Skip to contents

Calculate Log Odds Ratio Between Categories

Source: R/lexical_analysis.R
calculate_log_odds_ratio.Rd

Computes log odds ratio to compare word frequencies between categories. Identifies words that are distinctively used in one category vs another. Uses Laplace smoothing to handle zero counts.

Usage

calculate_log_odds_ratio(
  dfm_object,
  group_var,
  comparison_mode = c("binary", "one_vs_rest", "pairwise"),
  reference_level = NULL,
  top_n = 10,
  min_count = 5
)

Arguments

dfm_object

A quanteda dfm object

group_var

Character, name of the grouping variable in docvars

comparison_mode

Character, one of "binary", "one_vs_rest", or "pairwise"

  • binary: Compare two categories directly

  • one_vs_rest: Compare each category against all others combined

  • pairwise: Compare all pairs of categories

reference_level

Character, reference category for binary comparison (default: first level)

top_n

Number of top terms per comparison (default: 10)

min_count

Minimum word count to include (default: 5)

Value

Data frame with columns:

  • term: The word/feature

  • category1: First category in comparison

  • category2: Second category in comparison

  • count1: Count in category 1

  • count2: Count in category 2

  • odds1: Odds in category 1

  • odds2: Odds in category 2

  • odds_ratio: Ratio of odds

  • log_odds_ratio: Log of odds ratio (positive = more in compared category)

See also

Other lexical: calculate_dispersion_metrics(), calculate_lexical_dispersion(), calculate_text_readability(), clear_lexdiv_cache(), detect_multi_words(), extract_keywords_keyness(), extract_keywords_tfidf(), extract_morphology(), extract_named_entities(), extract_noun_chunks(), extract_pos_tags(), extract_subjects_objects(), find_similar_words(), get_sentences(), get_spacy_embeddings(), get_spacy_model_info(), get_word_similarity(), init_spacy_nlp(), lexical_analysis, lexical_diversity_analysis(), lexical_frequency_analysis(), parse_morphology_string(), plot_keyness_keywords(), plot_keyword_comparison(), plot_lexical_diversity_distribution(), plot_morphology_feature(), plot_readability_by_group(), plot_readability_distribution(), plot_tfidf_keywords(), plot_top_readability_documents(), render_displacy_dep(), render_displacy_ent(), spacy_extract_entities(), spacy_has_vectors(), spacy_initialized(), spacy_parse_full(), summarize_morphology()

Examples

if (FALSE) { # \dontrun{
library(quanteda)
corp <- corpus(c("The cat runs fast", "Dogs are loyal pets",
                 "Cats sleep all day", "My dog loves walks"),
               docvars = data.frame(animal = c("cat", "dog", "cat", "dog")))
dfm <- tokens(corp) %>% dfm()
log_odds <- calculate_log_odds_ratio(dfm, "animal")
} # }