A vector field v(x) = arg min_y c(x)^T y s.t. q(x) <= A(x)y <= p(x), l(x) <= y <= u(x) where c,A,q,p,l, and u are smooth functions. More...
#include <LPOptimumFunction.h>
Inheritance diagram for Optimization::LPOptimizerFunction:
Public Member Functions | |
| virtual int | NumDimensions () const |
| virtual void | InitLP (const Vector &x) |
| virtual void | UpdateLP (const Vector &x) |
| virtual bool | LPJacobianA_i (const Vector &x, int i, Matrix &JAi) |
| virtual bool | LPJacobianA_i (const Vector &x, int i, SparseMatrix &JAi) |
| virtual bool | LPJacobianP_i (const Vector &x, int i, Vector &Jpi) |
| virtual bool | LPJacobianQ_i (const Vector &x, int i, Vector &Jqi) |
| virtual bool | LPJacobianL_j (const Vector &x, int j, Vector &Jlj) |
| virtual bool | LPJacobianU_j (const Vector &x, int j, Vector &Juj) |
| virtual void | UpdateGLPK (const Vector &x) |
| virtual std::string | Label () const |
| virtual void | PreEval (const Vector &x) |
| virtual void | Eval (const Vector &x, Vector &v) |
| virtual void | Jacobian (const Vector &x, Matrix &J) |
| virtual void | Jacobian_i (const Vector &x, int i, Vector &Ji) |
 Public Member Functions inherited from Math::VectorFieldFunction | |
| virtual std::string | Label (int i) const |
| virtual std::string | VariableLabel (int i) const |
| virtual void | operator() (const Vector &x, Vector &v) |
| virtual Real | Eval_i (const Vector &x, int i) |
| virtual Real | Jacobian_ij (const Vector &x, int i, int j) |
| virtual void | Jacobian_j (const Vector &x, int j, Vector &Jj) |
| virtual void | DirectionalDeriv (const Vector &x, const Vector &h, Vector &v) |
| virtual Real | Divergence (const Vector &x) |
| virtual void | Hessian_i (const Vector &x, int i, Matrix &Hi) |
| virtual Real | Hessian_ijk (const Vector &x, int i, int j, int k) |
Public Attributes | |
| bool | initialized |
| bool | sparse |
| LinearProgram | lp |
| LinearProgram_Sparse | lps |
| GLPKInterface | glpk |
| LinearProgram::Result | solveResult |
| Vector | yopt |
Detailed Description
A vector field v(x) = arg min_y c(x)^T y s.t. q(x) <= A(x)y <= p(x), l(x) <= y <= u(x) where c,A,q,p,l, and u are smooth functions.
Subclasses must overload UpdateLP and LPJacobian[Z] if Z depends on x. For faster updating, use InitLP to set up all constant members of the LP and in UpdateLP, only change the members that depend on x.
Member Function Documentation
|
virtual |
Subclasses may overload this for a sparse update – may also help with warm start.
The documentation for this class was generated from the following files:
- LPOptimumFunction.h
- LPOptimumFunction.cpp
