toLDAvisJson: Wrapper to create Json mapping for LDAvis. This can be useful...

In bstewart/stm: Estimation of the Structural Topic Model

Wrapper to create Json mapping for LDAvis. This can be useful in indirect render e.g. Shiny Dashboards

Description

Tool for exploring topic/word distributions using LDAvis topic browser.

Usage

toLDAvisJson(
  mod,
  docs,
  R = 30,
  plot.opts = list(xlab = "PC1", ylab = "PC2"),
  lambda.step = 0.1,
  reorder.topics = TRUE
)

Arguments

mod	STM object. Output from stm function.
docs	Documents object passed to stm in this package's standard format (see the documentation in stm.
R	Passed to createJSON "integer, the number of terms to display in the barcharts of the interactive viz. Default is 30. Recommended to be roughly between 10 and 50."
plot.opts	Passed to createJSON "a named list used to customize various plot elements. By default, the x and y axes are labeled 'PC1' and 'PC2' (principal components 1 and 2), since jsPCA is the default scaling method. "
lambda.step	Passed to createJSON "a value between 0 and 1. Determines the interstep distance in the grid of lambda values over which to iterate when computing relevance. Default is 0.01. Recommended to be between 0.01 and 0.1."
reorder.topics	Passed to createJSON "Should LDAvis re-order the K topics in order of decreasing proportion? Default is True"

Details

Tool for exploring topic/word distributions using LDAvis topic browser. Development build of LDAvis available at https://github.com/cpsievert/LDAvis or download from CRAN. Note: LDAvis may renumber the topics.

References

Carson Sievert and Kenny Shirley. LDAvis: Interactive Visualization of Topic Models. R package version 0.3.1. https://github.com/cpsievert/LDAvis

Examples

mod <- stm(poliblog5k.docs, poliblog5k.voc, K=25,
           prevalence=~rating, data=poliblog5k.meta,
           max.em.its=2, init.type="Spectral") 
#please don't run a model with 2 iterations
#this is done here to make it run quickly.
toLDAvisJson(mod=mod, docs=poliblog5k.docs)

bstewart/stm documentation built on Jan. 3, 2024, 6:58 p.m.