Reaktoro
A unified framework for modeling chemically reactive systems
OptimumSolverBase Class Referenceabstract

The base class for all optimization algorithms. More...

#include <OptimumSolverBase.hpp>

Inheritance diagram for OptimumSolverBase:
[legend]

Public Member Functions

virtual ~OptimumSolverBase ()=0
 Pure virtual destructor.
 
virtual auto solve (const OptimumProblem &problem, OptimumState &state) -> OptimumResult=0
 Solve an optimisation problem. More...
 
virtual auto solve (const OptimumProblem &problem, OptimumState &state, const OptimumOptions &options) -> OptimumResult=0
 Solve an optimisation problem with given options. More...
 
virtual auto dxdp (VectorConstRef dgdp, VectorConstRef dbdp) -> Vector=0
 Return the sensitivity dx/dp of the solution x with respect to a vector of parameters p. More...
 
virtual auto clone () const -> OptimumSolverBase *=0
 Return a clone of this instance.
 

Detailed Description

The base class for all optimization algorithms.

Member Function Documentation

◆ solve() [1/2]

virtual auto solve ( const OptimumProblem problem,
OptimumState state 
) -> OptimumResult
pure virtual

Solve an optimisation problem.

Parameters
problemThe definition of the optimisation problem
state[in,out]The initial guess and the final state of the optimisation calculation

Implemented in OptimumSolverSimplex, OptimumSolverActNewton, OptimumSolverIpAction, OptimumSolverIpActive, OptimumSolverIpOpt, OptimumSolverIpBounds, OptimumSolverIpNewton, OptimumSolverKarpov, and OptimumSolverRefiner.

◆ solve() [2/2]

virtual auto solve ( const OptimumProblem problem,
OptimumState state,
const OptimumOptions options 
) -> OptimumResult
pure virtual

Solve an optimisation problem with given options.

Parameters
problemThe definition of the optimisation problem
state[in,out]The initial guess and the final state of the optimisation calculation
optionsThe options for the optimisation calculation

Implemented in OptimumSolverSimplex, OptimumSolverActNewton, OptimumSolverIpAction, OptimumSolverIpActive, OptimumSolverIpOpt, OptimumSolverIpBounds, OptimumSolverIpNewton, OptimumSolverKarpov, and OptimumSolverRefiner.

◆ dxdp()

virtual auto dxdp ( VectorConstRef  dgdp,
VectorConstRef  dbdp 
) -> Vector
pure virtual

Return the sensitivity dx/dp of the solution x with respect to a vector of parameters p.

Parameters
dgdpThe derivatives dg/dp of the objective gradient grad(f) with respect to the parameters p
dbdpThe derivatives db/dp of the vector b with respect to the parameters p

Implemented in OptimumSolverSimplex, OptimumSolverActNewton, OptimumSolverIpAction, OptimumSolverIpActive, OptimumSolverIpOpt, OptimumSolverIpBounds, OptimumSolverIpNewton, OptimumSolverKarpov, and OptimumSolverRefiner.


The documentation for this class was generated from the following files: