create_topic_model_report: Create Standard Results Report for a Structural Topic Model
In dtburk/texanaaid: An Aide for Text Analysis

View source: R/create_topic_model_report.R

Create standard results report for a structural topic model, including FREX and PROB words, topic prevalence plot, topic time trend plots, and most representative documents.

1
2
3

create_topic_model_report(model, report_dir, cov_dataset, raw_texts,
  diff_cov = NULL, diff_cov.value1 = NULL, diff_cov.value2 = NULL,
  representative_docs = TRUE)

`model`	An object of class STM.
`report_dir`	Directory in which to save the report.
`cov_dataset`	A data frame containing any document-level covariates. Must contain a variable named "year".
`raw_texts`	A character vector containing the texts analyzed in the topic model in a human-readable format, to output most representative documents for each topic.
`diff_cov`	The name of a (factor or character) variable in `cov_dataset` with which to produce a difference plot. If NULL, no plot will be produced.
`diff_cov.value1`	The first level of `diff_cov` to be used to create a difference plot.
`diff_cov.value2`	The second level of `diff_cov` to be used to create a difference plot.
`representative_docs`	If `TRUE` (the default), save full texts of most representative documents. If `FALSE`, save only a list of titles of the most representative documents.