Changes in version 1.0.0 (2025-11-21)                  

Major update

  - PERMANOVA refactor: The implementation used by prep_data() now uses
    a distance-based MANOVA (dbManova...()) pipeline using
    Gower-centered distance matrices. This replaces the previous SS
    calculations based on Huygens' theorem
      - The new approach is numerically more stable, especially with
        non-Euclidean distances (e.g., Bray–Curtis), where negative
        eigenvalues can invalidate naive SS partitions.
      - Sums of squares are obtained from projection matrices in the
        PCoA space (SSCP traces), with correct handling of negative
        eigenvalues (sign weighting), avoiding drift due to
        rounding/centering errors.
      - Results are therefore more robust and interpretable across a
        wider range of ecological distances.

Notes on compatibility

The user API of prep_data() remains the same. Small numerical
differences may occur due to more correct SS partitioning.

                 Changes in version 0.13.0 (2025-08-23)                 

This is a maintenance and feature-update release, preparing the package
for resubmission to CRAN.

NEW FEATURES

  - Updated the function sim_cbo() by changing the structure to an
    empirical optimization based on precision and then cost. The
    function evaluates for the sampling designs that best approximate to
    $(1-\alpha)$ and then finds the cost for each one of the selected
    sampling efforts. Lastly, the function marks the best options. The
    function is documented and includes examples.
  - Added the function underwood_cbo() to keep the possibility of using
    Underwood's optimization that was used in previous versions of the
    package. The function is documented and includes examples.

BUG FIXES

  - Resolved NOTEs related to "no visible binding for global variable"
    and "no visible global function definition". This was addressed by
    explicitly importing functions with @importFrom and declaring global
    variables with utils::globalVariables(), making the package more
    robust and compliant with CRAN policies.
  - Corrected potential miscalculations in internal simulation functions
    by ensuring all variable scopes are handled correctly.

IMPROVEMENTS

  - The main vignette has been updated to include a detailed example and
    workflow for the new permanova_twoway() function.
  - Internal code has been refactored to reduce dependencies and improve
    clarity, for instance, by favoring base R functions where
    appropriate.

                 Changes in version 0.12.0 (2024-08-21)                 

  - 'ecocbo' now can work with either single-factor and nested-symmetric
    experiments.

  - A new function, prep_data(), is added. This allows the user to
    select which model to use and then prepares the data for using with
    the rest of the functions.

  - plot_power() was updated. Readability of the power curve is improved
    by differentiating the optimal, or user selected, experimental
    design.

                       Changes in version 0.11.0                        

                 Changes in version 0.10.2 (2023-08-18)                 

This is a resubmission, in this version we have:

  - corrected the presentation for function names in the description
    texts by removing single quotes and adding () after each name.

  - added ISBN to the reference for Underwood (1997) in the DESCRIPTION,
    as well as in the manual files and vignette.

  - changed \dontrun{} to \donttest{} for examples involving
    'sim_beta()' as it takes more than 5 seconds to run them as they
    are. Changing the parameter values to make the function run faster
    would not be instructional to the final user, as it would not
    demonstrate the function's actual runtime and functionality. The
    example code will not be run automatically, but it can still be run
    manually if desired.

                       Changes in version 0.10.1                        

  - We added '\donttest{}' to the 'sim_beta()' example in the
    DESCRIPTION file because it takes more than 5 seconds to run.
    Changing the parameter values to make the function run faster would
    not be instructional to the final user, as it would not demonstrate
    the function's actual runtime and functionality. The example code
    will not be run automatically, but it can still be run manually if
    desired.

                        Changes in version 0.9.1                        

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