SplitMerge_SAMS Class Reference

Sequential Allocation Merge-Split (SAMS) sampler for Bayesian nonparametric models. More...

#include <splitmerge_SAMS.hpp>

Inheritance diagram for SplitMerge_SAMS:

Public Member Functions
	SplitMerge_SAMS (Data &d, Params &p, Likelihood &l, Process &pr, bool shuffle)
	Constructor for SAMS (Sequential Allocation Merge-Split) sampler.
void	step () override
	Perform one iteration of the SAMS algorithm.
int	get_accepted_split () const
	Get number of accepted split moves for diagnostics.
int	get_accepted_merge () const
	Get number of accepted merge moves for diagnostics.
int	get_accepted_shuffle () const
	Get number of accepted shuffle moves for diagnostics.
Public Member Functions inherited from Sampler
	Sampler (Data &d, Params &p, Likelihood &l, Process &pr)
	Constructor initializing sampler with required components.
virtual	~Sampler ()=default
	Virtual destructor for proper cleanup of derived classes.

Additional Inherited Members
Protected Attributes inherited from Sampler
Data &	data
	Reference to the data object containing observations and current allocations.
Params &	params
	Reference to the parameters object containing model hyperparameters and MCMC settings.
Likelihood &	likelihood
	Reference to the likelihood computation object for evaluating cluster assignments.
Process &	process
	Reference to the stochastic process object (DP, NGGP, DPW, NGGPW)
std::random_device	rd
	Random device for generating random numbers across sampling algorithms.

Detailed Description

Sequential Allocation Merge-Split (SAMS) sampler for Bayesian nonparametric models.

This class implements the SAMS algorithm, a variant of the split-merge sampler that uses sequential allocation instead of restricted Gibbs sampling for proposal generation. SAMS can be more efficient than standard split-merge for certain model configurations and provides an alternative proposal mechanism.

Key differences from standard Split-Merge:

• Sequential Allocation: Observations are allocated one by one in random order
• Simplified Proposals: No need for full restricted Gibbs chains
• Efficient Computation: Faster proposal generation for large clusters
• Maintained Ergodicity: Preserves theoretical properties of split-merge

The algorithm maintains the same three types of moves (split, merge, shuffle) but uses a different mechanism for generating proposals within each move type.

@reference Dahl, D. B. and Newcomb, S. (2022). "Sequentially allocated merge-split samplers for conjugate Bayesian nonparametric models" @reference Martinez, A. F. and Mena, R. H. (2014). "On a Nonparametric Change Point Detection Model in Markovian Regimes"

See also

Sampler, SplitMerge

Constructor & Destructor Documentation

SplitMerge_SAMS()

SplitMerge_SAMS::SplitMerge_SAMS ( Data & d, Params & p, Likelihood & l, Process & pr, bool shuffle )

inline

Constructor for SAMS (Sequential Allocation Merge-Split) sampler.

Parameters

d	Reference to Data object containing observations
p	Reference to Params object with hyperparameters
l	Reference to Likelihood object for probability computations
pr	Reference to Process object defining the prior
shuffle	Flag to enable shuffle moves in addition to split-merge

Initializes the SAMS sampler, which uses sequential allocation instead of restricted Gibbs sampling for generating proposals. This can provide computational advantages for certain model configurations.

Member Function Documentation

get_accepted_merge()

int SplitMerge_SAMS::get_accepted_merge ( ) const

inline

Get number of accepted merge moves for diagnostics.

Returns

Count of accepted merge moves

get_accepted_shuffle()

int SplitMerge_SAMS::get_accepted_shuffle ( ) const

inline

Get number of accepted shuffle moves for diagnostics.

Returns

Count of accepted shuffle moves

get_accepted_split()

int SplitMerge_SAMS::get_accepted_split ( ) const

inline

Get number of accepted split moves for diagnostics.

Returns

Count of accepted split moves

step()

void SplitMerge_SAMS::step ( )

overridevirtual

Perform one iteration of the SAMS algorithm.

Executes one step of the SAMS sampler:

1. Randomly select two observations as anchors
2. Determine move type based on their current assignments
3. Generate proposal using sequential allocation
4. Compute acceptance ratio and accept/reject the proposal

The sequential allocation approach can be more efficient than restricted Gibbs sampling while maintaining the theoretical properties of split-merge.

Perform a single split-merge MCMC step. Randomly choose two indices and decide whether to propose a split or merge move.

Implements Sampler.

---

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

• src/samplers/splitmerge_SAMS.hpp
• src/samplers/splitmerge_SAMS.cpp