Skip to contents

Changelog

Source: NEWS.md

taxdiv 1.0.0

New Features

  • batch_analysis() gains an indices parameter for selective index computation. Three groups available: "classical" (Shannon, Simpson), "clarke_warwick" (Delta, Delta*, AvTD, VarTD), and "ozkan_pto" (8 pTO indices). Default computes all. Unambiguous abbreviations accepted (e.g., "clas", "clark", "oz").

  • Westhoff-Maarel scale: recommended for compatibility with Ozkan

    1. and the original Excel macro, but no longer enforced as a hard requirement. The 1-9 limit in the original macro was a practical constraint of VBA (the macro froze on larger values), not a mathematical requirement of the algorithm. In R, ozkan_pto() and related functions accept any positive numeric abundance values.

  • n_iter minimum lowered from 101 to 1: ozkan_pto_resample(), ozkan_pto_sensitivity(), and ozkan_pto_full() now accept any positive integer for n_iter. The old >= 101 floor was an inherited Excel-macro convention with no mathematical basis. With n_iter = 1, only the deterministic first iteration runs, so the result equals ozkan_pto() (Run 1). Note: because the procedure returns the maximum across iterations, the result is non-decreasing in n_iter — keep n_iter fixed when comparing sites and use a high value (e.g. 500-1000) for final estimates.

  • Natural sort for site names: sites are now ordered as Site1, Site2, …, Site9, Site10 instead of the previous alphabetical order (Site1, Site10, Site11, Site2, …). Works with any naming pattern (A1-A10, Plot1-Plot20, etc.).

Shiny Dashboard (taxdiv Explorer)

  • Index group selection: three individual checkboxes with descriptions replace the previous “compute everything” approach. Each checkbox shows what the index group measures and any requirements.
  • Conditional settings panel: n_iter and seed fields only appear when Ozkan pTO is selected, reducing UI clutter.
  • Adaptive bar chart and summary boxes: when pTO is not selected, the chart and summary boxes automatically switch to Shannon or Delta instead of showing nothing.
  • Natural sort in Site Filter: sidebar site checkboxes now follow natural sort order matching the results table.
  • Template file rebuilt: 4-sheet Excel template (ENTER_DATA, EXAMPLE_text, EXAMPLE_numeric, INSTRUCTIONS) with Westhoff-Maarel abundances, styled headers (Calibri 12 bold / 11 normal), and English instructions with em dash formatting.
  • Download buttons fixed: the Excel/CSV download buttons on the Results card now sit in a card_footer() so they remain visible regardless of how tall the results table grows (previously the table’s fill behavior could push the buttons out of view).
  • n_iter default raised to 500 and its minimum lowered to 1, matching the relaxed function-level limit.

Documentation & Internalization

  • Translated all Turkish comments in test-ozkan_pto_resample.R to English.
  • Roxygen documentation for batch_analysis() updated with indices parameter details, Westhoff-Maarel requirements, and correct abbreviation examples.
  • All vignette examples updated to Westhoff-Maarel scale abundances (giris_rehberi.Rmd, introduction.Rmd, ozkan-pto.Rmd, visualization.Rmd, workflow.Rmd).
  • All Roxygen @examples updated to Westhoff-Maarel scale across 7 R source files.
  • inst/extdata/mediterranean_forest.csv converted from raw counts to Westhoff-Maarel scale.
  • Spelling wordlist updated (+11 terms: clarke, warwick, RStudio, etc.).

Bug Fixes

  • Fixed pmatch() ambiguity: documented that "cl" and "cla" are ambiguous between “classical” and “clarke_warwick”; minimum unambiguous prefixes are "clas" and "clark".

Testing

  • 668 assertions across 12 test files (up from 610).
  • New tests: index selection (9 tests), Westhoff-Maarel validation (4 tests), natural sort (3 tests).
  • All test data updated to Westhoff-Maarel scale across 5 test files (test-batch_analysis.R, test-s3_methods.R, test-parallel.R, test-ozkan_pto.R, test-compare_indices.R).
  • R CMD check –as-cran: 0 errors, 0 warnings, 0 notes.

taxdiv 0.1.0

CRAN release: 2026-04-01

Initial release of taxdiv — taxonomic diversity indices using Deng entropy.

Core Functions

  • Deng entropy at any taxonomic level (deng_entropy_level())
  • Ozkan (2018) pTO with slicing procedure (ozkan_pto(), pto_components())
    • 8 indices: uTO, TO, uTO+, TO+, uTO_max, TO_max, uTO+_max, TO+_max
    • max_level parameter for controlling taxonomic depth (NULL, “auto”, integer)
    • Presence-based entropy: equal weight (1/S) at each slice
  • Ozkan & Mert (2022) reinforced estimators:
  • Stochastic resampling — Run 2 (ozkan_pto_resample())
    • Random species inclusion/exclusion (50% probability per species)
    • Configurable iteration count and seed for reproducibility
  • Sensitivity analysis — Run 3 (ozkan_pto_sensitivity())
    • Species-specific inclusion probabilities derived from Run 2
    • Overall maximum across Run 1 + 2 + 3
  • Jackknife leave-one-out analysis (ozkan_pto_jackknife())
  • Full pipeline in one call (ozkan_pto_full())
    • Combines Run 1 + Run 2 + Run 3 + jackknife
  • Classical indices: Shannon H’ (shannon()) and Gini-Simpson (simpson())
    • Bias correction methods: Miller-Madow, Grassberger, Chao-Shen
  • Clarke & Warwick taxonomic distinctness:
    • Taxonomic diversity Delta (delta())
    • Taxonomic distinctness Delta* (delta_star())
    • Average taxonomic distinctness AvTD/Delta+ (avtd())
    • Variation in taxonomic distinctness VarTD/Lambda+ (vartd())
  • Multi-community comparison (compare_indices()) — 14 indices side by side
  • Multi-site batch analysis (batch_analysis()) — automatic column detection
  • Simulation-based significance testing (simulate_td())
    • Random subsampling from species pool for AvTD/VarTD confidence funnels
  • Taxonomic rarefaction with bootstrap CI (rarefaction_taxonomic())
    • 8 index choices, configurable sample sizes and iterations
  • Taxonomic distance matrix (tax_distance_matrix())
  • Taxonomy tree builder (build_tax_tree())
  • Parallel computing support for simulate_td(), rarefaction_taxonomic(), and batch_analysis() via parallel and n_cores parameters

S3 Class System

  • 6 S3 classes: compare_indices, batch_analysis, ozkan_pto, ozkan_pto_resample, ozkan_pto_sensitivity, rarefaction_taxonomic, td_simulation
  • 13 S3 methods: print(), summary(), and plot() for all main outputs

Visualization

Example Datasets

  • anatolian_trees — Anatolian tree species with full taxonomic hierarchy
  • gazi_comm — Community abundance data from Gazi University campus
  • gazi_gytk — Taxonomic classification for Gazi campus species

Documentation

  • English vignette with 7 plots and worked examples
  • Turkish (Turkce) vignette with 8 plots, formulas, and full walkthrough
  • pkgdown website: https://mgorgoz.github.io/taxonomic-diversity-r/
  • Pseudocode for pTO Run 1/2/3 (inst/pseudocode/taxonomic_diversity_pseudocode.md)
  • Pseudocode for Clarke & Warwick (inst/pseudocode/taxonomic_distance_pseudocode.md)

Testing

  • 610 assertions across 12 test files
  • Integration tests covering full Run 1 -> 2 -> 3 pipeline
  • Reproducibility tests with seed control
  • Mathematical relationship checks (ordering constraints, boundary cases)
  • R CMD check: 0 errors, 0 warnings, 0 notes

Validation

  • Verified against Ozkan (2018) hypothetical examples
  • Verified against Ozkan & Mert (2022) reinforced estimator methodology
  • Cross-validated with Kursad Ozkan’s Excel macro (TD_OMD.xlsm)
  • Documented Excel macro bug: Application.Calculate missing inside tekerur2() and tekerur3() loops