gtsam::HybridGaussianConditional Class Reference

Detailed Description

A conditional of gaussian conditionals indexed by discrete variables, as part of a Bayes Network.

a density over continuous variables given discrete/continuous parents.

This is the result of the elimination of a continuous variable in a hybrid scheme, such that the remaining variables are discrete+continuous.

Represents the conditional density P(X | M, Z) where X is the set of continuous random variables, M is the selection of discrete variables corresponding to a subset of the Gaussian variables and Z is parent of this node .

The probability P(x|y,z,...) is proportional to \( \sum_i k_i \exp - \frac{1}{2} |R_i x - (d_i - S_i y - T_i z - ...)|^2 \) where i indexes the components and k_i is a component-wise normalization constant.

• Symbolic factors, used to represent a graph structure, implemented in

Inheritance diagram for gtsam::HybridGaussianConditional:

Public Member Functions
Constructors
	HybridGaussianConditional ()=default
	Default constructor, mainly for serialization.
	HybridGaussianConditional (const DiscreteKey &discreteParent, const std::vector< GaussianConditional::shared_ptr > &conditionals)
	Construct from one discrete key and vector of conditionals.
	HybridGaussianConditional (const DiscreteKey &discreteParent, Key key, const std::vector< std::pair< Vector, double > > &parameters)
	Constructs a HybridGaussianConditional with means mu_i and standard deviations sigma_i.
	HybridGaussianConditional (const DiscreteKey &discreteParent, Key key, const Matrix &A, Key parent, const std::vector< std::pair< Vector, double > > &parameters)
	Constructs a HybridGaussianConditional with conditional means A × parent + b_i and standard deviations sigma_i.
	HybridGaussianConditional (const DiscreteKey &discreteParent, Key key, const Matrix &A1, Key parent1, const Matrix &A2, Key parent2, const std::vector< std::pair< Vector, double > > &parameters)
	Constructs a HybridGaussianConditional with conditional means A1 × parent1 + A2 × parent2 + b_i and standard deviations sigma_i.
	HybridGaussianConditional (const DiscreteKeys &discreteParents, const Conditionals &conditionals)
	Construct from multiple discrete keys and conditional tree.
	HybridGaussianConditional (const DiscreteKeys &discreteParents, const FactorValuePairs &pairs, bool pruned=false)
	Construct from multiple discrete keys M and a tree of factor/scalar pairs, where the scalar is assumed to be the the negative log constant for each assignment m, up to a constant.
Testable
bool	equals (const HybridFactor &lf, double tol=1e-9) const override
	Test equality with base HybridFactor.
void	print (const std::string &s="HybridGaussianConditional\n", const KeyFormatter &formatter=DefaultKeyFormatter) const override
	Print utility.
Standard API
GaussianConditional::shared_ptr	choose (const DiscreteValues &discreteValues) const
	Return the conditional Gaussian for the given discrete assignment.
GaussianConditional::shared_ptr	operator() (const DiscreteValues &discreteValues) const
	Syntactic sugar for choose.
size_t	nrComponents () const
	Returns the total number of continuous components.
KeyVector	continuousParents () const
	Returns the continuous keys among the parents.
double	negLogConstant () const override
	Return log normalization constant in negative log space.
std::shared_ptr< HybridGaussianFactor >	likelihood (const VectorValues &given) const
	Create a likelihood factor for a hybrid Gaussian conditional, return a hybrid Gaussian factor on the parents.
const Conditionals	conditionals () const
	Get Conditionals DecisionTree (dynamic cast from factors).
double	logProbability (const HybridValues &values) const override
	Compute the logProbability of this hybrid Gaussian conditional.
double	evaluate (const HybridValues &values) const override
	Calculate probability density for given values.
double	operator() (const HybridValues &values) const
	Evaluate probability density, sugar.
HybridGaussianConditional::shared_ptr	prune (const DiscreteConditional &discreteProbs) const
	Prune the decision tree of Gaussian factors as per the discrete discreteProbs.
bool	pruned () const
	Return true if the conditional has already been pruned.
std::shared_ptr< Factor >	restrict (const DiscreteValues &discreteValues) const override
	Restrict to the given discrete values.
Public Member Functions inherited from gtsam::HybridGaussianFactor
	HybridGaussianFactor ()=default
	Default constructor, mainly for serialization.
	HybridGaussianFactor (const DiscreteKey &discreteKey, const std::vector< GaussianFactor::shared_ptr > &factors)
	Construct a new HybridGaussianFactor on a single discrete key, providing the factors for each mode m as a vector of factors ϕ_m(x).
	HybridGaussianFactor (const DiscreteKey &discreteKey, const std::vector< GaussianFactorValuePair > &factorPairs)
	Construct a new HybridGaussianFactor on a single discrete key, including a scalar error value for each mode m.
	HybridGaussianFactor (const DiscreteKeys &discreteKeys, const FactorValuePairs &factorPairs)
	Construct a new HybridGaussianFactor on a several discrete keys M, including a scalar error value for each assignment m.
bool	equals (const HybridFactor &lf, double tol=1e-9) const override
	equals
void	print (const std::string &s="", const KeyFormatter &formatter=DefaultKeyFormatter) const override
	print
GaussianFactorValuePair	operator() (const DiscreteValues &assignment) const
	Get factor at a given discrete assignment.
AlgebraicDecisionTree< Key >	errorTree (const VectorValues &continuousValues) const override
	Compute error of the HybridGaussianFactor as a tree.
double	error (const HybridValues &hybridValues) const override
	Compute the log-likelihood, including the log-normalizing constant.
const FactorValuePairs &	factors () const
	Getter for GaussianFactor decision tree.
virtual HybridGaussianProductFactor	asProductFactor () const
	Helper function to return factors and functional to create a DecisionTree of Gaussian Factor Graphs.
std::shared_ptr< Factor >	restrict (const DiscreteValues &discreteValues) const override
	Restrict the factor to the given discrete values.
Public Member Functions inherited from gtsam::HybridFactor
	HybridFactor ()=default
	Default constructor creates empty factor.
	HybridFactor (const KeyVector &keys)
	Construct hybrid factor from continuous keys.
	HybridFactor (const DiscreteKeys &discreteKeys)
	Construct hybrid factor from discrete keys.
	HybridFactor (const KeyVector &continuousKeys, const DiscreteKeys &discreteKeys)
	Construct a new Hybrid Factor object.
bool	isDiscrete () const
	True if this is a factor of discrete variables only.
bool	isContinuous () const
	True if this is a factor of continuous variables only.
bool	isHybrid () const
	True is this is a Discrete-Continuous factor.
size_t	nrContinuous () const
	Return the number of continuous variables in this factor.
const DiscreteKeys &	discreteKeys () const
	Return the discrete keys for this factor.
const KeyVector &	continuousKeys () const
	Return only the continuous keys for this factor.
Public Member Functions inherited from gtsam::Factor
virtual	~Factor ()=default
	Default destructor.
bool	empty () const
	Whether the factor is empty (involves zero variables).
Key	front () const
	First key.
Key	back () const
	Last key.
const_iterator	find (Key key) const
	find
const KeyVector &	keys () const
	Access the factor's involved variable keys.
const_iterator	begin () const
	Iterator at beginning of involved variable keys.
const_iterator	end () const
	Iterator at end of involved variable keys.
size_t	size () const
virtual void	printKeys (const std::string &s="Factor", const KeyFormatter &formatter=DefaultKeyFormatter) const
	print only keys
bool	equals (const This &other, double tol=1e-9) const
	check equality
KeyVector &	keys ()
iterator	begin ()
	Iterator at beginning of involved variable keys.
iterator	end ()
	Iterator at end of involved variable keys.
Public Member Functions inherited from gtsam::Conditional< HybridGaussianFactor, HybridGaussianConditional >
void	print (const std::string &s="Conditional", const KeyFormatter &formatter=DefaultKeyFormatter) const
	print with optional formatter
bool	equals (const This &c, double tol=1e-9) const
	check equality
size_t	nrFrontals () const
	return the number of frontals
size_t	nrParents () const
	return the number of parents
Key	firstFrontalKey () const
	Convenience function to get the first frontal key.
Frontals	frontals () const
	return a view of the frontal keys
Parents	parents () const
	return a view of the parent keys
double	operator() (const HybridValues &x) const
	Evaluate probability density, sugar.
HybridGaussianFactor::const_iterator	beginFrontals () const
	Iterator pointing to first frontal key.
HybridGaussianFactor::const_iterator	endFrontals () const
	Iterator pointing past the last frontal key.
HybridGaussianFactor::const_iterator	beginParents () const
	Iterator pointing to the first parent key.
HybridGaussianFactor::const_iterator	endParents () const
	Iterator pointing past the last parent key.

Public Types
using	This = HybridGaussianConditional
using	shared_ptr = std::shared_ptr<This>
using	BaseFactor = HybridGaussianFactor
using	BaseConditional = Conditional<BaseFactor, HybridGaussianConditional>
using	Conditionals = DecisionTree<Key, GaussianConditional::shared_ptr>
	typedef for Decision Tree of Gaussian Conditionals
Public Types inherited from gtsam::HybridGaussianFactor
using	Base = HybridFactor
using	This = HybridGaussianFactor
using	shared_ptr = std::shared_ptr<This>
using	sharedFactor = std::shared_ptr<GaussianFactor>
using	FactorValuePairs = DecisionTree<Key, GaussianFactorValuePair>
	typedef for Decision Tree of Gaussian factors and arbitrary value.
Public Types inherited from gtsam::HybridFactor
enum class	Category { None , Discrete , Continuous , Hybrid }
	Enum to help with categorizing hybrid factors.
typedef HybridFactor	This
	This class.
typedef std::shared_ptr< HybridFactor >	shared_ptr
	shared_ptr to this class
typedef Factor	Base
	Our base class.
Public Types inherited from gtsam::Factor
typedef KeyVector::iterator	iterator
	Iterator over keys.
typedef KeyVector::const_iterator	const_iterator
	Const iterator over keys.
Public Types inherited from gtsam::Conditional< HybridGaussianFactor, HybridGaussianConditional >
typedef std::pair< typename HybridGaussianFactor::const_iterator, typename HybridGaussianFactor::const_iterator >	ConstFactorRange
	A mini implementation of an iterator range, to share const views of frontals and parents.
typedef ConstFactorRangeIterator	Frontals
	View of the frontal keys (call frontals()).
typedef ConstFactorRangeIterator	Parents
	View of the separator keys (call parents()).

Additional Inherited Members
static bool	CheckInvariants (const HybridGaussianConditional &conditional, const VALUES &x)
	Check invariants of this conditional, given the values x.
	Factor ()
	Default constructor for I/O.
template<typename CONTAINER>
	Factor (const CONTAINER &keys)
	Construct factor from container of keys.
template<typename ITERATOR>
	Factor (ITERATOR first, ITERATOR last)
	Construct factor from iterator keys.
Protected Member Functions inherited from gtsam::Conditional< HybridGaussianFactor, HybridGaussianConditional >
	Conditional ()
	Empty Constructor to make serialization possible.
template<typename CONTAINER>
static Factor	FromKeys (const CONTAINER &keys)
	Construct factor from container of keys.
template<typename ITERATOR>
static Factor	FromIterators (ITERATOR first, ITERATOR last)
	Construct factor from iterator keys.
Protected Attributes inherited from gtsam::HybridFactor
DiscreteKeys	discreteKeys_
KeyVector	continuousKeys_
	Record continuous keys for book-keeping.
Protected Attributes inherited from gtsam::Factor
KeyVector	keys_
	The keys involved in this factor.
Protected Attributes inherited from gtsam::Conditional< HybridGaussianFactor, HybridGaussianConditional >
size_t	nrFrontals_
	The first nrFrontal variables are frontal and the rest are parents.

Constructor & Destructor Documentation

HybridGaussianConditional() [1/6]

gtsam::HybridGaussianConditional::HybridGaussianConditional	(	const DiscreteKey &	discreteParent,
		const std::vector< GaussianConditional::shared_ptr > &	conditionals )

Construct from one discrete key and vector of conditionals.

Parameters

discreteParent	Single discrete parent variable
conditionals	Vector of conditionals with the same size as the cardinality of the discrete parent.

HybridGaussianConditional() [2/6]

gtsam::HybridGaussianConditional::HybridGaussianConditional	(	const DiscreteKey &	discreteParent,
		Key	key,
		const std::vector< std::pair< Vector, double > > &	parameters )

Constructs a HybridGaussianConditional with means mu_i and standard deviations sigma_i.

Parameters

discreteParent	The discrete parent or "mode" key.
key	The key for this conditional variable.
parameters	A vector of pairs (mu_i, sigma_i).

HybridGaussianConditional() [3/6]

gtsam::HybridGaussianConditional::HybridGaussianConditional	(	const DiscreteKey &	discreteParent,
		Key	key,
		const Matrix &	A,
		Key	parent,
		const std::vector< std::pair< Vector, double > > &	parameters )

Constructs a HybridGaussianConditional with conditional means A × parent + b_i and standard deviations sigma_i.

Parameters

discreteParent	The discrete parent or "mode" key.
key	The key for this conditional variable.
A	The matrix A.
parent	The key of the parent variable.
parameters	A vector of pairs (b_i, sigma_i).

HybridGaussianConditional() [4/6]

gtsam::HybridGaussianConditional::HybridGaussianConditional	(	const DiscreteKey &	discreteParent,
		Key	key,
		const Matrix &	A1,
		Key	parent1,
		const Matrix &	A2,
		Key	parent2,
		const std::vector< std::pair< Vector, double > > &	parameters )

Constructs a HybridGaussianConditional with conditional means A1 × parent1 + A2 × parent2 + b_i and standard deviations sigma_i.

Parameters

discreteParent	The discrete parent or "mode" key.
key	The key for this conditional variable.
A1	The first matrix.
parent1	The key of the first parent variable.
A2	The second matrix.
parent2	The key of the second parent variable.
parameters	A vector of pairs (b_i, sigma_i).

HybridGaussianConditional() [5/6]

gtsam::HybridGaussianConditional::HybridGaussianConditional	(	const DiscreteKeys &	discreteParents,
		const Conditionals &	conditionals )

Construct from multiple discrete keys and conditional tree.

Parameters

discreteParents	the discrete parents. Will be placed last.
conditionals	a decision tree of GaussianConditionals. The number of conditionals should be C^(number of discrete parents), where C is the cardinality of the DiscreteKeys in discreteParents, since the discreteParents will be used as the labels in the decision tree.

HybridGaussianConditional() [6/6]

gtsam::HybridGaussianConditional::HybridGaussianConditional	(	const DiscreteKeys &	discreteParents,
		const FactorValuePairs &	pairs,
		bool	pruned = false )

Construct from multiple discrete keys M and a tree of factor/scalar pairs, where the scalar is assumed to be the the negative log constant for each assignment m, up to a constant.

Note

Will throw if factors are not actually conditionals.

Parameters

discreteParents	the discrete parents. Will be placed last.
conditionalPairs	Decision tree of GaussianFactor/scalar pairs.
pruned	Flag indicating if conditional has been pruned.

Member Function Documentation

conditionals()

const HybridGaussianConditional::Conditionals gtsam::HybridGaussianConditional::conditionals

(

)

const

Get Conditionals DecisionTree (dynamic cast from factors).

Note

Slow: avoid using in favor of factors(), which uses existing tree.

equals()

bool gtsam::HybridGaussianConditional::equals ( const HybridFactor & lf, double tol = 1e-9 ) const

overridevirtual

Test equality with base HybridFactor.

Reimplemented from gtsam::HybridFactor.

evaluate()

double gtsam::HybridGaussianConditional::evaluate ( const HybridValues & values ) const

overridevirtual

Calculate probability density for given values.

Reimplemented from gtsam::Conditional< HybridGaussianFactor, HybridGaussianConditional >.

logProbability()

double gtsam::HybridGaussianConditional::logProbability ( const HybridValues & values ) const

overridevirtual

Compute the logProbability of this hybrid Gaussian conditional.

Parameters

values

Continuous values and discrete assignment.

Returns

double

Reimplemented from gtsam::Conditional< HybridGaussianFactor, HybridGaussianConditional >.

negLogConstant()

double gtsam::HybridGaussianConditional::negLogConstant ( ) const

inlineoverridevirtual

Return log normalization constant in negative log space.

The log normalization constant is the min of the individual log-normalization constants.

Returns

double

Reimplemented from gtsam::Conditional< HybridGaussianFactor, HybridGaussianConditional >.

print()

void gtsam::HybridGaussianConditional::print ( const std::string & s = "HybridGaussianConditional\n", const KeyFormatter & formatter = DefaultKeyFormatter ) const

overridevirtual

Print utility.

Reimplemented from gtsam::HybridFactor.

prune()

HybridGaussianConditional::shared_ptr gtsam::HybridGaussianConditional::prune

(

const DiscreteConditional &

discreteProbs

)

const

Prune the decision tree of Gaussian factors as per the discrete discreteProbs.

Parameters

discreteProbs

A pruned set of probabilities for the discrete keys.

Returns

Shared pointer to possibly a pruned HybridGaussianConditional

restrict()

std::shared_ptr< Factor > gtsam::HybridGaussianConditional::restrict ( const DiscreteValues & discreteValues ) const

overridevirtual

Restrict to the given discrete values.

Implements gtsam::HybridFactor.

---

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

• /tmp/gtsam-4.3.0-doxygen.rsXPUS/source/gtsam/hybrid/HybridGaussianConditional.h
• /tmp/gtsam-4.3.0-doxygen.rsXPUS/source/gtsam/hybrid/HybridGaussianConditional.cpp