gtsam::EliminateableFactorGraph< FACTORGRAPH > Class Template Reference

Detailed Description

template<class FACTORGRAPH>
class gtsam::EliminateableFactorGraph< FACTORGRAPH >

EliminateableFactorGraph is a base class for factor graphs that contains elimination algorithms.

Any factor graph holding eliminateable factors can derive from this class to expose functions for computing marginals, conditional marginals, doing multifrontal and sequential elimination, etc.

Public Member Functions
boost::shared_ptr< BayesNetType >	eliminateSequential (OptionalOrderingType orderingType=boost::none, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Do sequential elimination of all variables to produce a Bayes net. More...
boost::shared_ptr< BayesNetType >	eliminateSequential (const Ordering &ordering, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Do sequential elimination of all variables to produce a Bayes net. More...
boost::shared_ptr< BayesTreeType >	eliminateMultifrontal (OptionalOrderingType orderingType=boost::none, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Do multifrontal elimination of all variables to produce a Bayes tree. More...
boost::shared_ptr< BayesTreeType >	eliminateMultifrontal (const Ordering &ordering, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Do multifrontal elimination of all variables to produce a Bayes tree. More...
std::pair< boost::shared_ptr< BayesNetType >, boost::shared_ptr< FactorGraphType > >	eliminatePartialSequential (const Ordering &ordering, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Do sequential elimination of some variables, in ordering provided, to produce a Bayes net and a remaining factor graph. More...
std::pair< boost::shared_ptr< BayesNetType >, boost::shared_ptr< FactorGraphType > >	eliminatePartialSequential (const KeyVector &variables, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Do sequential elimination of the given variables in an ordering computed by COLAMD to produce a Bayes net and a remaining factor graph. More...
std::pair< boost::shared_ptr< BayesTreeType >, boost::shared_ptr< FactorGraphType > >	eliminatePartialMultifrontal (const Ordering &ordering, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Do multifrontal elimination of some variables, in ordering provided, to produce a Bayes tree and a remaining factor graph. More...
std::pair< boost::shared_ptr< BayesTreeType >, boost::shared_ptr< FactorGraphType > >	eliminatePartialMultifrontal (const KeyVector &variables, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Do multifrontal elimination of the given variables in an ordering computed by COLAMD to produce a Bayes net and a remaining factor graph. More...
boost::shared_ptr< BayesNetType >	marginalMultifrontalBayesNet (boost::variant< const Ordering &, const KeyVector & > variables, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Compute the marginal of the requested variables and return the result as a Bayes net. More...
boost::shared_ptr< BayesNetType >	marginalMultifrontalBayesNet (boost::variant< const Ordering &, const KeyVector & > variables, const Ordering &marginalizedVariableOrdering, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Compute the marginal of the requested variables and return the result as a Bayes net. More...
boost::shared_ptr< BayesTreeType >	marginalMultifrontalBayesTree (boost::variant< const Ordering &, const KeyVector & > variables, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Compute the marginal of the requested variables and return the result as a Bayes tree. More...
boost::shared_ptr< BayesTreeType >	marginalMultifrontalBayesTree (boost::variant< const Ordering &, const KeyVector & > variables, const Ordering &marginalizedVariableOrdering, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Compute the marginal of the requested variables and return the result as a Bayes tree. More...
boost::shared_ptr< FactorGraphType >	marginal (const KeyVector &variables, const Eliminate &function=EliminationTraitsType::DefaultEliminate, OptionalVariableIndex variableIndex=boost::none) const
	Compute the marginal factor graph of the requested variables.

Public Types
typedef EliminationTraits< FactorGraphType >	EliminationTraitsType
	Typedef to the specific EliminationTraits for this graph.
typedef EliminationTraitsType::ConditionalType	ConditionalType
	Conditional type stored in the Bayes net produced by elimination.
typedef EliminationTraitsType::BayesNetType	BayesNetType
	Bayes net type produced by sequential elimination.
typedef EliminationTraitsType::EliminationTreeType	EliminationTreeType
	Elimination tree type that can do sequential elimination of this graph.
typedef EliminationTraitsType::BayesTreeType	BayesTreeType
	Bayes tree type produced by multifrontal elimination.
typedef EliminationTraitsType::JunctionTreeType	JunctionTreeType
	Junction tree type that can do multifrontal elimination of this graph.
typedef std::pair< boost::shared_ptr< ConditionalType >, boost::shared_ptr< _FactorType > >	EliminationResult
	The pair of conditional and remaining factor produced by a single dense elimination step on a subgraph.
typedef std::function< EliminationResult(const FactorGraphType &, const Ordering &)>	Eliminate
	The function type that does a single dense elimination step on a subgraph.
typedef boost::optional< const VariableIndex & >	OptionalVariableIndex
	Typedef for an optional variable index as an argument to elimination functions.
typedef boost::optional< Ordering::OrderingType >	OptionalOrderingType
	Typedef for an optional ordering type.

Member Function Documentation

eliminateMultifrontal() [1/2]

template<class FACTORGRAPH >

boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesTreeType > gtsam::EliminateableFactorGraph< FACTORGRAPH >::eliminateMultifrontal	(	const Ordering &	ordering,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Do multifrontal elimination of all variables to produce a Bayes tree.

If an ordering is not provided, the ordering will be computed using either COLAMD or METIS, dependeing on the parameter orderingType (Ordering::COLAMD or Ordering::METIS)

Example - Full QR elimination in specified order:

boost::shared_ptr<GaussianBayesTree> result = graph.eliminateMultifrontal(EliminateQR, myOrdering);

eliminateMultifrontal() [2/2]

template<class FACTORGRAPH >

boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesTreeType > gtsam::EliminateableFactorGraph< FACTORGRAPH >::eliminateMultifrontal	(	OptionalOrderingType	orderingType = boost::none,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Do multifrontal elimination of all variables to produce a Bayes tree.

If an ordering is not provided, the ordering will be computed using either COLAMD or METIS, dependeing on the parameter orderingType (Ordering::COLAMD or Ordering::METIS)

Example - Full Cholesky elimination in COLAMD order:

boost::shared_ptr<GaussianBayesTree> result = graph.eliminateMultifrontal(EliminateCholesky);

Example - Reusing an existing VariableIndex to improve performance, and using COLAMD ordering:

VariableIndex varIndex(graph); // Build variable index
Data data = otherFunctionUsingVariableIndex(graph, varIndex); // Other code that uses variable index
boost::shared_ptr<GaussianBayesTree> result = graph.eliminateMultifrontal(EliminateQR, boost::none, varIndex);

eliminatePartialMultifrontal() [1/2]

template<class FACTORGRAPH >

std::pair< boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesTreeType >, boost::shared_ptr< FACTORGRAPH > > gtsam::EliminateableFactorGraph< FACTORGRAPH >::eliminatePartialMultifrontal	(	const KeyVector &	variables,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Do multifrontal elimination of the given variables in an ordering computed by COLAMD to produce a Bayes net and a remaining factor graph.

This computes the factorization \( p(X) = p(A|B) p(B) \), where \( A = \) variables, \( X \) is all the variables in the factor graph, and \( B = X\backslash A \).

eliminatePartialMultifrontal() [2/2]

template<class FACTORGRAPH >

std::pair< boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesTreeType >, boost::shared_ptr< FACTORGRAPH > > gtsam::EliminateableFactorGraph< FACTORGRAPH >::eliminatePartialMultifrontal	(	const Ordering &	ordering,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Do multifrontal elimination of some variables, in ordering provided, to produce a Bayes tree and a remaining factor graph.

This computes the factorization \( p(X) = p(A|B) p(B) \), where \( A = \) variables, \( X \) is all the variables in the factor graph, and \( B = X\backslash A \).

eliminatePartialSequential() [1/2]

template<class FACTORGRAPH >

std::pair< boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesNetType >, boost::shared_ptr< FACTORGRAPH > > gtsam::EliminateableFactorGraph< FACTORGRAPH >::eliminatePartialSequential	(	const KeyVector &	variables,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Do sequential elimination of the given variables in an ordering computed by COLAMD to produce a Bayes net and a remaining factor graph.

This computes the factorization \( p(X) = p(A|B) p(B) \), where \( A = \) variables, \( X \) is all the variables in the factor graph, and \( B = X\backslash A \).

eliminatePartialSequential() [2/2]

template<class FACTORGRAPH >

std::pair< boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesNetType >, boost::shared_ptr< FACTORGRAPH > > gtsam::EliminateableFactorGraph< FACTORGRAPH >::eliminatePartialSequential	(	const Ordering &	ordering,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Do sequential elimination of some variables, in ordering provided, to produce a Bayes net and a remaining factor graph.

This computes the factorization \( p(X) = p(A|B) p(B) \), where \( A = \) variables, \( X \) is all the variables in the factor graph, and \( B = X\backslash A \).

eliminateSequential() [1/2]

template<class FACTORGRAPH >

boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesNetType > gtsam::EliminateableFactorGraph< FACTORGRAPH >::eliminateSequential	(	const Ordering &	ordering,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Do sequential elimination of all variables to produce a Bayes net.

Example - Full QR elimination in specified order:

boost::shared_ptr<GaussianBayesNet> result = graph.eliminateSequential(myOrdering, EliminateQR);

Example - Reusing an existing VariableIndex to improve performance:

VariableIndex varIndex(graph); // Build variable index
Data data = otherFunctionUsingVariableIndex(graph, varIndex); // Other code that uses variable index
boost::shared_ptr<GaussianBayesNet> result = graph.eliminateSequential(myOrdering, EliminateQR, varIndex, boost::none);

eliminateSequential() [2/2]

template<class FACTORGRAPH >

boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesNetType > gtsam::EliminateableFactorGraph< FACTORGRAPH >::eliminateSequential	(	OptionalOrderingType	orderingType = boost::none,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Do sequential elimination of all variables to produce a Bayes net.

If an ordering is not provided, the ordering provided by COLAMD will be used.

Example - Full Cholesky elimination in COLAMD order:

boost::shared_ptr<GaussianBayesNet> result = graph.eliminateSequential(EliminateCholesky);

Example - METIS ordering for elimination

boost::shared_ptr<GaussianBayesNet> result = graph.eliminateSequential(OrderingType::METIS);

Example - Reusing an existing VariableIndex to improve performance, and using COLAMD ordering:

VariableIndex varIndex(graph); // Build variable index
Data data = otherFunctionUsingVariableIndex(graph, varIndex); // Other code that uses variable index
boost::shared_ptr<GaussianBayesNet> result = graph.eliminateSequential(EliminateQR, varIndex, boost::none);

marginalMultifrontalBayesNet() [1/2]

template<class FACTORGRAPH >

boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesNetType > gtsam::EliminateableFactorGraph< FACTORGRAPH >::marginalMultifrontalBayesNet	(	boost::variant< const Ordering &, const KeyVector & >	variables,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Compute the marginal of the requested variables and return the result as a Bayes net.

Uses COLAMD marginalization ordering by default

Parameters

variables	Determines the variables whose marginal to compute, if provided as an Ordering they will be ordered in the returned BayesNet as specified, and if provided as a KeyVector they will be ordered using constrained COLAMD.
function	Optional dense elimination function, if not provided the default will be used.
variableIndex	Optional pre-computed VariableIndex for the factor graph, if not provided one will be computed.

marginalMultifrontalBayesNet() [2/2]

template<class FACTORGRAPH >

boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesNetType > gtsam::EliminateableFactorGraph< FACTORGRAPH >::marginalMultifrontalBayesNet	(	boost::variant< const Ordering &, const KeyVector & >	variables,
		const Ordering &	marginalizedVariableOrdering,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Compute the marginal of the requested variables and return the result as a Bayes net.

Parameters

variables	Determines the variables whose marginal to compute, if provided as an Ordering they will be ordered in the returned BayesNet as specified, and if provided as a KeyVector they will be ordered using constrained COLAMD.
marginalizedVariableOrdering	Ordering for the variables being marginalized out, i.e. all variables not in variables.
function	Optional dense elimination function, if not provided the default will be used.
variableIndex	Optional pre-computed VariableIndex for the factor graph, if not provided one will be computed.

marginalMultifrontalBayesTree() [1/2]

template<class FACTORGRAPH >

boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesTreeType > gtsam::EliminateableFactorGraph< FACTORGRAPH >::marginalMultifrontalBayesTree	(	boost::variant< const Ordering &, const KeyVector & >	variables,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Compute the marginal of the requested variables and return the result as a Bayes tree.

Uses COLAMD marginalization order by default

Parameters

variables	Determines the variables whose marginal to compute, if provided as an Ordering they will be ordered in the returned BayesNet as specified, and if provided as a KeyVector they will be ordered using constrained COLAMD.
function	Optional dense elimination function, if not provided the default will be used.
variableIndex	Optional pre-computed VariableIndex for the factor graph, if not provided one will be computed.

marginalMultifrontalBayesTree() [2/2]

template<class FACTORGRAPH >

boost::shared_ptr< typename EliminateableFactorGraph< FACTORGRAPH >::BayesTreeType > gtsam::EliminateableFactorGraph< FACTORGRAPH >::marginalMultifrontalBayesTree	(	boost::variant< const Ordering &, const KeyVector & >	variables,
		const Ordering &	marginalizedVariableOrdering,
		const Eliminate &	function = EliminationTraitsType::DefaultEliminate,
		OptionalVariableIndex	variableIndex = boost::none
	)		const

Compute the marginal of the requested variables and return the result as a Bayes tree.

Parameters

variables	Determines the variables whose marginal to compute, if provided as an Ordering they will be ordered in the returned BayesNet as specified, and if provided as a KeyVector they will be ordered using constrained COLAMD.
marginalizedVariableOrdering	Ordering for the variables being marginalized out, i.e. all variables not in variables.
function	Optional dense elimination function, if not provided the default will be used.
variableIndex	Optional pre-computed VariableIndex for the factor graph, if not provided one will be computed.

---

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

• /Users/dellaert/git/github/gtsam/inference/EliminateableFactorGraph.h
• /Users/dellaert/git/github/gtsam/inference/EliminateableFactorGraph-inst.h