Likelihood Class Reference

Computes log-likelihood for clusters based on distance-based cohesion and repulsion. More...

#include <Likelihood.hpp>

Public Member Functions
	Likelihood (const Data &data, const Params &param)
	Constructs a Likelihood object.
double	cluster_loglikelihood (int cluster_index) const
	Calculates the full log-likelihood of a cluster.
double	cluster_loglikelihood (int cluster_index, const Eigen::VectorXi &cls_ass_k) const
	Calculates the full log-likelihood of a cluster with explicit assignments.
double	point_loglikelihood_cond (int point_index, int cluster_index) const
	Calculates the conditional log-likelihood of assigning a point to a cluster.
double	point_loglikelihood_cond (int point_index, int cluster_index, const Eigen::VectorXi &cls_ass_k) const
	Calculates the conditional log-likelihood with explicit cluster assignments.

Detailed Description

Computes log-likelihood for clusters based on distance-based cohesion and repulsion.

This class implements a likelihood model that encourages points within clusters to be close (cohesion) while clusters are pushed apart (repulsion). The model uses gamma priors on distance distributions.

Constructor & Destructor Documentation

Likelihood()

Likelihood::Likelihood ( const Data & data, const Params & param )

inline

Constructs a Likelihood object.

Parameters

data	Reference to Data object with distances and allocations
param	Reference to model parameters

Member Function Documentation

cluster_loglikelihood() [1/2]

double Likelihood::cluster_loglikelihood

(

int

cluster_index

)

const

Calculates the full log-likelihood of a cluster.

Computes both cohesion (within-cluster) and repulsion (between-cluster) components for all points in the specified cluster

Parameters

cluster_index

Index of the cluster to evaluate

Returns

Total log-likelihood of the cluster

cluster_loglikelihood() [2/2]

double Likelihood::cluster_loglikelihood	(	int	cluster_index,
		const Eigen::VectorXi &	cls_ass_k ) const

Calculates the full log-likelihood of a cluster with explicit assignments.

Overloaded version that accepts cluster assignments directly

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

point_loglikelihood_cond() [1/2]

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

Calculates the conditional log-likelihood of assigning a point to a cluster.

Computes how likely a point is to belong to a cluster, considering both its cohesion with the cluster and repulsion from other clusters

Parameters

point_index	Index of the point to evaluate
cluster_index	Index of the target cluster (can be K for new cluster)

Returns

Conditional log-likelihood of the point given the cluster

point_loglikelihood_cond() [2/2]

double Likelihood::point_loglikelihood_cond	(	int	point_index,
		int	cluster_index,
		const Eigen::VectorXi &	cls_ass_k ) const

Calculates the conditional log-likelihood with explicit cluster assignments.

Overloaded version that accepts cluster assignments directly

Parameters

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

Returns

Conditional log-likelihood of the point given the cluster

---

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

• src/utils/Likelihood.hpp
• src/utils/Likelihood.cpp