                 Changes in version 2.2.4 (2026-09-11)                  

  - released: 2026-09-11
  - merged fixes from upstream arules related to computation of interest
    measures (#55)
  - fixed computation of support with non-idempotent t-norms (#56)

                 Changes in version 2.2.3 (2026-08-20)                  

  - released: 2026-08-19
  - added permute()
  - improved filtering by axioms specified in excluded argument of
    dig*()
  - improved the core search algorithm by better handling of condition
    prefixes and caching of conjunction supports
  - added "search_stats" attribute to each result of dig*() functions
  - improved documentation, added vignettes

                 Changes in version 2.2.2 (2026-07-22)                  

  - released: 2026-07-22
  - optimized dig() on sparse crisp data (by implementing sparse bit
    chain)
  - added clustering characteristics to explore() for association rules
  - fixed handling of axioms in dig() and dig_*() function (arg
    excluded)
  - fixed handling of NA values in partition()

                 Changes in version 2.2.1 (2026-06-10)                  

  - released: 2026-06-10
  - fixed failing unit tests because of upstream fixes in R's
    wilcox.test()
  - enhanced placement of nodes in geom_diamond() to reduce edge
    crossing
  - added .subsets argument to partition() that enables partitioning by
    subsets of factor levels

                 Changes in version 2.2.0 (2026-03-11)                  

  - released: 2026-03-11
  - fixed bugs in explore()
  - added special value highlighting in explore()'s table of rules
  - added contingency table view to explore.associations()
  - added explore() methods for baseline contrasts, complement
    contrasts, paired baseline contrasts, and correlations
  - added dig_ancestors()
  - added plot_contingency()
  - enhanced info and error messages
  - added arguments for line type and colour to geom_diamond()
  - enhanced dig_*_contrasts() - statistical test errors are now stored
    in the "comment" column in the returned data frame

                 Changes in version 2.1.2 (2026-02-05)                  

  - released: 2026-02-04
  - added is_logicalish() to check if a vector is logical or can be
    coerced to logical
  - fixed issue with required version of R (>= 4.4.0)
  - fixed critical bug in explore()

                 Changes in version 2.1.1 (2026-02-03)                  

  - released: 2026-02-03
  - performance improvements
  - updated XSIMD library to 14.0.0
  - moved Shiny-related packages (shiny, shinyjs, shinyWidgets, DT,
    htmltools, htmlwidgets, jsonlite) from Imports to Suggests
  - removed deprecated measures argument from dig_associations() and
    dig_tautologies()
  - deprecated argument contingency_table in dig_associations() and
    dig_tautologies()
  - removed dependency on BH and RcppThread

                 Changes in version 2.1.0 (2025-11-05)                  

  - released: 2025-11-05
  - added cluster_associations() to cluster association rules
  - added add_interest() for associations flavour of nugget to compute
    additional interest measures (GUHA and arules measures)
  - deprecated measures argument in dig_associations() and
    dig_tautologies()
  - parse_condition() rewritten in C++ for better performance
  - enhanced UI layout of explore() for associations
  - added "Cluster" tab to explore() for associations
  - added selection of columns to show in explore() for associations
  - fixed rchk protection stack imbalance in CombinatorialProgress
    constructor

                 Changes in version 2.0.2 (2025-10-30)                  

  - released: 2025-10-31
  - attempt to fix rchk protection stack imbalance in
    CombinatorialProgress constructor
  - created vignette "Data Preparation"
  - updated main vignette

                 Changes in version 2.0.1 (2025-10-13)                  

  - released: 2025-10-13
  - fixed problem with C++20 and testthat by downgrading system
    requirements to C++17
  - added association_matrix()

                 Changes in version 2.0.0 (2025-10-13)                  

  - released: 2025-10-13
  - completely rewritten the core algorithm in dig()
  - all dig_*() functions now return nugget S3 objects
  - added is_almost_constant(), remove_almost_constant(),
    parse_condition(), fire(), is_condition(), remove_ill_conditions(),
    shorten_condition()
  - added dig_tautologies()
  - added geom_diamond()
  - added .span and .inc arguments to partition()
  - added various styles (quantile, k-means, hclust, bclust, ...) of
    crisp partitioning to partition()
  - added explore() function for interactive exploration of patterns
  - added exclude argument to dig() and other dig_*() functions
  - added support for the disjoint parameter to var_grid() and
    dig_grid()
  - added progress bar to the dig() function
  - added "dummy" method to partition()

                 Changes in version 1.4.0 (2025-01-08)                  

  - released: 2025-01-08
  - added var_names(), dig_baseline_contrasts(),
    dig_complement_contrasts()
  - dig_contrasts() renamed to dig_paired_baseline_contrasts()
  - dig_implications() renamed to dig_associations()
  - dichotomize() is deprecated (use partition() instead)
  - added max_support argument to dig()
  - added max_results argument to dig()
  - optimized performance of dig()
  - added min_conditional_focus_support argument to dig()
  - fixed handling of NULL returned by a callback function in dig_grid()
  - argument d of the callback function in dig_grid() renamed to pd
  - added handling of nd argument of the callback function in dig_grid()
  - added max_p_value argument to dig_paired_baseline_contrasts()
  - improved error messages
  - added nuggets vignette
  - started using lifecycle and pkgdown

                 Changes in version 1.3.0 (2024-11-13)                  

  - released: 2024-11-13
  - added is_degree(), dig_contrasts(), partition()
  - implemented fuzzy variant of dig_grid()
  - fixed crash when mixing logical (crisp) and numeric (fuzzy) inputs
    to dig()

                 Changes in version 1.2.0 (2024-10-11)                  

  - released: 2024-10-11
  - added var_grid(), dig_grid()
  - added the measures argument to dig_implications()
  - fixed contingency table arguments computation (pp, pn, np, nn) -
    previously, they were computed as relative frequencies, now they are
    computed as counts
  - fixed new-delete-type-mismatch ASAN error caused by wrong
    implementation of AlignedAllocator
  - fixed memory leaks

                 Changes in version 1.1.0 (2024-10-08)                  

  - released: 2024-10-08
  - added .other argument to dichotomize()
  - fixed handling of xvars, yvars tidy-selectors in dig_correlations())
  - added filtering of foci by their support
  - added handling of callback function arguments related to contingency
    tables (pp, pn, np, nn arguments)

                 Changes in version 1.0.2 (2024-01-09)                  

  - released: 2024-01-09
  - fixed handling of arguments min_coverage and min_support in
    dig_implications()
  - attempt to fix LTO error related to run_testthat_tests() - fixed by
    using an RC version of Rcpp (1.0.11.6)

                 Changes in version 1.0.1 (2023-11-29)                  

  - released: 2023-11-29
  - attempt to fix tests failing on R-devel

                 Changes in version 1.0.0 (2023-11-28)                  

  - released: 2023-11-28
  - first version of the package
  - implemented: dichotomize(), dig(), dig_implications(),
    dig_correlations(), which_antichain(), format_condition(),
    is_subset()
