // Store the arcs in a mixed order
1040 1228
      int k = std::max(int(sqrt(_arc_num)), 10);
1041 1229
      int i = 0;
1042 1230
      for (ArcIt e(_graph); e != INVALID; ++e) {
1043 1231
        _arc_ref[i] = e;
1044 1232
        if ((i += k) >= _arc_num) i = (i % k) + 1;
1045 1233
      }
1046 1234

	
1047 1235
      // Initialize arc maps
1048
      Flow inf_cap =
1049
        std::numeric_limits<Flow>::has_infinity ?
1050
        std::numeric_limits<Flow>::infinity() :
1051
        std::numeric_limits<Flow>::max();
1052 1236
      if (_pupper && _pcost) {
1053 1237
        for (int i = 0; i != _arc_num; ++i) {
1054 1238
          Arc e = _arc_ref[i];
1055 1239
          _source[i] = _node_id[_graph.source(e)];
1056 1240
          _target[i] = _node_id[_graph.target(e)];
1057 1241
          _cap[i] = (*_pupper)[e];
... ...
@@ -1080,24 +1264,12 @@
1080 1264
        } else {
1081 1265
          for (int i = 0; i != _arc_num; ++i)
1082 1266
            _cost[i] = 1;
1083 1267
        }
1084 1268
      }
1085 1269
      
1086
      // Initialize artifical cost
1087
      Cost art_cost;
1088
      if (std::numeric_limits<Cost>::is_exact) {
1089
        art_cost = std::numeric_limits<Cost>::max() / 4 + 1;
1090
      } else {
1091
        art_cost = std::numeric_limits<Cost>::min();
1092
        for (int i = 0; i != _arc_num; ++i) {
1093
          if (_cost[i] > art_cost) art_cost = _cost[i];
1094
        }
1095
        art_cost = (art_cost + 1) * _node_num;
1096
      }
1097

	
1098 1270
      // Remove non-zero lower bounds
1099 1271
      if (_plower) {
1100 1272
        for (int i = 0; i != _arc_num; ++i) {
1101 1273
          Flow c = (*_plower)[_arc_ref[i]];
1102 1274
          if (c != 0) {
1103 1275
            _cap[i] -= c;
... ...
@@ -1115,20 +1287,20 @@
1115 1287
        _last_succ[u] = u;
1116 1288
        _parent[u] = _root;
1117 1289
        _pred[u] = e;
1118 1290
        _cost[e] = art_cost;
1119 1291
        _cap[e] = inf_cap;
1120 1292
        _state[e] = STATE_TREE;
1121
        if (_supply[u] >= 0) {
1293
        if (_supply[u] > 0 || (_supply[u] == 0 && sum_supply <= 0)) {
1122 1294
          _flow[e] = _supply[u];
1123 1295
          _forward[u] = true;
1124
          _pi[u] = -art_cost;
1296
          _pi[u] = -art_cost + _pi[_root];
1125 1297
        } else {
1126 1298
          _flow[e] = -_supply[u];
1127 1299
          _forward[u] = false;
1128
          _pi[u] = art_cost;
1300
          _pi[u] = art_cost + _pi[_root];
1129 1301
        }
1130 1302
      }
1131 1303

	
1132 1304
      return true;
1133 1305
    }
1134 1306

	
... ...
@@ -1379,17 +1551,12 @@
1379 1551
        if (change) {
1380 1552
          updateTreeStructure();
1381 1553
          updatePotential();
1382 1554
        }
1383 1555
      }
1384 1556

	
1385
      // Check if the flow amount equals zero on all the artificial arcs
1386
      for (int e = _arc_num; e != _arc_num + _node_num; ++e) {
1387
        if (_flow[e] > 0) return false;
1388
      }
1389

	
1390 1557
      // Copy flow values to _flow_map
1391 1558
      if (_plower) {
1392 1559
        for (int i = 0; i != _arc_num; ++i) {
1393 1560
          Arc e = _arc_ref[i];
1394 1561
          _flow_map->set(e, (*_plower)[e] + _flow[i]);
1395 1562
        }
Ignore white space 6 line context
... ...
@@ -30,25 +30,25 @@
30 30
#include "test_tools.h"
31 31

	
32 32
using namespace lemon;
33 33

	
34 34
char test_lgf[] =
35 35
  "@nodes\n"
36
  "label  sup1 sup2 sup3\n"
37
  "    1    20   27    0\n"
38
  "    2    -4    0    0\n"
39
  "    3     0    0    0\n"
40
  "    4     0    0    0\n"
41
  "    5     9    0    0\n"
42
  "    6    -6    0    0\n"
43
  "    7     0    0    0\n"
44
  "    8     0    0    0\n"
45
  "    9     3    0    0\n"
46
  "   10    -2    0    0\n"
47
  "   11     0    0    0\n"
48
  "   12   -20  -27    0\n"
36
  "label  sup1 sup2 sup3 sup4 sup5\n"
37
  "    1    20   27    0   20   30\n"
38
  "    2    -4    0    0   -8   -3\n"
39
  "    3     0    0    0    0    0\n"
40
  "    4     0    0    0    0    0\n"
41
  "    5     9    0    0    6   11\n"
42
  "    6    -6    0    0   -5   -6\n"
43
  "    7     0    0    0    0    0\n"
44
  "    8     0    0    0    0    3\n"
45
  "    9     3    0    0    0    0\n"
46
  "   10    -2    0    0   -7   -2\n"
47
  "   11     0    0    0  -10    0\n"
48
  "   12   -20  -27    0  -30  -20\n"
49 49
  "\n"
50 50
  "@arcs\n"
51 51
  "       cost  cap low1 low2\n"
52 52
  " 1  2    70   11    0    8\n"
53 53
  " 1  3   150    3    0    1\n"
54 54
  " 1  4    80   15    0    2\n"
... ...
@@ -73,12 +73,18 @@
73 73
  "\n"
74 74
  "@attributes\n"
75 75
  "source 1\n"
76 76
  "target 12\n";
77 77

	
78 78

	
79
enum ProblemType {
80
  EQ,
81
  GEQ,
82
  LEQ
83
};
84

	
79 85
// Check the interface of an MCF algorithm
80 86
template <typename GR, typename Flow, typename Cost>
81 87
class McfClassConcept
82 88
{
83 89
public:
84 90

	
... ...
@@ -94,23 +100,25 @@
94 100
             .upperMap(upper)
95 101
             .capacityMap(upper)
96 102
             .boundMaps(lower, upper)
97 103
             .costMap(cost)
98 104
             .supplyMap(sup)
99 105
             .stSupply(n, n, k)
106
             .flowMap(flow)
107
             .potentialMap(pot)
100 108
             .run();
109
      
110
      const MCF& const_mcf = mcf;
101 111

	
102
      const typename MCF::FlowMap &fm = mcf.flowMap();
103
      const typename MCF::PotentialMap &pm = mcf.potentialMap();
112
      const typename MCF::FlowMap &fm = const_mcf.flowMap();
113
      const typename MCF::PotentialMap &pm = const_mcf.potentialMap();
104 114

	
105
      v = mcf.totalCost();
106
      double x = mcf.template totalCost<double>();
107
      v = mcf.flow(a);
108
      v = mcf.potential(n);
109
      mcf.flowMap(flow);
110
      mcf.potentialMap(pot);
115
      v = const_mcf.totalCost();
116
      double x = const_mcf.template totalCost<double>();
117
      v = const_mcf.flow(a);
118
      v = const_mcf.potential(n);
111 119

	
112 120
      ignore_unused_variable_warning(fm);
113 121
      ignore_unused_variable_warning(pm);
114 122
      ignore_unused_variable_warning(x);
115 123
    }
116 124

	
... ...
@@ -139,13 +147,14 @@
139 147

	
140 148

	
141 149
// Check the feasibility of the given flow (primal soluiton)
142 150
template < typename GR, typename LM, typename UM,
143 151
           typename SM, typename FM >
144 152
bool checkFlow( const GR& gr, const LM& lower, const UM& upper,
145
                const SM& supply, const FM& flow )
153
                const SM& supply, const FM& flow,
154
                ProblemType type = EQ )
146 155
{
147 156
  TEMPLATE_DIGRAPH_TYPEDEFS(GR);
148 157

	
149 158
  for (ArcIt e(gr); e != INVALID; ++e) {
150 159
    if (flow[e] < lower[e] || flow[e] > upper[e]) return false;
151 160
  }
... ...
@@ -153,54 +162,69 @@
153 162
  for (NodeIt n(gr); n != INVALID; ++n) {
154 163
    typename SM::Value sum = 0;
155 164
    for (OutArcIt e(gr, n); e != INVALID; ++e)
156 165
      sum += flow[e];
157 166
    for (InArcIt e(gr, n); e != INVALID; ++e)
158 167
      sum -= flow[e];
159
    if (sum != supply[n]) return false;
168
    bool b = (type ==  EQ && sum == supply[n]) ||
169
             (type == GEQ && sum >= supply[n]) ||
170
             (type == LEQ && sum <= supply[n]);
171
    if (!b) return false;
160 172
  }
161 173

	
162 174
  return true;
163 175
}
164 176

	
165 177
// Check the feasibility of the given potentials (dual soluiton)
166 178
// using the "Complementary Slackness" optimality condition
167 179
template < typename GR, typename LM, typename UM,
168
           typename CM, typename FM, typename PM >
180
           typename CM, typename SM, typename FM, typename PM >
169 181
bool checkPotential( const GR& gr, const LM& lower, const UM& upper,
170
                     const CM& cost, const FM& flow, const PM& pi )
182
                     const CM& cost, const SM& supply, const FM& flow, 
183
                     const PM& pi )
171 184
{
172 185
  TEMPLATE_DIGRAPH_TYPEDEFS(GR);
173 186

	
174 187
  bool opt = true;
175 188
  for (ArcIt e(gr); opt && e != INVALID; ++e) {
176 189
    typename CM::Value red_cost =
177 190
      cost[e] + pi[gr.source(e)] - pi[gr.target(e)];
178 191
    opt = red_cost == 0 ||
179 192
          (red_cost > 0 && flow[e] == lower[e]) ||
180 193
          (red_cost < 0 && flow[e] == upper[e]);
181 194
  }
195
  
196
  for (NodeIt n(gr); opt && n != INVALID; ++n) {
197
    typename SM::Value sum = 0;
198
    for (OutArcIt e(gr, n); e != INVALID; ++e)
199
      sum += flow[e];
200
    for (InArcIt e(gr, n); e != INVALID; ++e)
201
      sum -= flow[e];
202
    opt = (sum == supply[n]) || (pi[n] == 0);
203
  }
204
  
182 205
  return opt;
183 206
}
184 207

	
185 208
// Run a minimum cost flow algorithm and check the results
186 209
template < typename MCF, typename GR,
187 210
           typename LM, typename UM,
188 211
           typename CM, typename SM >
189 212
void checkMcf( const MCF& mcf, bool mcf_result,
190 213
               const GR& gr, const LM& lower, const UM& upper,
191 214
               const CM& cost, const SM& supply,
192 215
               bool result, typename CM::Value total,
193
               const std::string &test_id = "" )
216
               const std::string &test_id = "",
217
               ProblemType type = EQ )
194 218
{
195 219
  check(mcf_result == result, "Wrong result " + test_id);
196 220
  if (result) {
197
    check(checkFlow(gr, lower, upper, supply, mcf.flowMap()),
221
    check(checkFlow(gr, lower, upper, supply, mcf.flowMap(), type),
198 222
          "The flow is not feasible " + test_id);
199 223
    check(mcf.totalCost() == total, "The flow is not optimal " + test_id);
200
    check(checkPotential(gr, lower, upper, cost, mcf.flowMap(),
224
    check(checkPotential(gr, lower, upper, cost, supply, mcf.flowMap(),
201 225
                         mcf.potentialMap()),
202 226
          "Wrong potentials " + test_id);
203 227
  }
204 228
}
205 229

	
206 230
int main()
... ...
@@ -223,33 +247,36 @@
223 247
  typedef ListDigraph Digraph;
224 248
  DIGRAPH_TYPEDEFS(ListDigraph);
225 249

	
226 250
  // Read the test digraph
227 251
  Digraph gr;
228 252
  Digraph::ArcMap<int> c(gr), l1(gr), l2(gr), u(gr);
229
  Digraph::NodeMap<int> s1(gr), s2(gr), s3(gr);
253
  Digraph::NodeMap<int> s1(gr), s2(gr), s3(gr), s4(gr), s5(gr);
230 254
  ConstMap<Arc, int> cc(1), cu(std::numeric_limits<int>::max());
231 255
  Node v, w;
232 256

	
233 257
  std::istringstream input(test_lgf);
234 258
  DigraphReader<Digraph>(gr, input)
235 259
    .arcMap("cost", c)
236 260
    .arcMap("cap", u)
237 261
    .arcMap("low1", l1)
238 262
    .arcMap("low2", l2)
239 263
    .nodeMap("sup1", s1)
240 264
    .nodeMap("sup2", s2)
241 265
    .nodeMap("sup3", s3)
266
    .nodeMap("sup4", s4)
267
    .nodeMap("sup5", s5)
242 268
    .node("source", v)
243 269
    .node("target", w)
244 270
    .run();
245 271

	
246 272
  // A. Test NetworkSimplex with the default pivot rule
247 273
  {
248 274
    NetworkSimplex<Digraph> mcf(gr);
249 275

	
276
    // Check the equality form
250 277
    mcf.upperMap(u).costMap(c);
251 278
    checkMcf(mcf, mcf.supplyMap(s1).run(),
252 279
             gr, l1, u, c, s1, true,  5240, "#A1");
253 280
    checkMcf(mcf, mcf.stSupply(v, w, 27).run(),
254 281
             gr, l1, u, c, s2, true,  7620, "#A2");
255 282
    mcf.lowerMap(l2);
... ...
@@ -264,12 +291,34 @@
264 291
             gr, l2, cu, cc, s2, true,  94, "#A6");
265 292
    mcf.reset();
266 293
    checkMcf(mcf, mcf.run(),
267 294
             gr, l1, cu, cc, s3, true,   0, "#A7");
268 295
    checkMcf(mcf, mcf.boundMaps(l2, u).run(),
269 296
             gr, l2, u, cc, s3, false,   0, "#A8");
297

	
298
    // Check the GEQ form
299
    mcf.reset().upperMap(u).costMap(c).supplyMap(s4);
300
    checkMcf(mcf, mcf.run(),
301
             gr, l1, u, c, s4, true,  3530, "#A9", GEQ);
302
    mcf.problemType(mcf.GEQ);
303
    checkMcf(mcf, mcf.lowerMap(l2).run(),
304
             gr, l2, u, c, s4, true,  4540, "#A10", GEQ);
305
    mcf.problemType(mcf.CARRY_SUPPLIES).supplyMap(s5);
306
    checkMcf(mcf, mcf.run(),
307
             gr, l2, u, c, s5, false,    0, "#A11", GEQ);
308

	
309
    // Check the LEQ form
310
    mcf.reset().problemType(mcf.LEQ);
311
    mcf.upperMap(u).costMap(c).supplyMap(s5);
312
    checkMcf(mcf, mcf.run(),
313
             gr, l1, u, c, s5, true,  5080, "#A12", LEQ);
314
    checkMcf(mcf, mcf.lowerMap(l2).run(),
315
             gr, l2, u, c, s5, true,  5930, "#A13", LEQ);
316
    mcf.problemType(mcf.SATISFY_DEMANDS).supplyMap(s4);
317
    checkMcf(mcf, mcf.run(),
318
             gr, l2, u, c, s4, false,    0, "#A14", LEQ);
270 319
  }
271 320

	
272 321
  // B. Test NetworkSimplex with each pivot rule
273 322
  {
274 323
    NetworkSimplex<Digraph> mcf(gr);
275 324
    mcf.supplyMap(s1).costMap(c).capacityMap(u).lowerMap(l2);
0 comments (0 inline)