C++ Reference: class MPSolverInterface

Note: This documentation is automatically generated.

This class wraps the actual mathematical programming solvers. Each solver (GLOP, CLP, CBC, GLPK, SCIP) has its own interface class that derives from this abstract class. This class is never directly accessed by the user. @see glop_interface.cc @see cbc_interface.cc @see clp_interface.cc @see glpk_interface.cc @see scip_interface.cc

Method
AddIndicatorConstraint	Return type: virtual bool Arguments: MPConstraint* const ct Adds an indicator constraint. Returns true if the feature is supported by the underlying solver.
AddRowConstraint	Return type: virtual void Arguments: MPConstraint* const ct Adds a linear constraint.
AddVariable	Return type: virtual void Arguments: MPVariable* const var Add a variable.
best_objective_bound	Return type: double Returns the best objective bound. The problem must be discrete, otherwise it crashes, or returns trivial bound (+/- inf) in NDEBUG mode.
BranchingPriorityChangedForVariable	Return type: virtual void Arguments: int var_index
CheckSolutionExists	Return type: virtual bool Checks whether a feasible solution exists. The behavior is similar to CheckSolutionIsSynchronized() above.
CheckSolutionIsSynchronized	Return type: bool Checks whether the solution is synchronized with the model, i.e. whether the model has changed since the solution was computed last. If it isn't, it crashes in NDEBUG, and returns false othwerwise.
CheckSolutionIsSynchronizedAndExists	Return type: bool Handy shortcut to do both checks above (it is often used).
ClearConstraint	Return type: virtual void Arguments: MPConstraint* const constraint Clears a constraint from all its terms.
ClearObjective	Return type: virtual void Clears the objective from all its terms.
column_status	Return type: virtual MPSolver::BasisStatus Arguments: int variable_index Returns the basis status of a constraint.
ComputeExactConditionNumber	Return type: virtual double Computes exact condition number. Only available for continuous problems and only implemented in GLPK.
constraint_is_extracted	Return type: bool Arguments: int ct_index
DirectlySolveProto	Return type: virtual std::optional<MPSolutionResponse> Arguments: const MPModelRequest& request, // `interrupt` is non-const because the internal // solver may set it to true itself, in some cases. std::atomic<bool>* interrupt Attempts to directly solve a MPModelRequest, bypassing the MPSolver data structures entirely. Like MPSolver::SolveWithProto(), optionally takes in an 'interrupt' boolean. Returns {} (eg. absl::nullopt) if direct-solve is not supported by the underlying solver (possibly because interrupt != nullptr), in which case the user should fall back to using MPSolver.
InterruptSolve	Return type: virtual bool
IsContinuous	Return type: virtual bool ----- Misc ----- Queries problem type. For simplicity, the distinction between continuous and discrete is based on the declaration of the user when the solver is created (example: GLPK_LINEAR_PROGRAMMING vs. GLPK_MIXED_INTEGER_PROGRAMMING), not on the actual content of the model. Returns true if the problem is continuous.
IsLP	Return type: virtual bool Returns true if the problem is continuous and linear.
IsMIP	Return type: virtual bool Returns true if the problem is discrete and linear.
iterations	Return type: virtual int64_t ------ Query statistics on the solution and the solve ------ Returns the number of simplex iterations. The problem must be discrete, otherwise it crashes, or returns kUnknownNumberOfIterations in NDEBUG mode.
last_variable_index	Return type: int Returns the index of the last variable extracted.
MPSolverInterface	Return type: explicit Arguments: MPSolver* const solver Constructor. The user will access the MPSolverInterface through the MPSolver passed as argument.
~MPSolverInterface	Return type: virtual
NextSolution	Return type: virtual bool See MPSolver::NextSolution() for contract.
nodes	Return type: virtual int64_t Returns the number of branch-and-bound nodes. The problem must be discrete, otherwise it crashes, or returns kUnknownNumberOfNodes in NDEBUG mode.
objective_value	Return type: double Returns the objective value of the best solution found so far.
quiet	Return type: bool Returns the boolean indicating the verbosity of the solver output.
Reset	Return type: virtual void ----- Model modifications and extraction ----- Resets extracted model.
result_status	Return type: MPSolver::ResultStatus Returns the result status of the last solve.
row_status	Return type: virtual MPSolver::BasisStatus Arguments: int constraint_index Returns the basis status of a row.
set_constraint_as_extracted	Return type: void Arguments: int ct_index, bool extracted
set_quiet	Return type: void Arguments: bool quiet_value Sets the boolean indicating the verbosity of the solver output.
set_variable_as_extracted	Return type: void Arguments: int var_index, bool extracted
SetCallback	Return type: virtual void Arguments: MPCallback* mp_callback See MPSolver::SetCallback() for details.
SetCoefficient	Return type: virtual void Arguments: MPConstraint* const constraint, const MPVariable* const variable, double new_value, double old_value Changes a coefficient in a constraint.
SetConstraintBounds	Return type: virtual void Arguments: int index, double lb, double ub Modify bounds of an extracted variable.
SetObjectiveCoefficient	Return type: virtual void Arguments: const MPVariable* const variable, double coefficient Changes a coefficient in the linear objective.
SetObjectiveOffset	Return type: virtual void Arguments: double value Changes the constant term in the linear objective.
SetOptimizationDirection	Return type: virtual void Arguments: bool maximize Sets the optimization direction (min/max).
SetStartingLpBasis	Return type: virtual void Arguments: const std::vector<MPSolver::BasisStatus>& variable_statuses, const std::vector<MPSolver::BasisStatus>& constraint_statuses See MPSolver::SetStartingLpBasis().
SetVariableBounds	Return type: virtual void Arguments: int index, double lb, double ub Modifies bounds of an extracted variable.
SetVariableInteger	Return type: virtual void Arguments: int index, bool integer Modifies integrality of an extracted variable.
Solve	Return type: virtual MPSolver::ResultStatus Arguments: const MPSolverParameters& param ----- Solve ----- Solves problem with specified parameter values. Returns true if the solution is optimal.
SolverVersion	Return type: virtual std::string Returns a string describing the underlying solver and its version.
SupportsCallbacks	Return type: virtual bool
underlying_solver	Return type: virtual void* Returns the underlying solver.
variable_is_extracted	Return type: bool Arguments: int var_index
Write	Return type: virtual void Arguments: const std::string& filename Writes the model using the solver internal write function. Currently only available for GurobiInterface.