ModErn Text Analysis
META Enumerates Textual Applications
Public Member Functions | Protected Member Functions | Protected Attributes | List of all members
meta::topics::lda_gibbs Class Reference

A LDA topic model implemented using a collapsed gibbs sampler. More...

#include <lda_gibbs.h>

Inheritance diagram for meta::topics::lda_gibbs:
meta::topics::lda_model meta::topics::parallel_lda_gibbs

Public Member Functions

 lda_gibbs (std::shared_ptr< index::forward_index > idx, uint64_t num_topics, double alpha, double beta)
 Constructs the lda model over the given documents, with the given number of topics, and hyperparameters \(\alpha\) and \(\beta\) for the priors on \(\phi\) (topic distributions) and \(\theta\) (topic proportions), respectively. More...
 
virtual ~lda_gibbs ()=default
 Destructor: virtual for potential subclassing.
 
virtual void run (uint64_t num_iters, double convergence=1e-6) override
 Runs the sampler for a maximum number of iterations, or until the given convergence criterion is met. More...
 
virtual double compute_term_topic_probability (term_id term, topic_id topic) const override
 
virtual double compute_doc_topic_probability (doc_id doc, topic_id topic) const override
 
- Public Member Functions inherited from meta::topics::lda_model
 lda_model (std::shared_ptr< index::forward_index > idx, uint64_t num_topics)
 Constructs an lda_model over the given set of documents and with a fixed number of topics. More...
 
virtual ~lda_model ()=default
 Destructor. More...
 
void save_doc_topic_distributions (const std::string &filename) const
 Saves the topic proportions \(\theta_d\) for each document to the given file. More...
 
void save_topic_term_distributions (const std::string &filename) const
 Saves the term distributions \(\phi_j\) for each topic to the given file. More...
 
void save (const std::string &prefix) const
 Saves the current model to a set of files beginning with prefix: prefix.phi, prefix.theta, and prefix.terms. More...
 
uint64_t num_topics () const
 

Protected Member Functions

topic_id sample_topic (term_id term, doc_id doc)
 Samples a topic from the full conditional distribution \(P(z_i = j | w, \boldsymbol{z})\). More...
 
virtual double compute_sampling_weight (term_id term, doc_id doc, topic_id topic) const
 Computes a weight proportional to \(P(z_i = j | w, \boldsymbol{z})\). More...
 
virtual void initialize ()
 Initializes the first set of topic assignments for inference. More...
 
virtual void perform_iteration (uint64_t iter, bool init=false)
 Performs a sampling iteration. More...
 
virtual void decrease_counts (topic_id topic, term_id term, doc_id doc)
 Decreases all counts associated with the given topic, term, and document by one. More...
 
virtual void increase_counts (topic_id topic, term_id term, doc_id doc)
 Increases all counts associated with the given topic, term, and document by one. More...
 
double corpus_log_likelihood () const
 
lda_gibbsoperator= (const lda_gibbs &)=delete
 lda_gibbs cannot be copy assigned.
 
 lda_gibbs (const lda_gibbs &other)=delete
 lda_gibbs cannot be copy constructed.
 
- Protected Member Functions inherited from meta::topics::lda_model
lda_modeloperator= (const lda_model &)=delete
 lda_models cannot be copy assigned.
 
 lda_model (const lda_model &)=delete
 lda_models cannot be copy constructed.
 

Protected Attributes

std::vector< std::vector< topic_id > > doc_word_topic_
 The topic assignment for every word in every document. More...
 
std::vector< stats::multinomial< term_id > > phi_
 The word distributions for each topic, \(\phi_t\).
 
std::vector< stats::multinomial< topic_id > > theta_
 The topic distributions for each document, \(\theta_d\).
 
std::mt19937_64 rng_
 The random number generator for the sampler.
 
- Protected Attributes inherited from meta::topics::lda_model
std::shared_ptr< index::forward_indexidx_
 The index containing the documents for the model.
 
size_t num_topics_
 The number of topics.
 
size_t num_words_
 The number of total unique words.
 

Detailed Description

A LDA topic model implemented using a collapsed gibbs sampler.

See also
http://www.pnas.org/content/101/suppl.1/5228.full.pdf

Constructor & Destructor Documentation

meta::topics::lda_gibbs::lda_gibbs ( std::shared_ptr< index::forward_index idx,
uint64_t  num_topics,
double  alpha,
double  beta 
)

Constructs the lda model over the given documents, with the given number of topics, and hyperparameters \(\alpha\) and \(\beta\) for the priors on \(\phi\) (topic distributions) and \(\theta\) (topic proportions), respectively.

Parameters
idxThe index that contains the documents to model
num_topicsThe number of topics to infer
alphaThe hyperparameter for the Dirichlet prior over \(\phi\)
betaThe hyperparameter for the Dirichlet prior over \(\theta\)

Member Function Documentation

void meta::topics::lda_gibbs::run ( uint64_t  num_iters,
double  convergence = 1e-6 
)
overridevirtual

Runs the sampler for a maximum number of iterations, or until the given convergence criterion is met.

The convergence criterion is determined as the relative difference in log corpus likelihood between two iterations.

Parameters
num_itersThe maximum number of iterations to run the sampler for
convergenceThe lowest relative difference in \(\log P(\mathbf{w} \mid \mathbf{z})\) to be allowed before considering the sampler to have converged

Implements meta::topics::lda_model.

double meta::topics::lda_gibbs::compute_term_topic_probability ( term_id  term,
topic_id  topic 
) const
overridevirtual
Returns
the probability that the given term appears in the given topic
Parameters
termThe term we are concerned with.
topicThe topic we are concerned with.

Implements meta::topics::lda_model.

double meta::topics::lda_gibbs::compute_doc_topic_probability ( doc_id  doc,
topic_id  topic 
) const
overridevirtual
Returns
the probability that the given topic is picked for the given document
Parameters
docThe document we are concerned with.
topicThe topic we are concerned with.

Implements meta::topics::lda_model.

topic_id meta::topics::lda_gibbs::sample_topic ( term_id  term,
doc_id  doc 
)
protected

Samples a topic from the full conditional distribution \(P(z_i = j | w, \boldsymbol{z})\).

Used in both initialization and each normal iteration of the sampler, after removing the current value of \(z_i\) from the vector of assignments \(\boldsymbol{z}\).

Parameters
termThe term we are sampling a topic assignment for
docThe document the term resides in
Returns
the topic sampled the given (term, doc) pair
double meta::topics::lda_gibbs::compute_sampling_weight ( term_id  term,
doc_id  doc,
topic_id  topic 
) const
protectedvirtual

Computes a weight proportional to \(P(z_i = j | w, \boldsymbol{z})\).

Parameters
termThe current word we are sampling for
docThe document in which the term resides
topicThe topic \(j\) we want to compute the probability for
Returns
a weight proportional to the probability that the given term in the given document belongs to the given topic

Reimplemented in meta::topics::parallel_lda_gibbs.

void meta::topics::lda_gibbs::initialize ( )
protectedvirtual

Initializes the first set of topic assignments for inference.

Employs an online application of the sampler where counts are only considered for the words observed so far through the loop.

Reimplemented in meta::topics::parallel_lda_gibbs.

void meta::topics::lda_gibbs::perform_iteration ( uint64_t  iter,
bool  init = false 
)
protectedvirtual

Performs a sampling iteration.

Parameters
iterThe iteration number
initWhether or not to employ the online method (defaults to false)

Reimplemented in meta::topics::parallel_lda_gibbs.

void meta::topics::lda_gibbs::decrease_counts ( topic_id  topic,
term_id  term,
doc_id  doc 
)
protectedvirtual

Decreases all counts associated with the given topic, term, and document by one.

Parameters
topicThe topic in question
termThe term in question
docThe document in question

Reimplemented in meta::topics::parallel_lda_gibbs.

void meta::topics::lda_gibbs::increase_counts ( topic_id  topic,
term_id  term,
doc_id  doc 
)
protectedvirtual

Increases all counts associated with the given topic, term, and document by one.

Parameters
topicThe topic in question
termThe term in question
docThe document in question

Reimplemented in meta::topics::parallel_lda_gibbs.

double meta::topics::lda_gibbs::corpus_log_likelihood ( ) const
protected
Returns
\(\log P(\mathbf{w} \mid \mathbf{z})\)

Member Data Documentation

std::vector<std::vector<topic_id> > meta::topics::lda_gibbs::doc_word_topic_
protected

The topic assignment for every word in every document.

Note that the same word occurring multiple times in one document could potentially have many different topics assigned to it, so we are not using term_ids here, but our own contrived intra document term id.

Indexed as [doc_id][position].


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