gtsam 4.1.1
gtsam
gtsam::NonlinearOptimizer Class Referenceabstract

Detailed Description

This is the abstract interface for classes that can optimize for the maximum-likelihood estimate of a NonlinearFactorGraph.

To use a class derived from this interface, construct the class with a NonlinearFactorGraph and an initial Values variable assignment. Next, call the optimize() method which returns the optimized variable assignment.

Simple and compact example:

// One-liner to do full optimization and use the result.
Values result = DoglegOptimizer(graph, initialValues).optimize();
const NonlinearFactorGraph & graph() const
return the graph with nonlinear factors
Definition: NonlinearOptimizer.h:118

Example exposing more functionality and details:

// Create initial optimizer
DoglegOptimizer optimizer(graph, initialValues);
// Run full optimization until convergence.
Values result = optimizer->optimize();
// The new optimizer has results and statistics
cout << "Converged in " << optimizer.iterations() << " iterations "
"with final error " << optimizer.error() << endl;

Example of setting parameters before optimization:

// Each derived optimizer type has its own parameters class, which inherits from NonlinearOptimizerParams
DoglegParams params;
params.factorization = DoglegParams::QR;
params.relativeErrorTol = 1e-3;
params.absoluteErrorTol = 1e-3;
// Optimize
Values result = DoglegOptimizer(graph, initialValues, params).optimize();

This interface also exposes an iterate() method, which performs one iteration. The optimize() method simply calls iterate() multiple times, until the error changes less than a threshold. We expose iterate() so that you can easily control what happens between iterations, such as drawing or printing, moving points from behind the camera to in front, etc.

For more flexibility you may override virtual methods in your own derived class.

+ Inheritance diagram for gtsam::NonlinearOptimizer:

Public Member Functions

Standard interface
virtual const Valuesoptimize ()
 Optimize for the maximum-likelihood estimate, returning a the optimized variable assignments. More...
 
const ValuesoptimizeSafely ()
 Optimize, but return empty result if any uncaught exception is thrown Intended for MATLAB. More...
 
double error () const
 return error in current optimizer state
 
size_t iterations () const
 return number of iterations in current optimizer state
 
const Valuesvalues () const
 return values in current optimizer state
 
const NonlinearFactorGraphgraph () const
 return the graph with nonlinear factors
 
Advanced interface
virtual ~NonlinearOptimizer ()
 Virtual destructor.
 
virtual VectorValues solve (const GaussianFactorGraph &gfg, const NonlinearOptimizerParams &params) const
 Default function to do linear solve, i.e. More...
 
virtual GaussianFactorGraph::shared_ptr iterate ()=0
 Perform a single iteration, returning GaussianFactorGraph corresponding to the linearized factor graph. More...
 

Public Types

using shared_ptr = boost::shared_ptr< const NonlinearOptimizer >
 A shared pointer to this class.
 

Protected Member Functions

void defaultOptimize ()
 A default implementation of the optimization loop, which calls iterate() until checkConvergence returns true.
 
virtual const NonlinearOptimizerParams_params () const =0
 
 NonlinearOptimizer (const NonlinearFactorGraph &graph, std::unique_ptr< internal::NonlinearOptimizerState > state)
 Constructor for initial construction of base classes. More...
 

Protected Attributes

NonlinearFactorGraph graph_
 The graph with nonlinear factors.
 
std::unique_ptr< internal::NonlinearOptimizerState > state_
 PIMPL'd state.
 

Constructor & Destructor Documentation

◆ NonlinearOptimizer()

gtsam::NonlinearOptimizer::NonlinearOptimizer ( const NonlinearFactorGraph graph,
std::unique_ptr< internal::NonlinearOptimizerState >  state 
)
protected

Constructor for initial construction of base classes.

Takes ownership of state.

Member Function Documentation

◆ _params()

virtual const NonlinearOptimizerParams & gtsam::NonlinearOptimizer::_params ( ) const
protectedpure virtual

◆ iterate()

virtual GaussianFactorGraph::shared_ptr gtsam::NonlinearOptimizer::iterate ( )
pure virtual

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

Implemented in gtsam::DoglegOptimizer, gtsam::GaussNewtonOptimizer, gtsam::LevenbergMarquardtOptimizer, and gtsam::NonlinearConjugateGradientOptimizer.

◆ optimize()

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

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 ( )

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
optimizera non-linear optimizer

◆ solve()

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

Default function to do linear solve, i.e.

optimize a GaussianFactorGraph


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