Coherence and exclusivity of each topic of one model — plot_topic

The per-topic version of plot_k_frontier(), as stm::topicQuality(): each topic is a point (coherence x exclusivity), sized by prevalence and labelled by its number. Topics at the bottom left are poor on both criteria: read them first, they are often junk, mixed or background topics.

Usage

plot_topic_diagnostics(
  x,
  corpus = NULL,
  K = NULL,
  run = 1L,
  coherence = "coherence_umass",
  exclusivity = "exclusivity",
  base_size = 16,
  ...
)

Arguments

x: A table from topic_diagnostics(), or the topics table of a topic_model_evaluation (then give corpus, K and run).
corpus, K, run: Model to draw, when x holds several.
coherence, exclusivity: Column names of the axes.
base_size: Numeric scalar, base font size of the theme.
...: Not used: an argument renamed in the development version of mintR gives an error naming its replacement.

Value

A ggplot.