NGGP clustering process with module-based covariates. More...

#include <NGGPx.hpp>

Inheritance diagram for NGGPx:

Public Member Functions
	NGGPx (Data &d, Params &p, U_sampler &U_sam, const std::vector< std::shared_ptr< Module > > &mods)
	Construct an NGGPWx process.
Gibbs Sampling Methods
double	gibbs_prior_existing_cluster (int cls_idx, int obs_idx=0) const override
	Computes the log prior probability of assigning a data point to an existing cluster.
Eigen::VectorXd	gibbs_prior_existing_clusters (int obs_idx) const override
	Computes the log prior for assigning an observation to each existing cluster.
double	gibbs_prior_new_cluster () const override
	Computes the log prior probability of assigning a data point to a new cluster.
double	gibbs_prior_new_cluster_obs (int obs_idx) const override
	Observation-specific new-cluster prior.
Split-Merge Algorithm Methods
double	prior_ratio_split (int ci, int cj) const override
	Computes the prior ratio for a split operation in a spatially-aware NGGP-based split-merge MCMC algorithm.
double	prior_ratio_merge (int size_old_ci, int size_old_cj) const override
	Computes the prior ratio for a merge operation in a spatially-aware NGGP-based split-merge MCMC algorithm.
double	prior_ratio_shuffle (int size_old_ci, int size_old_cj, int ci, int cj) const override
	Computes the prior ratio for a shuffle operation in a spatially-aware NGGP-based split-merge MCMC algorithm.
Parameter Update Methods
void	update_params () override
	Update process parameters.
Public Member Functions inherited from NGGP
	NGGP (Data &d, Params &p, U_sampler &mh)
	Constructor for the Normalized Generalized Gamma Process.
Public Member Functions inherited from Process
	Process (Data &d, const Params &p)
	Constructor initializing process with data and parameters.
void	set_old_allocations (const Eigen::VectorXi &new_allocations)
	Store current allocations for potential rollback.
void	set_old_cluster_members (const std::unordered_map< int, std::vector< int > > &new_cluster_members)
	Store current cluster members for potential rollback.
void	set_old_K (int new_K)
	Store current number of clusters for potential rollback.
void	restore_state ()
	Restores the cached state into the data object.
const Eigen::VectorXi &	old_allocations_view () const
	Provides read-only access to the stored previous allocations.
void	set_idx_i (int i)
	Set index of first observation in split-merge pair.
void	set_idx_j (int j)
	Set index of second observation in split-merge pair.
virtual	~Process ()
	Virtual destructor for proper cleanup of derived classes.

Protected Attributes
std::vector< std::shared_ptr< Module > >	modules
Protected Attributes inherited from NGGP
U_sampler &	U_sampler_method
	Reference to the U_sampler instance for updating the latent variable U.
std::random_device	rd
	Random device for seeding.
std::mt19937	gen
	Mersenne Twister random number generator.
Protected Attributes inherited from Process
Data &	data
	Reference to the data object containing observations and allocations.
const Params &	params
	Reference to the parameters object containing process hyperparameters.
Eigen::VectorXi	old_allocations
	Storage for previous allocations to enable rollback in case of rejection or for computation requiring it.
std::unordered_map< int, std::vector< int > >	old_cluster_members
	Storage for previous cluster members to enable rollback in case of rejection or for computation requiring it.
int	old_K = 0
	Number of clusters associated with the stored previous state.
int	idx_i
	Index of first observation involved in split-merge move.
int	idx_j
	Index of second observation involved in split-merge move.
const double	log_a = log(params.a)
	Precomputed logarithm of total mass parameter for efficiency.

Additional Inherited Members
Protected Member Functions inherited from Process
const std::unordered_map< int, std::vector< int > > &	old_cluster_members_view () const
	Provides read-only access to the stored previous cluster members.

Detailed Description

NGGP clustering process with module-based covariates.

This class extends the Normalized Generalized Gamma Process (NGGP) to incorporate module-based computations for covariates and other similarity terms. It augments Gibbs and split-merge moves with module contributions computed through the Module interface.

Constructor & Destructor Documentation

◆ NGGPx()

NGGPx::NGGPx	(	Data &	d,
		Params &	p,
		U_sampler &	U_sam,
		const std::vector< std::shared_ptr< Module > > &	mods )

inline

Construct an NGGPWx process.

Parameters

d	Data container.
p	Model and sampler parameters.
U_sam	Sampler for the latent NGGP auxiliary variable $U$.
mods	Vector of shared pointers to Module objects providing similarity computations.

Member Function Documentation

◆ gibbs_prior_existing_cluster()

double NGGPx::gibbs_prior_existing_cluster	(	int	cls_idx,
		int	obs_idx = 0 ) const

nodiscardoverridevirtual

Computes the log prior probability of assigning a data point to an existing cluster.

Combines the NGGP term for an existing cluster with spatial weighting and (when applicable in the implementation) covariate similarity contributions.

Parameters

cls_idx	Cluster index.
obs_idx	Observation index (used for observation-specific spatial/covariate terms).

Returns: Log prior (up to an additive constant) for assigning observation obs_idx to existing cluster cls_idx.

Reimplemented from NGGP.

◆ gibbs_prior_existing_clusters()

Eigen::VectorXd NGGPx::gibbs_prior_existing_clusters ( int obs_idx ) const

nodiscardoverridevirtual

Computes the log prior for assigning an observation to each existing cluster.

Parameters

obs_idx Observation index.

Returns: Vector of size $K$ containing log priors (up to an additive constant), one per existing cluster.

Reimplemented from NGGP.

◆ gibbs_prior_new_cluster()

double NGGPx::gibbs_prior_new_cluster ( ) const

nodiscardoverridevirtual

Computes the log prior probability of assigning a data point to a new cluster.

This is the baseline new-cluster mass from the NGGP model (without observation-specific covariate effects).

Returns: Log prior (up to an additive constant) for creating a new cluster.

Reimplemented from NGGP.

◆ gibbs_prior_new_cluster_obs()

double NGGPx::gibbs_prior_new_cluster_obs ( int obs_idx ) const

nodiscardoverridevirtual

Observation-specific new-cluster prior.

Extends the baseline new-cluster prior by including the covariate singleton similarity contribution for observation obs_idx.

Parameters

obs_idx Observation index.

Returns: Log prior (up to an additive constant) for assigning observation obs_idx to a newly created cluster.

Reimplemented from Process.

◆ prior_ratio_merge()

double NGGPx::prior_ratio_merge	(	int	size_old_ci,
		int	size_old_cj ) const

nodiscardoverridevirtual

Computes the prior ratio for a merge operation in a spatially-aware NGGP-based split-merge MCMC algorithm.

Accounts for the NGGP prior and spatial/covariate contributions (as implemented by NGGPW and CovariatesModule).

Parameters

size_old_ci	Size of the first cluster before the merge.
size_old_cj	Size of the second cluster before the merge.

Returns: Log prior ratio contribution for the merge move.

Reimplemented from NGGP.

◆ prior_ratio_shuffle()

double NGGPx::prior_ratio_shuffle	(	int	size_old_ci,
		int	size_old_cj,
		int	ci,
		int	cj ) const

nodiscardoverridevirtual

Computes the prior ratio for a shuffle operation in a spatially-aware NGGP-based split-merge MCMC algorithm.

Accounts for the NGGP prior and spatial/covariate contributions (as implemented by NGGPW and CovariatesModule).

Parameters

size_old_ci	Size of the first cluster before the shuffle.
size_old_cj	Size of the second cluster before the shuffle.
ci	Index of the first cluster involved in the shuffle.
cj	Index of the second cluster involved in the shuffle.

Returns: Log prior ratio contribution for the shuffle move.

Reimplemented from NGGP.

◆ prior_ratio_split()

double NGGPx::prior_ratio_split	(	int	ci,
		int	cj ) const

nodiscardoverridevirtual

Computes the prior ratio for a split operation in a spatially-aware NGGP-based split-merge MCMC algorithm.

Accounts for the NGGP prior and spatial/covariate contributions (as implemented by NGGPW and CovariatesModule).

Parameters

ci	Index of the first (proposed) split cluster.
cj	Index of the second (proposed) split cluster.

Returns: Log prior ratio contribution for the split move.

Reimplemented from NGGP.

◆ update_params()

void NGGPx::update_params ( )

inlineoverridevirtual

Update process parameters.

Delegates to the configured U_sampler implementation to update the latent auxiliary variable $U$ conditional on the current partition.

See also: U_sampler::update_U()

Reimplemented from NGGP.

Member Data Documentation

◆ modules

std::vector<std::shared_ptr<Module> > NGGPx::modules

protected

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

src/processes/NGGPx.hpp
src/processes/NGGPx.cpp

Public Member Functions

Protected Attributes

Additional Inherited Members

Detailed Description

Constructor & Destructor Documentation

◆ NGGPx()

Member Function Documentation

◆ gibbs_prior_existing_cluster()

◆ gibbs_prior_existing_clusters()

◆ gibbs_prior_new_cluster()

◆ gibbs_prior_new_cluster_obs()

◆ prior_ratio_merge()

◆ prior_ratio_shuffle()

◆ prior_ratio_split()

◆ update_params()

Member Data Documentation

◆ modules