Class mmin_constr_gencan (o2scl)¶
-
template<class param_t, class func_t, class dfunc_t = func_t, class hfunc_t = func_t, class vec_t = boost::numeric::ublas::vector<double>>
class mmin_constr_gencan : public ool_constr_min<param_t, func_t, func_t, func_t, boost::numeric::ublas::vector<double>>¶ Constrained minimization by the “GENCAN” method (OOL)
Note
Not working yet
Public Functions
-
inline mmin_constr_gencan()¶
-
inline virtual int alloc(const size_t n)¶
Allocate memory.
-
inline virtual int free()¶
Free previously allocated memory.
-
inline virtual int set(func_t &fn, dfunc_t &dfn, hfunc_t &hfn, vec_t &init, param_t &par)¶
Set the function, the initial guess, and the parameters.
-
inline virtual int restart()¶
Restart the minimizer.
-
inline virtual int iterate()¶
Perform an iteration.
-
inline virtual int is_optimal()¶
See if we’re finished.
-
inline const char *type()¶
Return string denoting type (“mmin_constr_gencan”)
Public Members
-
double epsgpen¶
Tolerance on Euclidean norm of projected gradient (default 1.0e-5)
-
double epsgpsn¶
Tolerance on infinite norm of projected gradient (default 1.0e-5)
-
double fmin¶
Minimum function value (default \( 10^{-99} \))
If the function value is below this value, then the algorithm assumes that the function is not bounded and exits.
-
double udelta0¶
Trust-region radius (default -1.0)
-
double ucgmia¶
Maximum interations per variable (default -1.0)
-
double ucgmib¶
Extra maximum iterations (default -1.0)
-
int cg_scre¶
Conjugate gradient condition type (default 1)
-
double cg_gpnf¶
Projected gradient norm (default 1.0e-5)
-
double cg_epsi¶
Desc (default 1.0e-1)
-
double cg_epsf¶
Desc (default 1.0e-5)
-
double cg_epsnqmp¶
Stopping fractional tolerance for conjugate gradient (default 1.0e-4)
-
int cg_maxitnqmp¶
Maximum iterations for conjugate gradient (default 5)
-
int nearlyq¶
Set to 1 if the function is nearly quadratic (default 0)
-
double nint¶
Interpolation constant (default 2.0)
-
double next¶
Extrapolation constant (default 2.0)
-
int mininterp¶
Minimum interpolation size (default 4)
-
int maxextrap¶
Maximum extrapolations in truncated Newton direction (default 100)
-
int trtype¶
Type of trust region (default 0)
-
double eta¶
Threshold for abandoning current face (default 0.9)
-
double delmin¶
Minimum trust region for truncated Newton direction (default 0.1)
-
double lspgmi¶
Minimum spectral steplength (default 1.0e-10)
-
double lspgma¶
Maximum spectral steplength (default 1.0e10)
-
double theta¶
Constant for the angle condition (default 1.0e-6)
-
double gamma¶
Constant for Armijo condition (default 1.0e-4)
-
double beta¶
Constant for beta condition (default 0.5)
-
double sigma1¶
Lower bound to the step length reduction (default 0.1)
-
double sigma2¶
Upper bound to the step length reduction (default 0.9)
-
double epsrel¶
Relative small number (default 1.0e-7)
-
double epsabs¶
Absolute small number (default 1.0e-10)
-
double infrel¶
Relative infinite number (default 1.0e20)
-
double infabs¶
Absolute infinite number (default 1.0e99)
-
inline mmin_constr_gencan()¶