Collection of objects used to extend the Constraint Solver library. More...

Classes
class	ArgumentHolder
	Argument Holder: useful when visiting a model. More...

class	ArrayWithOffset

class	Assignment
	An Assignment is a variable -> domains mapping, used to report solutions to the user. More...

class	AssignmentContainer

class	AssignmentElement

class	BaseIntExpr

class	BaseLns
	This is the base class for building an Lns operator. More...

class	BaseObject
	A BaseObject is the root of all reversibly allocated objects. More...

class	BooleanVar

class	CallMethod0
	Demon proxy to a method on the constraint with no arguments. More...

class	CallMethod1
	Demon proxy to a method on the constraint with one argument. More...

class	CallMethod2
	Demon proxy to a method on the constraint with two arguments. More...

class	CallMethod3
	Demon proxy to a method on the constraint with three arguments. More...

class	CastConstraint
	Cast constraints are special channeling constraints designed to keep a variable in sync with an expression. More...

class	ChangeValue
	Defines operators which change the value of variables; each neighbor corresponds to one modified variable. More...

class	Constraint
	A constraint is the main modeling object. More...

class	CumulBoundsPropagator

class	Decision
	A Decision represents a choice point in the search tree. More...

class	DecisionBuilder
	A DecisionBuilder is responsible for creating the search tree. More...

class	DecisionVisitor
	A DecisionVisitor is used to inspect a decision. More...

struct	DefaultPhaseParameters
	This struct holds all parameters for the default search. More...

class	DelayedCallMethod0
	Low-priority demon proxy to a method on the constraint with no arguments. More...

class	DelayedCallMethod1
	Low-priority demon proxy to a method on the constraint with one argument. More...

class	DelayedCallMethod2
	Low-priority demon proxy to a method on the constraint with two arguments. More...

class	Demon
	A Demon is the base element of a propagation queue. More...

class	DimensionCumulOptimizerCore

class	DisjunctiveConstraint

class	DisjunctivePropagator
	This class acts like a CP propagator: it takes a set of tasks given by their start/duration/end features, and reduces the range of possible values. More...

class	ExchangeSubtrip

class	FilteredHeuristicCloseNodesLNSOperator
	Filtered heuristic LNS operator, where the destruction phase consists of removing a node and the 'num_close_nodes' nodes closest to it, along with each of their corresponding sibling pickup/deliveries that are performed. More...

class	FilteredHeuristicExpensiveChainLNSOperator
	Similar to the heuristic path LNS above, but instead of removing one route entirely, the destruction phase consists of removing all nodes on an "expensive" chain from a route. More...

class	FilteredHeuristicLocalSearchOperator
	Class of operators using a RoutingFilteredHeuristic to insert unperformed nodes after changes have been made to the current solution. More...

class	FilteredHeuristicPathLNSOperator
	LNS-like operator based on a filtered first solution heuristic to rebuild the solution, after the destruction phase consisting of removing one route. More...

class	GlobalDimensionCumulOptimizer

class	GlobalVehicleBreaksConstraint
	GlobalVehicleBreaksConstraint ensures breaks constraints are enforced on all vehicles in the dimension passed to its constructor. More...

class	ImprovementSearchLimit

class	IndexPairSwapActiveOperator
	Operator which inserts inactive nodes into a path and makes a pair of active nodes inactive. More...

class	InitAndGetValues
	Utility class to encapsulate an IntVarIterator and use it in a range-based loop. More...

class	IntervalVar
	Interval variables are often used in scheduling. More...

class	IntervalVarElement

class	IntExpr
	The class IntExpr is the base of all integer expressions in constraint programming. More...

class	IntVar
	The class IntVar is a subset of IntExpr. More...

class	IntVarElement

class	IntVarIterator
	The class Iterator has two direct subclasses. More...

class	IntVarLocalSearchFilter

class	IntVarLocalSearchHandler

class	IntVarLocalSearchOperator
	Specialization of LocalSearchOperator built from an array of IntVars which specifies the scope of the operator. More...

class	LightPairRelocateOperator

class	LocalDimensionCumulOptimizer

class	LocalSearchFilter
	Local Search Filters are used for fast neighbor pruning. More...

class	LocalSearchFilterManager
	Filter manager: when a move is made, filters are executed to decide whether the solution is feasible and compute parts of the new cost. More...

class	LocalSearchMonitor

class	LocalSearchOperator
	The base class for all local search operators. More...

class	LocalSearchState

class	LocalSearchVariable

class	MakePairActiveOperator
	Pair-based neighborhood operators, designed to move nodes by pairs (pairs are static and given). More...

class	MakePairInactiveOperator
	Operator which makes pairs of active nodes inactive. More...

class	MakeRelocateNeighborsOperator
	Relocate neighborhood which moves chains of neighbors. More...

class	ModelCache
	Implements a complete cache for model elements: expressions and constraints. More...

class	ModelParser
	Model Parser. More...

class	ModelVisitor
	Model visitor. More...

class	NumericalRev
	Subclass of Rev<T> which adds numerical operations. More...

class	NumericalRevArray
	Subclass of RevArray<T> which adds numerical operations. More...

class	OptimizeVar
	This class encapsulates an objective. More...

class	Pack

class	PairExchangeOperator
	Operator which exchanges the position of two pairs; for both pairs the first node of the pair must be before the second node on the same path. More...

class	PairExchangeRelocateOperator
	Operator which exchanges the paths of two pairs (path have to be different). More...

class	PairNodeSwapActiveOperator
	Operator which inserts pairs of inactive nodes into a path and makes an active node inactive. More...

class	PairRelocateOperator
	Operator which moves a pair of nodes to another position where the first node of the pair must be before the second node on the same path. More...

class	PathOperator
	Base class of the local search operators dedicated to path modifications (a path is a set of nodes linked together by arcs). More...

class	PathState

class	ProfiledDecisionBuilder

class	PropagationBaseObject
	NOLINT. More...

class	PropagationMonitor

class	RegularLimit
	Usual limit based on wall_time, number of explored branches and number of failures in the search tree. More...

class	RelocateExpensiveChain
	RelocateExpensiveChain. More...

class	RelocatePathAndHeuristicInsertUnperformedOperator
	Heuristic-based local search operator which relocates an entire route to an empty vehicle of different vehicle class and then tries to insert unperformed nodes using the heuristic. More...

class	RelocateSubtrip
	Tries to move subtrips after an insertion node. More...

class	ResourceAssignmentOptimizer

class	Rev
	This class adds reversibility to a POD type. More...

class	RevArray
	Reversible array of POD types. More...

class	RevBitMatrix
	Matrix version of the RevBitSet class. More...

class	RevBitSet
	This class represents a reversible bitset. More...

class	RevGrowingArray
	This class is a reversible growing array. More...

class	RevImmutableMultiMap
	Reversible Immutable MultiMap class. More...

class	RevIntSet
	This is a special class to represent a 'residual' set of T. More...

class	RevPartialSequence
	--— RevPartialSequence --— More...

class	RevSwitch
	A reversible switch that can switch once from false to true. More...

class	RoutingCPSatWrapper

class	RoutingDimension
	Dimensions represent quantities accumulated at nodes along the routes. More...

class	RoutingGlopWrapper

class	RoutingIndexManager
	Manager for any NodeIndex <-> variable index conversion. More...

class	RoutingLinearSolverWrapper

class	RoutingModel

class	RoutingModelVisitor
	Routing model visitor. More...

class	SearchLimit
	Base class of all search limits. More...

class	SearchLog
	The base class of all search logs that periodically outputs information when the search is running. More...

class	SearchMonitor
	A search monitor is a simple set of callbacks to monitor all search events. More...

class	SequenceVar
	A sequence variable is a variable whose domain is a set of possible orderings of the interval variables. More...

class	SequenceVarElement
	The SequenceVarElement stores a partial representation of ranked interval variables in the underlying sequence variable. More...

class	SequenceVarLocalSearchHandler

class	SequenceVarLocalSearchOperator

class	SimpleBoundCosts
	A structure meant to store soft bounds and associated violation constants. More...

class	SimpleRevFIFO
	This class represent a reversible FIFO structure. More...

class	SmallRevBitSet
	This class represents a small reversible bitset (size <= 64). More...

class	SolutionCollector
	This class is the root class of all solution collectors. More...

class	SolutionPool
	This class is used to manage a pool of solutions. More...

class	Solver
	Solver Class. More...

class	SwapIndexPairOperator
	Operator which iterates through each alternative of a set of pairs. More...

class	SymmetryBreaker
	A symmetry breaker is an object that will visit a decision and create the 'symmetrical' decision in return. More...

struct	TravelBounds

class	TypeIncompatibilityChecker
	Checker for type incompatibilities. More...

class	TypeRegulationsChecker

class	TypeRegulationsConstraint
	The following constraint ensures that incompatibilities and requirements between types are respected. More...

class	TypeRequirementChecker
	Checker for type requirements. More...

class	UnaryDimensionChecker

class	UnsortedNullableRevBitset
	This class represents a reversible bitset. More...

class	VarLocalSearchOperator
	Base operator class for operators manipulating variables. More...

Typedefs
typedef VarLocalSearchOperator< SequenceVar, std::vector< int >, SequenceVarLocalSearchHandler >	SequenceVarLocalSearchOperatorTemplate

typedef std::function< int64_t(int64_t)>	RoutingTransitCallback1

typedef std::function< int64_t(int64_t, int64_t)>	RoutingTransitCallback2

typedef std::pair< std::vector< int64_t >, std::vector< int64_t > >	RoutingIndexPair

typedef std::vector< RoutingIndexPair >	RoutingIndexPairs

Enumerations
enum	VarTypes { UNSPECIFIED , DOMAIN_INT_VAR , BOOLEAN_VAR , CONST_VAR , VAR_ADD_CST , VAR_TIMES_CST , CST_SUB_VAR , OPP_VAR , TRACE_VAR }
	This enum is used internally to do dynamic typing on subclasses of integer variables. More...

enum class	DimensionSchedulingStatus { OPTIMAL , RELAXED_OPTIMAL_ONLY , INFEASIBLE }

Functions
int64_t	CpRandomSeed ()

std::ostream &	operator<< (std::ostream &out, const Solver *const s)

int64_t	Zero ()
	NOLINT. More...

int64_t	One ()
	This method returns 1. More...

std::ostream &	operator<< (std::ostream &out, const BaseObject *o)

std::ostream &	operator<< (std::ostream &out, const Assignment &assignment)

void	SetAssignmentFromAssignment (Assignment target_assignment, const std::vector< IntVar > &target_vars, const Assignment source_assignment, const std::vector< IntVar > &source_vars)
	NOLINT. More...

uint64_t	Hash1 (uint64_t value)
	Hash functions. More...

uint64_t	Hash1 (uint32_t value)

uint64_t	Hash1 (int64_t value)

uint64_t	Hash1 (int value)

uint64_t	Hash1 (void *const ptr)

template<class T >
uint64_t	Hash1 (const std::vector< T * > &ptrs)

uint64_t	Hash1 (const std::vector< int64_t > &ptrs)

template<class T >
LocalSearchOperator *	MakeLocalSearchOperator (Solver solver, const std::vector< IntVar > &vars, const std::vector< IntVar * > &secondary_vars, std::function< int(int64_t)> start_empty_path_class)
	Operator Factories. More...

template<class T >
bool	IsArrayConstant (const std::vector< T > &values, const T &value)

template<class T >
bool	IsArrayBoolean (const std::vector< T > &values)

template<class T >
bool	AreAllOnes (const std::vector< T > &values)

template<class T >
bool	AreAllNull (const std::vector< T > &values)

template<class T >
bool	AreAllGreaterOrEqual (const std::vector< T > &values, const T &value)

template<class T >
bool	AreAllLessOrEqual (const std::vector< T > &values, const T &value)

template<class T >
bool	AreAllPositive (const std::vector< T > &values)

template<class T >
bool	AreAllNegative (const std::vector< T > &values)

template<class T >
bool	AreAllStrictlyPositive (const std::vector< T > &values)

template<class T >
bool	AreAllStrictlyNegative (const std::vector< T > &values)

template<class T >
bool	IsIncreasingContiguous (const std::vector< T > &values)

template<class T >
bool	IsIncreasing (const std::vector< T > &values)

template<class T >
bool	IsArrayInRange (const std::vector< IntVar * > &vars, T range_min, T range_max)

bool	AreAllBound (const std::vector< IntVar * > &vars)

bool	AreAllBooleans (const std::vector< IntVar * > &vars)

template<class T >
bool	AreAllBoundOrNull (const std::vector< IntVar * > &vars, const std::vector< T > &values)
	Returns true if all the variables are assigned to a single value, or if their corresponding value is null. More...

bool	AreAllBoundTo (const std::vector< IntVar * > &vars, int64_t value)
	Returns true if all variables are assigned to 'value'. More...

int64_t	MaxVarArray (const std::vector< IntVar * > &vars)

int64_t	MinVarArray (const std::vector< IntVar * > &vars)

void	FillValues (const std::vector< IntVar * > &vars, std::vector< int64_t > *const values)

int64_t	PosIntDivUp (int64_t e, int64_t v)

int64_t	PosIntDivDown (int64_t e, int64_t v)

std::vector< int64_t >	ToInt64Vector (const std::vector< int > &input)

LocalSearchFilter *	MakePathStateFilter (Solver solver, std::unique_ptr< PathState > path_state, const std::vector< IntVar > &nexts)

LocalSearchFilter *	MakeUnaryDimensionFilter (Solver *solver, std::unique_ptr< UnaryDimensionChecker > checker, const std::string &dimension_name)

void	AppendTasksFromPath (const std::vector< int64_t > &path, const TravelBounds &travel_bounds, const RoutingDimension &dimension, DisjunctivePropagator::Tasks *tasks)

void	AppendTasksFromIntervals (const std::vector< IntervalVar * > &intervals, DisjunctivePropagator::Tasks *tasks)

void	FillPathEvaluation (const std::vector< int64_t > &path, const RoutingModel::TransitCallback2 &evaluator, std::vector< int64_t > *values)

void	FillTravelBoundsOfVehicle (int vehicle, const std::vector< int64_t > &path, const RoutingDimension &dimension, TravelBounds *travel_bounds)

DecisionBuilder *	MakeSetValuesFromTargets (Solver solver, std::vector< IntVar > variables, std::vector< int64_t > targets)
	A decision builder which tries to assign values to variables as close as possible to target values first. More...

bool	SolveModelWithSat (const RoutingModel &model, const RoutingSearchParameters &search_parameters, const Assignment initial_solution, Assignment solution)
	Attempts to solve the model using the cp-sat solver. More...

IntVarLocalSearchFilter *	MakeVehicleBreaksFilter (const RoutingModel &routing_model, const RoutingDimension &dimension)

DecisionBuilder *	MakeRestoreDimensionValuesForUnchangedRoutes (RoutingModel *model)

RoutingModelParameters	BuildModelParametersFromFlags ()
	Builds routing search parameters from flags. More...

RoutingSearchParameters	BuildSearchParametersFromFlags ()
	Builds routing search parameters from flags. More...

RoutingModelParameters	DefaultRoutingModelParameters ()

RoutingSearchParameters	DefaultRoutingSearchParameters ()

std::string	FindErrorInRoutingSearchParameters (const RoutingSearchParameters &search_parameters)
	Returns an empty std::string if the routing search parameters are valid, and a non-empty, human readable error description if they're not. More...

	DEFINE_STRONG_INT_TYPE (RoutingNodeIndex, int)
	Defining common types used in the routing library outside the main RoutingModel class has several purposes: 1) It allows some small libraries to avoid a dependency on routing. More...

	DEFINE_STRONG_INT_TYPE (RoutingCostClassIndex, int)

	DEFINE_STRONG_INT_TYPE (RoutingDimensionIndex, int)

	DEFINE_STRONG_INT_TYPE (RoutingDisjunctionIndex, int)

	DEFINE_STRONG_INT_TYPE (RoutingVehicleClassIndex, int)


template<class T >
Demon *	MakeConstraintDemon0 (Solver const s, T const ct, void(T::*method)(), const std::string &name)

template<class P >
std::string	ParameterDebugString (P param)

template<class P >
std::string	ParameterDebugString (P *param)
	Support limited to pointers to classes which define DebugString(). More...

template<class T , class P >
Demon *	MakeConstraintDemon1 (Solver const s, T const ct, void(T::*method)(P), const std::string &name, P param1)

template<class T , class P , class Q >
Demon *	MakeConstraintDemon2 (Solver const s, T const ct, void(T::*method)(P, Q), const std::string &name, P param1, Q param2)

template<class T , class P , class Q , class R >
Demon *	MakeConstraintDemon3 (Solver const s, T const ct, void(T::*method)(P, Q, R), const std::string &name, P param1, Q param2, R param3)


template<class T >
Demon *	MakeDelayedConstraintDemon0 (Solver const s, T const ct, void(T::*method)(), const std::string &name)

template<class T , class P >
Demon *	MakeDelayedConstraintDemon1 (Solver const s, T const ct, void(T::*method)(P), const std::string &name, P param1)

template<class T , class P , class Q >
Demon *	MakeDelayedConstraintDemon2 (Solver const s, T const ct, void(T::*method)(P, Q), const std::string &name, P param1, Q param2)

Detailed Description

Collection of objects used to extend the Constraint Solver library.

The vehicle routing library lets one model and solve generic vehicle routing problems ranging from the Traveling Salesman Problem to more complex problems such as the Capacitated Vehicle Routing Problem with Time Windows.

This file contains a set of objects that simplifies writing extensions of the library.

The main objects that define extensions are:

BaseIntExpr, the base class of all expressions that are not variables.
SimpleRevFIFO, a reversible FIFO list with templatized values. A reversible data structure is a data structure that reverts its modifications when the search is going up in the search tree, usually after a failure occurs.
RevImmutableMultiMap, a reversible immutable multimap.
MakeConstraintDemon<n> and MakeDelayedConstraintDemon<n> to wrap methods of a constraint as a demon.
RevSwitch, a reversible flip-once switch.
SmallRevBitSet, RevBitSet, and RevBitMatrix: reversible 1D or 2D bitsets.
LocalSearchOperator, IntVarLocalSearchOperator, ChangeValue and PathOperator, to create new local search operators.
LocalSearchFilter and IntVarLocalSearchFilter, to create new local search filters.
BaseLns, to write Large Neighborhood Search operators.
SymmetryBreaker, to describe model symmetries that will be broken during search using the 'Symmetry Breaking During Search' framework see Gent, I. P., Harvey, W., & Kelsey, T. (2002). Groups and Constraints: Symmetry Breaking During Search. Principles and Practice of Constraint Programming CP2002 (Vol. 2470, pp. 415-430). Springer. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.11.1442.

Then, there are some internal classes that are used throughout the solver and exposed in this file:

SearchLog, the root class of all periodic outputs during search.
ModelCache, A caching layer to avoid creating twice the same object.

The objective of a vehicle routing problem is to build routes covering a set of nodes minimizing the overall cost of the routes (usually proportional to the sum of the lengths of each segment of the routes) while respecting some problem-specific constraints (such as the length of a route). A route is equivalent to a path connecting nodes, starting/ending at specific starting/ending nodes.

The term "vehicle routing" is historical and the category of problems solved is not limited to the routing of vehicles: any problem involving finding routes visiting a given number of nodes optimally falls under this category of problems, such as finding the optimal sequence in a playlist. The literature around vehicle routing problems is extremely dense but one can find some basic introductions in the following links:

The vehicle routing library is a vertical layer above the constraint programming library (ortools/constraint_programming:cp). One has access to all underlying constrained variables of the vehicle routing model which can therefore be enriched by adding any constraint available in the constraint programming library.

There are two sets of variables available:

path variables:
- "next(i)" variables representing the immediate successor of the node corresponding to i; use IndexToNode() to get the node corresponding to a "next" variable value; note that node indices are strongly typed integers (cf. ortools/base/strong_int.h);
- "vehicle(i)" variables representing the vehicle route to which the node corresponding to i belongs;
- "active(i)" boolean variables, true if the node corresponding to i is visited and false if not; this can be false when nodes are either optional or part of a disjunction;
- The following relationships hold for all i: active(i) == 0 <=> next(i) == i <=> vehicle(i) == -1, next(i) == j => vehicle(j) == vehicle(i).
dimension variables, used when one is accumulating quantities along routes, such as weight or volume carried, distance or time:
- "cumul(i,d)" variables representing the quantity of dimension d when arriving at the node corresponding to i;
- "transit(i,d)" variables representing the quantity of dimension d added after visiting the node corresponding to i.
- The following relationship holds for all (i,d): next(i) == j => cumul(j,d) == cumul(i,d) + transit(i,d). Solving the vehicle routing problems is mainly done using approximate methods (namely local search, cf. http://en.wikipedia.org/wiki/Local_search_(optimization) ), potentially combined with exact techniques based on dynamic programming and exhaustive tree search. Advanced tips: Flags are available to tune the search used to solve routing problems. Here is a quick overview of the ones one might want to modify:
Limiting the search for solutions:
- routing_solution_limit (default: kint64max): stop the search after finding 'routing_solution_limit' improving solutions;
- routing_time_limit (default: kint64max): stop the search after 'routing_time_limit' milliseconds;
Customizing search:
- routing_first_solution (default: select the first node with an unbound successor and connect it to the first available node): selects the heuristic to build a first solution which will then be improved by local search; possible values are GlobalCheapestArc (iteratively connect two nodes which produce the cheapest route segment), LocalCheapestArc (select the first node with an unbound successor and connect it to the node which produces the cheapest route segment), PathCheapestArc (starting from a route "start" node, connect it to the node which produces the cheapest route segment, then extend the route by iterating on the last node added to the route).
- Local search neighborhoods:
  - routing_no_lns (default: false): forbids the use of Large Neighborhood Search (LNS); LNS can find good solutions but is usually very slow. Refer to the description of PATHLNS in the LocalSearchOperators enum in constraint_solver.h for more information.
  - routing_no_tsp (default: true): forbids the use of exact methods to solve "sub"-traveling salesman problems (TSPs) of the current model (such as sub-parts of a route, or one route in a multiple route problem). Uses dynamic programming to solve such TSPs with a maximum size (in number of nodes) up to cp_local_search_tsp_opt_size (flag with a default value of 13 nodes). It is not activated by default because it can slow down the search.
- Meta-heuristics: used to guide the search out of local minima found by local search. Note that, in general, a search with metaheuristics activated never stops, therefore one must specify a search limit. Several types of metaheuristics are provided:
  - routing_guided_local_search (default: false): activates guided local search (cf. http://en.wikipedia.org/wiki/Guided_Local_Search); this is generally the most efficient metaheuristic for vehicle routing;
  - routing_simulated_annealing (default: false): activates simulated annealing (cf. http://en.wikipedia.org/wiki/Simulated_annealing);
  - routing_tabu_search (default: false): activates tabu search (cf. http://en.wikipedia.org/wiki/Tabu_search).

Code sample: Here is a simple example solving a traveling salesman problem given a cost function callback (returns the cost of a route segment):

Define a custom distance/cost function from an index to another; in this example just returns the sum of the indices:

int64_t MyDistance(int64_t from, int64_t to) { return from + to; }
Create a routing model for a given problem size (int number of nodes) and number of routes (here, 1):

RoutingIndexManager manager(...number of nodes..., 1); RoutingModel routing(manager);
Set the cost function by registering an std::function<int64_t(int64_t, int64_t)> in the model and passing its index as the vehicle cost.

const int cost = routing.RegisterTransitCallback(MyDistance); routing.SetArcCostEvaluatorOfAllVehicles(cost);
Find a solution using Solve(), returns a solution if any (owned by routing):

const Assignment* solution = routing.Solve(); CHECK(solution != nullptr);
Inspect the solution cost and route (only one route here):

LOG(INFO) << "Cost " << solution->ObjectiveValue(); const int route_number = 0; for (int64_t node = routing.Start(route_number); !routing.IsEnd(node); node = solution->Value(routing.NextVar(node))) { LOG(INFO) << manager.IndexToNode(node); }

Keywords: Vehicle Routing, Traveling Salesman Problem, TSP, VRP, CVRPTW, PDP.

Typedef Documentation

◆ RoutingIndexPair

typedef std::pair<std::vector<int64_t>, std::vector<int64_t> > RoutingIndexPair

Definition at line 45 of file routing_types.h.

◆ RoutingIndexPairs

typedef std::vector<RoutingIndexPair> RoutingIndexPairs

Definition at line 46 of file routing_types.h.

◆ RoutingTransitCallback1

typedef std::function<int64_t(int64_t)> RoutingTransitCallback1

Definition at line 42 of file routing_types.h.

◆ RoutingTransitCallback2

typedef std::function<int64_t(int64_t, int64_t)> RoutingTransitCallback2

Definition at line 43 of file routing_types.h.

◆ SequenceVarLocalSearchOperatorTemplate

typedef VarLocalSearchOperator<SequenceVar, std::vector<int>, SequenceVarLocalSearchHandler> SequenceVarLocalSearchOperatorTemplate

Definition at line 1148 of file constraint_solveri.h.

Enumeration Type Documentation

◆ DimensionSchedulingStatus

enum class DimensionSchedulingStatus

strong

Enumerator
OPTIMAL
RELAXED_OPTIMAL_ONLY
INFEASIBLE

Definition at line 150 of file routing_lp_scheduling.h.

◆ VarTypes

enum VarTypes

This enum is used internally to do dynamic typing on subclasses of integer variables.

Enumerator
UNSPECIFIED
DOMAIN_INT_VAR
BOOLEAN_VAR
CONST_VAR
VAR_ADD_CST
VAR_TIMES_CST
CST_SUB_VAR
OPP_VAR
TRACE_VAR

Definition at line 115 of file constraint_solveri.h.

Function Documentation

◆ AppendTasksFromIntervals()

void operations_research::AppendTasksFromIntervals	(	const std::vector< IntervalVar * > &	intervals,
		DisjunctivePropagator::Tasks *	tasks
	)

◆ AppendTasksFromPath()

void operations_research::AppendTasksFromPath	(	const std::vector< int64_t > &	path,
		const TravelBounds &	travel_bounds,
		const RoutingDimension &	dimension,
		DisjunctivePropagator::Tasks *	tasks
	)

◆ AreAllBooleans()

bool operations_research::AreAllBooleans ( const std::vector< IntVar * > & vars )

inline

Definition at line 2955 of file constraint_solveri.h.

◆ AreAllBound()

bool operations_research::AreAllBound ( const std::vector< IntVar * > & vars )

inline

Definition at line 2946 of file constraint_solveri.h.

◆ AreAllBoundOrNull()

bool operations_research::AreAllBoundOrNull	(	const std::vector< IntVar * > &	vars,
		const std::vector< T > &	values
	)

Returns true if all the variables are assigned to a single value, or if their corresponding value is null.

Definition at line 2962 of file constraint_solveri.h.

◆ AreAllBoundTo()

bool operations_research::AreAllBoundTo	(	const std::vector< IntVar * > &	vars,
		int64_t	value
	)

inline

Returns true if all variables are assigned to 'value'.

Definition at line 2973 of file constraint_solveri.h.

◆ AreAllGreaterOrEqual()

bool operations_research::AreAllGreaterOrEqual	(	const std::vector< T > &	values,
		const T &	value
	)

Definition at line 2876 of file constraint_solveri.h.

◆ AreAllLessOrEqual()

bool operations_research::AreAllLessOrEqual	(	const std::vector< T > &	values,
		const T &	value
	)

Definition at line 2886 of file constraint_solveri.h.

◆ AreAllNegative()

bool operations_research::AreAllNegative ( const std::vector< T > & values )

Definition at line 2901 of file constraint_solveri.h.

◆ AreAllNull()

bool operations_research::AreAllNull ( const std::vector< T > & values )

Definition at line 2871 of file constraint_solveri.h.

◆ AreAllOnes()

bool operations_research::AreAllOnes ( const std::vector< T > & values )

Definition at line 2866 of file constraint_solveri.h.

◆ AreAllPositive()

bool operations_research::AreAllPositive ( const std::vector< T > & values )

Definition at line 2896 of file constraint_solveri.h.

◆ AreAllStrictlyNegative()

bool operations_research::AreAllStrictlyNegative ( const std::vector< T > & values )

Definition at line 2911 of file constraint_solveri.h.

◆ AreAllStrictlyPositive()

bool operations_research::AreAllStrictlyPositive ( const std::vector< T > & values )

Definition at line 2906 of file constraint_solveri.h.

◆ BuildModelParametersFromFlags()

RoutingModelParameters operations_research::BuildModelParametersFromFlags ( )

Builds routing search parameters from flags.

◆ BuildSearchParametersFromFlags()

RoutingSearchParameters operations_research::BuildSearchParametersFromFlags ( )

Builds routing search parameters from flags.

describe a valid set of routing search parameters.

◆ CpRandomSeed()

int64_t operations_research::CpRandomSeed ( )

inline

Definition at line 160 of file constraint_solver.h.

◆ DefaultRoutingModelParameters()

RoutingModelParameters operations_research::DefaultRoutingModelParameters ( )

◆ DefaultRoutingSearchParameters()

RoutingSearchParameters operations_research::DefaultRoutingSearchParameters ( )

◆ DEFINE_STRONG_INT_TYPE() [1/5]

operations_research::DEFINE_STRONG_INT_TYPE	(	RoutingCostClassIndex	,
		int
	)

◆ DEFINE_STRONG_INT_TYPE() [2/5]

operations_research::DEFINE_STRONG_INT_TYPE	(	RoutingDimensionIndex	,
		int
	)

◆ DEFINE_STRONG_INT_TYPE() [3/5]

operations_research::DEFINE_STRONG_INT_TYPE	(	RoutingDisjunctionIndex	,
		int
	)

◆ DEFINE_STRONG_INT_TYPE() [4/5]

operations_research::DEFINE_STRONG_INT_TYPE	(	RoutingNodeIndex	,
		int
	)

Defining common types used in the routing library outside the main RoutingModel class has several purposes: 1) It allows some small libraries to avoid a dependency on routing.

{h,cc}, eg. routing_neighborhoods.h. 2) It allows an easier wrapping via SWIG, which can have issues with intra-class types.

Users that depend on routing.{h,cc} should just use the RoutingModel:: equivalent, eg. RoutingModel::NodeIndex.

◆ DEFINE_STRONG_INT_TYPE() [5/5]

operations_research::DEFINE_STRONG_INT_TYPE	(	RoutingVehicleClassIndex	,
		int
	)

◆ FillPathEvaluation()

void operations_research::FillPathEvaluation	(	const std::vector< int64_t > &	path,
		const RoutingModel::TransitCallback2 &	evaluator,
		std::vector< int64_t > *	values
	)

◆ FillTravelBoundsOfVehicle()

void operations_research::FillTravelBoundsOfVehicle	(	int	vehicle,
		const std::vector< int64_t > &	path,
		const RoutingDimension &	dimension,
		TravelBounds *	travel_bounds
	)

◆ FillValues()

void operations_research::FillValues	(	const std::vector< IntVar * > &	vars,
		std::vector< int64_t > *const	values
	)

inline

Definition at line 3002 of file constraint_solveri.h.

◆ FindErrorInRoutingSearchParameters()

std::string operations_research::FindErrorInRoutingSearchParameters ( const RoutingSearchParameters & search_parameters )

Returns an empty std::string if the routing search parameters are valid, and a non-empty, human readable error description if they're not.

◆ Hash1() [1/7]

uint64_t operations_research::Hash1 ( const std::vector< int64_t > & ptrs )

inline

Definition at line 262 of file constraint_solveri.h.

◆ Hash1() [2/7]

uint64_t operations_research::Hash1 ( const std::vector< T * > & ptrs )

Definition at line 252 of file constraint_solveri.h.

◆ Hash1() [3/7]

uint64_t operations_research::Hash1 ( int value )

inline

Definition at line 240 of file constraint_solveri.h.

◆ Hash1() [4/7]

uint64_t operations_research::Hash1 ( int64_t value )

inline

Definition at line 236 of file constraint_solveri.h.

◆ Hash1() [5/7]

uint64_t operations_research::Hash1 ( uint32_t value )

inline

Definition at line 225 of file constraint_solveri.h.

◆ Hash1() [6/7]

uint64_t operations_research::Hash1 ( uint64_t value )

inline

Hash functions.

value = (value << 21) - value - 1;

value * 265

value * 21

Definition at line 214 of file constraint_solveri.h.

◆ Hash1() [7/7]

uint64_t operations_research::Hash1 ( void *const ptr )

inline

Definition at line 242 of file constraint_solveri.h.

◆ IsArrayBoolean()

bool operations_research::IsArrayBoolean ( const std::vector< T > & values )

Definition at line 2856 of file constraint_solveri.h.

◆ IsArrayConstant()

bool operations_research::IsArrayConstant	(	const std::vector< T > &	values,
		const T &	value
	)

Definition at line 2846 of file constraint_solveri.h.

◆ IsArrayInRange()

bool operations_research::IsArrayInRange	(	const std::vector< IntVar * > &	vars,
		T	range_min,
		T	range_max
	)

Definition at line 2936 of file constraint_solveri.h.

◆ IsIncreasing()

bool operations_research::IsIncreasing ( const std::vector< T > & values )

Definition at line 2926 of file constraint_solveri.h.

◆ IsIncreasingContiguous()

bool operations_research::IsIncreasingContiguous ( const std::vector< T > & values )

Definition at line 2916 of file constraint_solveri.h.

◆ MakeConstraintDemon0()

Demon * operations_research::MakeConstraintDemon0	(	Solver *const	s,
		T *const	ct,
		void(T::*)()	method,
		const std::string &	name
	)

Definition at line 519 of file constraint_solveri.h.

◆ MakeConstraintDemon1()

Demon * operations_research::MakeConstraintDemon1	(	Solver *const	s,
		T *const	ct,
		void(T::*)(P)	method,
		const std::string &	name,
		P	param1
	)

Definition at line 560 of file constraint_solveri.h.

◆ MakeConstraintDemon2()

Demon * operations_research::MakeConstraintDemon2	(	Solver *const	s,
		T *const	ct,
		void(T::*)(P, Q)	method,
		const std::string &	name,
		P	param1,
		Q	param2
	)

Definition at line 599 of file constraint_solveri.h.

◆ MakeConstraintDemon3()

Demon * operations_research::MakeConstraintDemon3	(	Solver *const	s,
		T *const	ct,
		void(T::*)(P, Q, R)	method,
		const std::string &	name,
		P	param1,
		Q	param2,
		R	param3
	)

Definition at line 642 of file constraint_solveri.h.

◆ MakeDelayedConstraintDemon0()

Demon * operations_research::MakeDelayedConstraintDemon0	(	Solver *const	s,
		T *const	ct,
		void(T::*)()	method,
		const std::string &	name
	)

Definition at line 682 of file constraint_solveri.h.

◆ MakeDelayedConstraintDemon1()

Demon * operations_research::MakeDelayedConstraintDemon1	(	Solver *const	s,
		T *const	ct,
		void(T::*)(P)	method,
		const std::string &	name,
		P	param1
	)

Definition at line 718 of file constraint_solveri.h.

◆ MakeDelayedConstraintDemon2()

Demon * operations_research::MakeDelayedConstraintDemon2	(	Solver *const	s,
		T *const	ct,
		void(T::*)(P, Q)	method,
		const std::string &	name,
		P	param1,
		Q	param2
	)

Definition at line 762 of file constraint_solveri.h.

◆ MakeLocalSearchOperator()

LocalSearchOperator * operations_research::MakeLocalSearchOperator	(	Solver *	solver,
		const std::vector< IntVar * > &	vars,
		const std::vector< IntVar * > &	secondary_vars,
		std::function< int(int64_t)>	start_empty_path_class
	)

Operator Factories.

◆ MakePathStateFilter()

LocalSearchFilter * operations_research::MakePathStateFilter	(	Solver *	solver,
		std::unique_ptr< PathState >	path_state,
		const std::vector< IntVar * > &	nexts
	)

◆ MakeRestoreDimensionValuesForUnchangedRoutes()

DecisionBuilder * operations_research::MakeRestoreDimensionValuesForUnchangedRoutes ( RoutingModel * model )

◆ MakeSetValuesFromTargets()

DecisionBuilder * operations_research::MakeSetValuesFromTargets	(	Solver *	solver,
		std::vector< IntVar * >	variables,
		std::vector< int64_t >	targets
	)

A decision builder which tries to assign values to variables as close as possible to target values first.

◆ MakeUnaryDimensionFilter()

LocalSearchFilter * operations_research::MakeUnaryDimensionFilter	(	Solver *	solver,
		std::unique_ptr< UnaryDimensionChecker >	checker,
		const std::string &	dimension_name
	)

◆ MakeVehicleBreaksFilter()

IntVarLocalSearchFilter * operations_research::MakeVehicleBreaksFilter	(	const RoutingModel &	routing_model,
		const RoutingDimension &	dimension
	)

◆ MaxVarArray()

int64_t operations_research::MaxVarArray ( const std::vector< IntVar * > & vars )

inline

The std::max<int64_t> is needed for compilation on MSVC.

Definition at line 2982 of file constraint_solveri.h.

◆ MinVarArray()

int64_t operations_research::MinVarArray ( const std::vector< IntVar * > & vars )

inline

The std::min<int64_t> is needed for compilation on MSVC.

Definition at line 2992 of file constraint_solveri.h.

◆ One()

int64_t operations_research::One ( )

inline

This method returns 1.

Definition at line 3164 of file constraint_solver.h.

◆ operator<<() [1/3]

std::ostream & operations_research::operator<<	(	std::ostream &	out,
		const Assignment &	assignment
	)

◆ operator<<() [2/3]

std::ostream & operations_research::operator<<	(	std::ostream &	out,
		const BaseObject *	o
	)

◆ operator<<() [3/3]

std::ostream & operations_research::operator<<	(	std::ostream &	out,
		const Solver *const	s
	)

◆ ParameterDebugString() [1/2]

std::string operations_research::ParameterDebugString ( P * param )

Support limited to pointers to classes which define DebugString().

Definition at line 531 of file constraint_solveri.h.

◆ ParameterDebugString() [2/2]

std::string operations_research::ParameterDebugString ( P param )

Definition at line 525 of file constraint_solveri.h.

◆ PosIntDivDown()

int64_t operations_research::PosIntDivDown	(	int64_t	e,
		int64_t	v
	)

inline

Definition at line 3016 of file constraint_solveri.h.

◆ PosIntDivUp()

int64_t operations_research::PosIntDivUp	(	int64_t	e,
		int64_t	v
	)

inline

Definition at line 3011 of file constraint_solveri.h.

◆ SetAssignmentFromAssignment()

void operations_research::SetAssignmentFromAssignment	(	Assignment *	target_assignment,
		const std::vector< IntVar * > &	target_vars,
		const Assignment *	source_assignment,
		const std::vector< IntVar * > &	source_vars
	)

NOLINT.

Given a "source_assignment", clears the "target_assignment" and adds all IntVars in "target_vars", with the values of the variables set according to the corresponding values of "source_vars" in "source_assignment". source_vars and target_vars must have the same number of elements. The source and target assignments can belong to different Solvers.

◆ SolveModelWithSat()

bool operations_research::SolveModelWithSat	(	const RoutingModel &	model,
		const RoutingSearchParameters &	search_parameters,
		const Assignment *	initial_solution,
		Assignment *	solution
	)

Attempts to solve the model using the cp-sat solver.

As of 5/2019, will solve the TSP corresponding to the model if it has a single vehicle. Therefore the resulting solution might not actually be feasible. Will return false if a solution could not be found.

◆ ToInt64Vector()

std::vector< int64_t > operations_research::ToInt64Vector ( const std::vector< int > & input )

◆ Zero()

int64_t operations_research::Zero ( )

inline

NOLINT.

This method returns 0. It is useful when 0 can be cast either as a pointer or as an integer value and thus lead to an ambiguous function call.

Definition at line 3161 of file constraint_solver.h.

C++ Reference

C++ Reference: Routing

Classes

Typedefs

Enumerations

Functions

Detailed Description

Typedef Documentation

◆ RoutingIndexPair

◆ RoutingIndexPairs

◆ RoutingTransitCallback1

◆ RoutingTransitCallback2

◆ SequenceVarLocalSearchOperatorTemplate

Enumeration Type Documentation

◆ DimensionSchedulingStatus

◆ VarTypes

Function Documentation

◆ AppendTasksFromIntervals()

◆ AppendTasksFromPath()

◆ AreAllBooleans()

◆ AreAllBound()

◆ AreAllBoundOrNull()

◆ AreAllBoundTo()

◆ AreAllGreaterOrEqual()

◆ AreAllLessOrEqual()

◆ AreAllNegative()

◆ AreAllNull()

◆ AreAllOnes()

◆ AreAllPositive()

◆ AreAllStrictlyNegative()

◆ AreAllStrictlyPositive()

◆ BuildModelParametersFromFlags()

◆ BuildSearchParametersFromFlags()

◆ CpRandomSeed()

◆ DefaultRoutingModelParameters()

◆ DefaultRoutingSearchParameters()

◆ DEFINE_STRONG_INT_TYPE() [1/5]

◆ DEFINE_STRONG_INT_TYPE() [2/5]

◆ DEFINE_STRONG_INT_TYPE() [3/5]

◆ DEFINE_STRONG_INT_TYPE() [4/5]

◆ DEFINE_STRONG_INT_TYPE() [5/5]

◆ FillPathEvaluation()

◆ FillTravelBoundsOfVehicle()

◆ FillValues()

◆ FindErrorInRoutingSearchParameters()

◆ Hash1() [1/7]

◆ Hash1() [2/7]

◆ Hash1() [3/7]

◆ Hash1() [4/7]

◆ Hash1() [5/7]

◆ Hash1() [6/7]

◆ Hash1() [7/7]

◆ IsArrayBoolean()

◆ IsArrayConstant()

◆ IsArrayInRange()

◆ IsIncreasing()

◆ IsIncreasingContiguous()

◆ MakeConstraintDemon0()

◆ MakeConstraintDemon1()

◆ MakeConstraintDemon2()

◆ MakeConstraintDemon3()

◆ MakeDelayedConstraintDemon0()

◆ MakeDelayedConstraintDemon1()

◆ MakeDelayedConstraintDemon2()

◆ MakeLocalSearchOperator()

◆ MakePathStateFilter()

◆ MakeRestoreDimensionValuesForUnchangedRoutes()

◆ MakeSetValuesFromTargets()

◆ MakeUnaryDimensionFilter()

◆ MakeVehicleBreaksFilter()

◆ MaxVarArray()

◆ MinVarArray()

◆ One()

◆ operator<<() [1/3]

◆ operator<<() [2/3]

◆ operator<<() [3/3]

◆ ParameterDebugString() [1/2]

◆ ParameterDebugString() [2/2]

◆ PosIntDivDown()

◆ PosIntDivUp()