Prepare Text Data and Analyze Topic Models — sniff_groups_stm

Processes text data for structural topic modeling and performs topic number selection analysis, returning both the processed data and diagnostic plots.

Usage

sniff_groups_stm_prepare(
  groups,
  group_to_stm = "g01",
  search_topics = c(5:40, 45, 50, 55, 60),
  seed = 1234,
  cores = 1
)

Arguments

groups: A list containing network data with a 'network' component
group_to_stm: Character string specifying which research group to process (default: 'g01')
search_topics: Numeric vector of topic numbers to evaluate (default: c(5:40, 45, 50, 55, 60))
seed: Random seed for reproducibility (default: 1234)
cores: Number of CPU cores to use (default: 1)

Value

A list containing:

result: The searchK results object
plots: A list containing two ggplot objects (p1: metrics by K, p2: exclusivity vs coherence)
df_prep: Output from stm::textProcessor
df_doc: Output from stm::prepDocuments
df: Original filtered data

Examples

if (FALSE) { # \dontrun{
output <- sniff_groups_stm_prepare(network_data)
output$plots$p1 # View first plot
output$result # Access search results
} # }