Skip to contents

Calculate word frequencies from text data

Source: R/helpers.R
word_freqs.Rd

Analyzes a collection of texts (such as user profile descriptions) to identify the most frequently used words or word pairs (bigrams). Useful for understanding the common themes and topics within network communities.

Usage

word_freqs(
  texts,
  bigrams = FALSE,
  top = 30,
  remove_stopwords = TRUE,
  language = "en"
)

Arguments

texts

Character vector. Texts to analyze (e.g., user profile descriptions)

bigrams

Logical. If TRUE, count word pairs instead of individual words (default FALSE)

top

Integer. Number of top frequent words/bigrams to return (default 30)

remove_stopwords

Logical. Remove common stopwords like "the", "and" (default TRUE)

language

Character. Language for stopword removal (default "en" for English)

Value

A data frame with word frequency statistics:

feature

Character. The word or bigram

frequency

Integer. Number of occurrences

rank

Integer. Rank by frequency (1 = most frequent)

docfreq

Integer. Number of documents containing the feature

group

Character. Always "all" for this function

See also

com_labels

Examples

# Sample user descriptions
descriptions <- c(
  "Data scientist working on machine learning projects",
  "Climate researcher studying global warming effects",
  "Marine biologist researching ocean ecosystems",
  "Environmental data scientist analyzing climate patterns"
)

# Get top words
word_freq <- word_freqs(descriptions, top = 10)
head(word_freq)
#>     feature frequency rank docfreq group
#> 1      data         2    1       2   all
#> 2 scientist         2    1       2   all
#> 3   climate         2    1       2   all
#> 4   working         1    4       1   all
#> 5   machine         1    4       1   all
#> 6  learning         1    4       1   all

# Get top bigrams
bigram_freq <- word_freqs(descriptions, bigrams = TRUE, top = 5)
head(bigram_freq)
#>             feature frequency rank docfreq group
#> 1    data scientist         2    1       2   all
#> 2 scientist working         1    2       1   all
#> 3   working machine         1    2       1   all
#> 4  machine learning         1    2       1   all
#> 5 learning projects         1    2       1   all

if (FALSE) { # \dontrun{
# With real profile data from network analysis
auth <- get_token("your.handle.bsky.social", "your-app-password")
profiles <- get_profiles(c("user1.bsky.social", "user2.bsky.social"), auth$accessJwt)
word_analysis <- word_freqs(profiles$description, top = 50)
} # }