gtsam::LevenbergMarquardtOptimizer Class Reference

#include <LevenbergMarquardtOptimizer.h>

Inheritance diagram for gtsam::LevenbergMarquardtOptimizer:

Collaboration diagram for gtsam::LevenbergMarquardtOptimizer:

Public Types
typedef std::shared_ptr< LevenbergMarquardtOptimizer >	shared_ptr

Public Member Functions
Constructors/Destructor
	LevenbergMarquardtOptimizer (const NonlinearFactorGraph &graph, const Values &initialValues, const LevenbergMarquardtParams &params=LevenbergMarquardtParams())
	LevenbergMarquardtOptimizer (const NonlinearFactorGraph &graph, const Values &initialValues, const Ordering &ordering, const LevenbergMarquardtParams &params=LevenbergMarquardtParams())
	~LevenbergMarquardtOptimizer () override
Standard interface
double	lambda () const
	Access the current damping value.
int	getInnerIterations () const
	Access the current number of inner iterations.
void	print (const std::string &str="") const
	print
Advanced interface
GaussianFactorGraph::shared_ptr	iterate () override
const LevenbergMarquardtParams &	params () const
void	writeLogFile (double currentError)
virtual GaussianFactorGraph::shared_ptr	linearize () const
GaussianFactorGraph	buildDampedSystem (const GaussianFactorGraph &linear, const VectorValues &sqrtHessianDiagonal) const
bool	tryLambda (const GaussianFactorGraph &linear, const VectorValues &sqrtHessianDiagonal)
Standard interface
virtual const Values &	optimize ()
const Values &	optimizeSafely ()
double	error () const
	return error in current optimizer state
size_t	iterations () const
	return number of iterations in current optimizer state
const Values &	values () const
	return values in current optimizer state
const NonlinearFactorGraph &	graph () const
	return the graph with nonlinear factors
Advanced interface
virtual VectorValues	solve (const GaussianFactorGraph &gfg, const NonlinearOptimizerParams &params) const

Protected Member Functions
void	initTime ()
const NonlinearOptimizerParams &	_params () const override
void	defaultOptimize ()

Protected Attributes
const LevenbergMarquardtParams	params_
	LM parameters.
std::chrono::time_point< std::chrono::high_resolution_clock >	startTime_
	time when optimization started
NonlinearFactorGraph	graph_
	The graph with nonlinear factors.
std::unique_ptr< internal::NonlinearOptimizerState >	state_
	PIMPL'd state.

Detailed Description

This class performs Levenberg-Marquardt nonlinear optimization

Constructor & Destructor Documentation

LevenbergMarquardtOptimizer() [1/2]

gtsam::LevenbergMarquardtOptimizer::LevenbergMarquardtOptimizer	(	const NonlinearFactorGraph &	graph,
		const Values &	initialValues,
		const LevenbergMarquardtParams &	params = LevenbergMarquardtParams()
	)

Standard constructor, requires a nonlinear factor graph, initial variable assignments, and optimization parameters. For convenience this version takes plain objects instead of shared pointers, but internally copies the objects.

Parameters

graph	The nonlinear factor graph to optimize
initialValues	The initial variable assignments
params	The optimization parameters

LevenbergMarquardtOptimizer() [2/2]

gtsam::LevenbergMarquardtOptimizer::LevenbergMarquardtOptimizer	(	const NonlinearFactorGraph &	graph,
		const Values &	initialValues,
		const Ordering &	ordering,
		const LevenbergMarquardtParams &	params = LevenbergMarquardtParams()
	)

Standard constructor, requires a nonlinear factor graph, initial variable assignments, and optimization parameters. For convenience this version takes plain objects instead of shared pointers, but internally copies the objects.

Parameters

graph	The nonlinear factor graph to optimize
initialValues	The initial variable assignments

~LevenbergMarquardtOptimizer()

gtsam::LevenbergMarquardtOptimizer::~LevenbergMarquardtOptimizer ( )

inlineoverride

Virtual destructor

Member Function Documentation

_params()

const NonlinearOptimizerParams& gtsam::LevenbergMarquardtOptimizer::_params ( ) const

inlineoverrideprotectedvirtual

Access the parameters (base class version)

Implements gtsam::NonlinearOptimizer.

buildDampedSystem()

GaussianFactorGraph gtsam::LevenbergMarquardtOptimizer::buildDampedSystem	(	const GaussianFactorGraph &	linear,
		const VectorValues &	sqrtHessianDiagonal
	)		const

Build a damped system for a specific lambda – for testing only

defaultOptimize()

void gtsam::NonlinearOptimizer::defaultOptimize ( )

protectedinherited

A default implementation of the optimization loop, which calls iterate() until checkConvergence returns true.

iterate()

GaussianFactorGraph::shared_ptr gtsam::LevenbergMarquardtOptimizer::iterate ( )

overridevirtual

Perform a single iteration, returning GaussianFactorGraph corresponding to the linearized factor graph.

Implements gtsam::NonlinearOptimizer.

linearize()

virtual GaussianFactorGraph::shared_ptr gtsam::LevenbergMarquardtOptimizer::linearize ( ) const

virtual

linearize, can be overwritten

optimize()

virtual const Values& gtsam::NonlinearOptimizer::optimize ( )

inlinevirtualinherited

Optimize for the maximum-likelihood estimate, returning a the optimized variable assignments.

This function simply calls iterate() in a loop, checking for convergence with check_convergence(). For fine-grain control over the optimization process, you may call iterate() and check_convergence() yourself, and if needed modify the optimization state between iterations.

Reimplemented in gtsam::NonlinearConjugateGradientOptimizer.

optimizeSafely()

const Values& gtsam::NonlinearOptimizer::optimizeSafely ( )

inherited

Optimize, but return empty result if any uncaught exception is thrown Intended for MATLAB. In C++, use above and catch exceptions. No message is printed: it is up to the caller to check the result

Parameters

optimizer

a non-linear optimizer

params()

const LevenbergMarquardtParams& gtsam::LevenbergMarquardtOptimizer::params ( ) const

inline

Read-only access the parameters

solve()

virtual VectorValues gtsam::NonlinearOptimizer::solve ( const GaussianFactorGraph &  gfg, const NonlinearOptimizerParams &  params  ) const

virtualinherited

Default function to do linear solve, i.e. optimize a GaussianFactorGraph

tryLambda()

bool gtsam::LevenbergMarquardtOptimizer::tryLambda	(	const GaussianFactorGraph &	linear,
		const VectorValues &	sqrtHessianDiagonal
	)

Inner loop, changes state, returns true if successful or giving up

---

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

• /home/docs/checkouts/readthedocs.org/user_builds/gtsam-jlblanco-docs/checkouts/latest/gtsam/nonlinear/LevenbergMarquardtOptimizer.h