Likelihood Class Reference

Abstract base class for likelihood computation. More...

#include <Likelihood.hpp>

Inheritance diagram for Likelihood:

Public Member Functions
	Likelihood (const Data &data)
virtual double	cluster_loglikelihood (int cluster_index) const =0
	Computes the log-likelihood for a cluster.
virtual double	cluster_loglikelihood (int cluster_index, const Eigen::Ref< const Eigen::VectorXi > &cls_ass_k) const =0
	Computes the log-likelihood for a cluster with given assignments.
virtual double	clusters_loglikelihood (int c1, int c2) const
	Computes the joint log-likelihood for two clusters.
virtual double	point_loglikelihood_cond (int point_index, int cluster_index) const =0
	Conditional log-likelihood of a point in a particular cluster.
virtual	~Likelihood ()=default

Protected Attributes
const Data &	data
	Reference to Data object with distances and allocations.

Detailed Description

Abstract base class for likelihood computation.

This class defines the interface for computing likelihoods in clustering models. Derived classes must implement methods for computing both cluster-level and point-level conditional log-likelihoods, which are essential for Gibbs and split-merge MCMC algorithms.

Constructor & Destructor Documentation

Likelihood()

Likelihood::Likelihood ( const Data & data )

inline

~Likelihood()

virtual Likelihood::~Likelihood ( )

virtualdefault

Member Function Documentation

cluster_loglikelihood() [1/2]

virtual double Likelihood::cluster_loglikelihood ( int cluster_index ) const

pure virtual

Computes the log-likelihood for a cluster.

Parameters

cluster_index

Index of the cluster to evaluate

Returns

Total log-likelihood of the cluster

Note

Useful for split-merge algorithms

Implemented in Gamma_likelihood, Gamma_likelihood_distances, GaussianMixtureModel_likelihood, Natarajan_likelihood, Natarajan_likelihood_summaryStats, and Null_likelihood.

cluster_loglikelihood() [2/2]

virtual double Likelihood::cluster_loglikelihood ( int cluster_index, const Eigen::Ref< const Eigen::VectorXi > & cls_ass_k ) const

pure virtual

Computes the log-likelihood for a cluster with given assignments.

Parameters

cluster_index	Index of the cluster to evaluate
cls_ass_k	Vector of point indices in the cluster

Returns

Total log-likelihood of the cluster

Note

Useful for split-merge algorithms

Implemented in Gamma_likelihood, Gamma_likelihood_distances, GaussianMixtureModel_likelihood, Natarajan_likelihood, Natarajan_likelihood_summaryStats, and Null_likelihood.

clusters_loglikelihood()

virtual double Likelihood::clusters_loglikelihood ( int c1, int c2 ) const

inlinevirtual

Computes the joint log-likelihood for two clusters.

Parameters

c1	Index of first cluster
c2	Index of second cluster

Returns

Joint log-likelihood of clusters c1 and c2

Note

Default implementation assumes cluster independence. Derived classes with inter-cluster interactions can override this.

Reimplemented in Natarajan_likelihood, and Natarajan_likelihood_summaryStats.

point_loglikelihood_cond()

virtual double Likelihood::point_loglikelihood_cond ( int point_index, int cluster_index ) const

pure virtual

Conditional log-likelihood of a point in a particular cluster.

Parameters

point_index	Index of the point to evaluate
cluster_index	Index of the cluster

Returns

Conditional log-likelihood

Note

Useful for Gibbs sampling

Implemented in Gamma_likelihood, Gamma_likelihood_distances, GaussianMixtureModel_likelihood, Natarajan_likelihood, Natarajan_likelihood_summaryStats, and Null_likelihood.

Member Data Documentation

data

const Data& Likelihood::data

protected

Reference to Data object with distances and allocations.

---

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

• BNPClust/utils/Likelihood.hpp