|
noparama
v0.0.1
Nonparametric Bayesian models
|
#include <np_neal_algorithm8.h>


Public Member Functions | |
| NealAlgorithm8 (random_engine_t &generator, distribution_t &likelihood, dirichlet_process &nonparametrics) | |
| void | update (membertrix &cluster_matrix, const data_ids_t &data_ids) |
| void | printStatistics () |
Public Member Functions inherited from UpdateClusterPopulation | |
| UpdateClusterPopulation () | |
| virtual | ~UpdateClusterPopulation () |
Additional Inherited Members | |
Protected Attributes inherited from UpdateClusterPopulation | |
| std::default_random_engine | _generator |
| std::uniform_real_distribution< double > | _distribution |
| char | _verbosity |
This class NealAlgorithm8 updates the cluster population assuming a dirichlet process as nonparametric prior. It is a so-called auxiliary variables method. There is not just one new cluster proposed, but several at once. This number
| NealAlgorithm8::NealAlgorithm8 | ( | random_engine_t & | generator, |
| distribution_t & | likelihood, | ||
| dirichlet_process & | nonparametrics | ||
| ) |
Construct update method for cluster population.
| [in] | generator | Random generator |
| [in] | likelihood | Likelihood function to be used in update() |
| [in] | nonparametrics | Sufficient statistics of the nonparametric prior (e.g. Dirichlet) |
|
virtual |
MCMC statistics. How many times are steps rejected, etc.
Implements UpdateClusterPopulation.
|
virtual |
Update the cluster population. The observation has to be deleted beforehand.
| [in,out] | cluster_matrix | Cluster-observation membership matrix |
| [in] | data_ids | Observations to be considered for existing and new cluster |
Implements UpdateClusterPopulation.
1.8.13