Changes in version 2.0.1.9000                      

                 Changes in version 2.0.1 (2026-03-08)                  

  - Fixed numerical stability in ppg_esim tests where eigen() could
    return complex values on certain hardware architectures (e.g., macOS
    ARM64).
  - Added automated documentation spell-checking via the spelling
    package in a dedicated GitHub Actions workflow.
  - Expanded the GitHub Actions CI matrix to 17 runners, mirroring all
    official CRAN check flavors.

                 Changes in version 2.0.0 (2026-02-28)                  

Breaking changes

  - All primary user-facing functions have transitioned from
    dot-notation to snake_case notation for better clarity and to avoid
    S3 method dispatch conflicts.
      - phen.varcov() is now phen_varcov()
      - gen.varcov() is now gen_varcov()
      - weight.mat() is now weight_mat()
      - gen.advance() is now gen_advance()
  - The base combinatorial selection functions have been replaced with
    the dedicated lpsi() function (Linear Phenotypic Selection Index).

Major changes

  - The variance-covariance calculation engine, system solvers, and
    combinatorial index builders are now fully powered by Rcpp and
    RcppEigen.
  - Index evaluations for large trait configurations are now
    exponentially faster due to primitive matrix operations.

New features

  - Introduced new suites for evaluating linear genomic selection
    indices (lgsi()) and combining genomic/phenotypic data (ppg_lgsi()).
  - Introduced new modules for tracking continuous index performance
    across multiple stages of breeding trials (mlpsi(), mlgsi()).
  - Implemented specialized solvers to maximize genetic advance while
    specifically constraining genetic gain for restrictive/undesired
    traits to zero (rlpsi(), dg_lpsi()).
  - Adopted Eigen-decomposition based selection methods for phenotypic
    (esim()) and genomic (gesim()) evaluations.
  - Added specialized infrastructure for evaluating linear marker
    selection indices (lmsi()).
  - Added new comprehensive toolsets to actively simulate and visualize
    multi-cycle genetic advance over time under varied selection index
    intensities and environmental variances
    (simulate_selection_cycles()).
  - Created a complete pkgdown official website.
  - Added a complete vignette suite detailing the mathematical
    foundation and code usage for all new marker, genomic, multi-stage,
    constrained, and stochastic features.

                 Changes in version 1.2.1 (2026-01-26)                  

  - improve the overall performance of the package

                 Changes in version 1.2.0 (2023-09-19)                  

  - add new function meanPerformance() for calculating the mean
    performance in randomized block design data
  - remove some typos from the documentation

                 Changes in version 1.1.4 (2022-06-13)                  

  - removed two functions sel.index() and sel.score.rank() for easy
    implementation of comb.indices()

                 Changes in version 1.1.3 (2021-09-25)                  

  - Added a new function gen.advance() for genetic advance calculation

                 Changes in version 1.1.2 (2021-07-26)                  

  - rank column added in rcomb.indices() for ranking of index

                 Changes in version 1.1.1 (2021-06-03)                  

  - Added a new function rcomb.indices() for calculated possible
    selection indices excluding single character or particular index
    from all possible selection indices

                 Changes in version 1.1.0 (2021-04-23)                  

  - Bug fixes in phen.varcov()
  - Added a new function comb.indices() for calculating possible
    selection indices with the group/pairs of traits/characters
  - removed function: rank.index()

                 Changes in version 1.0.0 (2021-04-06)                  

  - Added a NEWS.md file to track changes to the package.