Changes in version 0.1.0 (2026-08-08)                  

Initial CRAN submission.

Features

  - tokenize_posts() — tokenizes a text column into a long-format token
    table, with optional English stopword removal.
  - cooccurrence_counts() — computes 2x2 contingency table counts (a, b,
    c, d) for every pair of words meeting a minimum document-frequency
    threshold.
  - word_orientation() — computes the phi coefficient for each word
    pair, tests it for significance (chi-squared or Fisher's exact test
    for sparse cells), applies a Benjamini-Hochberg correction across
    all tested pairs by default, and classifies each pair as
    "attraction", "repulsion", or "neutral".
  - plot_orientation_network() — plots the strongest attraction and
    repulsion word pairs as a network graph, using base igraph plotting.
  - analyze_word_orientation() — a single-call pipeline wrapping the
    four functions above, from raw text to a classified table and
    network plot.
  - example_social_posts() — a small built-in toy dataset for examples
    and tests.