                 Changes in version 1.1.0 (2026-10-05)                  

First version submitted to CRAN.

Breaking change

  - generate_and_apply() no longer writes the result into the global
    environment; it returns the data frame. Assign it: df2 <-
    generate_and_apply(df, lines). The new_name argument is kept but
    ignored.

Bug fixes

  - Calcule_Comparative() labelled the groups in order of appearance
    while t.test() orders them alphabetically, so a table could show one
    group's mean with the other group's label and standard deviation.
    Both now follow levels(factor(group)). Cohen's d is computed
    directly (pooled SD); the robust option still uses
    WRS2::akp.effect().
  - calcula_omega_all() failed and Fiabilidad() returned a list with
    recent semTools versions, where compRelSEM() returns a list; both
    return numbers again.
  - u_mann_whitney_superioridad() labelled large effects in favour of
    the second group (PS well below .50) as "No efecto"; the size now
    uses max(PS, 1 - PS). Cases with missing values are excluded before
    counting n.
  - calcular_correlaciones() assigned significance marks by comparing p
    values as text; it now uses the numeric p values.
  - normalize_carreras(remove_unmatched = TRUE) failed because it used
    the output before creating it.
  - Decimal commas ("3,5") were read as thousands separators (35) in
    clean_edad(), convert_age_to_years() and
    convert_age_to_years_months(); "un año" is now recognized as 1 year.
  - epsilon_cuadrado_kruskal() counts only the cases used by the test.

CRAN compliance

  - No function installs or attaches packages at run time; dependencies
    are declared in DESCRIPTION and imported in the NAMESPACE.
  - Documentation generated with roxygen2; every function has a runnable
    example and the two datasets are documented.
  - Progress is reported with message() instead of print().
  - Non-ASCII characters in the code are escaped.
