tinyopt::Options Struct Reference

Classes
struct	Optimization
struct	Hessian
struct	CostScaling
struct	StopCriteria
struct	LogOptions
struct	LM
struct	GD
struct	CG
struct	DL
struct	BFGS
struct	LBFGS

Public Member Functions
	Options (Solver type=Solver::LevenbergMarquardt)

Public Attributes
LinearSolverMethod	linear_solver = LinearSolverMethod::LDLT
	Linear system method. LDLT is enabled by default; methods require their CMake option.
double	svd_relative_threshold = 0.0
	Relative singular-value cutoff for SVD; 0 uses Eigen's default threshold.
struct tinyopt::Options::LM	lm
Optimization options
struct tinyopt::Options::Optimization	opt
Hessian Properties
struct tinyopt::Options::Hessian	hessian
Cost scaling options (mostly for NLLS solvers really)
struct tinyopt::Options::CostScaling	cost
Stop criteria
struct tinyopt::Options::StopCriteria	stop
Logging Options
struct tinyopt::Options::LogOptions	log
Gradient Descent options
struct tinyopt::Options::GD	gd
struct tinyopt::Options::CG	cg
struct tinyopt::Options::DL	dl
struct tinyopt::Options::BFGS	bfgs
struct tinyopt::Options::LBFGS	lbfgs

Solver Type
enum	Solver {   LevenbergMarquardt = 0 , GaussNewton , GradientDescent , ConjugateGradient ,   DogLeg , BFGS , LBFGS }
Solver	solver_type = Solver::LevenbergMarquardt
	Which solver to use. Default is LevenbergMarquardt for NLLS problems.

---

The documentation for this struct was generated from the following file:

• include/tinyopt/optimizers/options.h