Summarize Poisson NMF or Multinomial Topic Model Fit — summary.poisson_nmf

summary method for the “poisson_nmf_fit” and “multinom_topic_model_fit” classes.

# S3 method for poisson_nmf_fit
summary(object, ...)

# S3 method for multinom_topic_model_fit
summary(object, ...)

# S3 method for summary.poisson_nmf_fit
print(x, show.mixprops = FALSE, show.topic.reps = FALSE, ...)

# S3 method for summary.multinom_topic_model_fit
print(
  x,
  show.size.factors = FALSE,
  show.mixprops = FALSE,
  show.topic.reps = FALSE,
  ...
)

Arguments

object: An object of class “poisson_nmf_fit” or “multinom_topic_model_fit”. The former is usually the result of calling fit_poisson_nmf; the latter is usually the result of calling fit_topic_model or poisson2multinom.
...: Additional arguments passed to the generic summary or print.summary method.
x: An object of class “summary.poisson_nmf_fit”, usually a result of a call to summary.poisson_nmf_fit.
show.mixprops: If TRUE, print a summary of the mixture proportions.
show.topic.reps: If TRUE, print a summary of the topic representatives.
show.size.factors: If TRUE, print a summary of the size factors.

Value

The functions summary.poisson_nmf_fit and summary.multinom_topic_model_fit compute and return a list of statistics summarizing the model fit. The returned list includes some or all of the following elements:

n: The number of rows in the counts matrix, typically the number of samples.
m: The number of columns in the counts matrix, typically the number of observed counts per sample.
k: The rank of the Poisson NMF or the number of topics.
s: A vector of length n giving the "size factor" estimates; these estimates should be equal, or close to, the total counts in each row of the counts matrix.
numiter: The number of loadings and/or factor updates performed.
loglik: The Poisson NMF log-likelihood.
loglik.multinom: The multinomial topic model log-likelihood.
dev: The Poisson NMF deviance.
res: The maximum residual of the Karush-Kuhn-Tucker (KKT) first-order optimality conditions. This can be used to assess convergence of the updates to a (local) solution.
mixprops: Matrix giving a high-level summary of the mixture proportions, in which rows correspond to topics, and columns are ranges of mixture proportionss.
topic.reps: A matrix in which the ith row gives the mixture proportions for the sample "most representative" of topic i; by "most representative", we mean the row (or sample) with the highest proportion of counts drawn from the topic i.