Docs
Loading...
Searching...
No Matches
CategoricalCovariatesModule Class Reference

Module for categorical covariate-related computations within clustering processes. More...

#include <categorical_covariate_module.hpp>

Inheritance diagram for CategoricalCovariatesModule:
Module

Public Member Functions

 CategoricalCovariatesModule (const Data &data_, const Eigen::VectorXi categorical_covariate, std::vector< double > prior_alpha_, const Eigen::VectorXi *old_alloc_provider=nullptr, const std::unordered_map< int, std::vector< int > > *old_cluster_members_provider_=nullptr)
 Constructor for CategoricalCovariatesModule.
Similarity Computation Methods
double compute_similarity_cls (int cls_idx, bool old_allo=false) const override __attribute__((hot))
 Compute covariate similarity contribution for a cluster.
double compute_similarity_obs (int obs_idx, int cls_idx) const override __attribute__((hot))
 Compute covariate similarity for a single observation in a cluster.
Eigen::VectorXd compute_similarity_obs (int obs_idx) const override __attribute__((hot))
 Compute covariate similarity contributions for all existing clusters.
Public Member Functions inherited from Module
 Module (const Eigen::VectorXi *old_allocations_provider_=nullptr, const std::unordered_map< int, std::vector< int > > *old_cluster_members_provider_=nullptr)
void set_old_allocations_provider (const Eigen::VectorXi *provider)
void set_old_cluster_members_provider (const std::unordered_map< int, std::vector< int > > *provider)
virtual ~Module ()=default

Protected Attributes

Module References
const std::vector< double > prior_alpha
const Data & data
 Reference to data object with cluster assignments.
const Eigen::VectorXi categorical_covariate_data
Precomputed Values
const double alpha_0 = std::accumulate(prior_alpha.begin(), prior_alpha.end(), 0.0)
const double lgamma_alpha_0 = std::lgamma(alpha_0)
double prod_lgamma_prior = 0
Protected Attributes inherited from Module
const Eigen::VectorXi * old_allocations_provider
 Provider function for accessing old allocation state.
const std::unordered_map< int, std::vector< int > > * old_cluster_members_provider
 Provider function for accessing old cluster members map.

Detailed Description

Module for categorical covariate-related computations within clustering processes.

This class implements covariate-related computations for clustering processes where covariates are categorical ({1, ..., C}). It computes similarity measures based on the distribution of categorical covariates within clusters using a Dirichlet-Multinomial model.

alphas correspond to the concentration parameters for each category.

Constructor & Destructor Documentation

◆ CategoricalCovariatesModule()

CategoricalCovariatesModule::CategoricalCovariatesModule ( const Data & data_,
const Eigen::VectorXi categorical_covariate,
std::vector< double > prior_alpha_,
const Eigen::VectorXi * old_alloc_provider = nullptr,
const std::unordered_map< int, std::vector< int > > * old_cluster_members_provider_ = nullptr )
inline

Constructor for CategoricalCovariatesModule.

Parameters
data_Reference to data object with cluster assignments
categorical_covariateCategorical covariate values for observations
prior_alpha_Prior alpha parameters for Dirichlet-Multinomial model
old_alloc_providerOptional pointer to previous allocation provider
old_cluster_members_provider_Optional pointer to previous cluster members provider

Member Function Documentation

◆ compute_similarity_cls()

double CategoricalCovariatesModule::compute_similarity_cls ( int cls_idx,
bool old_allo = false ) const
overridevirtual

Compute covariate similarity contribution for a cluster.

Computes the log marginal likelihood of the covariates within a cluster under the Dirichlet-Multinomial conjugate model. Higher values indicate that observations in the cluster have similar covariate values. ...

The computation follows Müller et al. (2011):

  1. Compute log marginal likelihood using log-gamma functions

Implements Module.

◆ compute_similarity_obs() [1/2]

Eigen::VectorXd CategoricalCovariatesModule::compute_similarity_obs ( int obs_idx) const
overridevirtual

Compute covariate similarity contributions for all existing clusters.

Computes the predictive contributions for adding observation obs_idx to each existing cluster, considering covariate values.

Parameters
obs_idxIndex of the observation
Returns
Vector of log predictive density contributions for each cluster

Implements Module.

◆ compute_similarity_obs() [2/2]

double CategoricalCovariatesModule::compute_similarity_obs ( int obs_idx,
int cls_idx ) const
overridevirtual

Compute covariate similarity for a single observation in a cluster.

Computes the predictive contribution when adding observation obs_idx to cluster cls_idx, considering the covariate values.

Parameters
obs_idxIndex of the observation
cls_idxIndex of the cluster
Returns
Log predictive density contribution

Used in Gibbs sampling to compute the probability of assigning an observation to a cluster based on covariate similarity.

Implements Module.

Member Data Documentation

◆ alpha_0

const double CategoricalCovariatesModule::alpha_0 = std::accumulate(prior_alpha.begin(), prior_alpha.end(), 0.0)
protected

Sum of prior alphas

◆ categorical_covariate_data

const Eigen::VectorXi CategoricalCovariatesModule::categorical_covariate_data
protected

Categorical covariate values for observations

◆ data

const Data& CategoricalCovariatesModule::data
protected

Reference to data object with cluster assignments.

◆ lgamma_alpha_0

const double CategoricalCovariatesModule::lgamma_alpha_0 = std::lgamma(alpha_0)
protected

Log Gamma of sum of prior alphas

◆ prior_alpha

const std::vector<double> CategoricalCovariatesModule::prior_alpha
protected

Prior alpha parameters for Dirichlet-Multinomial model

◆ prod_lgamma_prior

double CategoricalCovariatesModule::prod_lgamma_prior = 0
protected

Sum of log Gamma of prior alphas

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