Changelog

Source: NEWS.md

RRatepol 1.3.0

Complete unit test suite

• tests/testthat/ created from scratch (~3 000 lines across 19 test files, one per function)
• shared helper-fixtures.R added for reusable test data
• every function now has tests for valid input, invalid input with informative errors, and edge cases

CI and test coverage

• new test-coverage.yaml workflow uploads coverage results to Codecov on every push
• R-CMD-check.yaml updated to current r-lib/actions conventions
• testthat (≥ 3.0.0) added to Suggests

Removed runtime dependencies

• RUtilpol (GitHub-only) and usethis removed from Imports
• Remotes: field removed; package is now installable from CRAN without any GitHub dependency

New internal utilities

• R/util_internal.R: drop-in replacements for all RUtilpol helpers used in this package
• R/util_search_parameter.R: Grimm-smoothing window-growth logic extracted into util_search_parameter() (#96)

New silent argument

• silent = FALSE added to estimate_roc() and all functions that produce console output; when TRUE, suppresses all messages and warnings without affecting verbose (#109)

Bug fixes

• extract_data(): fix NULL column names in age-uncertainty matrix (#48); fix inverted sort condition that left unsorted age data unsorted (#49); fix NA-detection condition that silently passed NA age rows downstream (#50)
• reduce_data(): add two-way sample alignment so community, age, and age-uncertainty stay in sync when NA age rows are dropped (#52)
• prepare_data(): remove orphaned res object that caused an error inside purrr::map() (#55)
• estimate_dissimilarity_coefficient(): fix copy-paste error where "chisq" called vegan::vegdist(method = "chord") (#56)
• subset_samples(): drop empty-bin rows instead of leaving NA-filled template rows in the result (#61)
• detect_peak_points(): adapt to renamed gratia::derivatives() columns (.lower_ci replaces lower) and switch from deprecated newdata to data argument (#69)
• check_data(): exclude NAs from summary statistics and report NA counts per column (#43, #47)
• smooth_community_data(): add assertions for logical parameters and smooth_n_points bounds; raise informative errors for shep with ≤ 2 points (#51, #83)
• plot_roc(): fix wrong object name (sequence_01 → sequence_02) in function example (#106)

Input validation — new assertions

• assertions added across all exported functions to catch invalid, empty, or mismatched inputs before they cause cryptic downstream errors
• extract_data(): additional pre- and post-processing checks (#81)
• reduce_data(): NULL/empty input guards; intersect()-based sample alignment fixes all-zero row handling (#82, #37)
• prepare_data(): NULL/empty guards; bin_size assertion skipped when working_units = "levels" (#84, #79)
• make_bins(): working_units must be a single value (#85)
• run_iteration(): logical-parameter assertions; time_standardisation must not be 0 or NA; standardisation-failure error is now unconditional (#86, #62)
• subset_samples(): bin_selection must be "first", "random", or NULL (#87)
• reduce_data_simple(): empty-input guard; logical-parameter assertions (#88)
• subset_community(): empty-input guard (#89)
• standardise_community_data(): guard against n_individuals = 0; message when minimum count is used as default (#90)
• transform_into_proportions(): empty-input guard; single-value sel_method check (#91)
• make_trend(): single-value sel_method check; return value coerced to numeric vector, fixing "non_linear" returning an array (#92, #78)
• detect_sni(): NA checks on CharData columns 2 and 3 (#93)
• plot_roc(): peaks must be a single non-NA value; warning when ROC contains NAs (#94)
• estimate_roc(): single-value and logical checks for standardise, tranform_to_proportions, verbose, interest_threshold, and use_parallel (#95)

Documentation and site

• Roxygen2 updated from 7.2.3 → 7.3.3; all man/*.Rd files regenerated
• pkgdown site rebuilt; reference index and article pages updated
• quarto added to Suggests to support vignette rendering

New contributor

• Friederike Wolke (@FriedaRosa) added as contributor (ctb) for authoring the test suite and filing the issues that drove this release (#60)

RRatepol 1.2.3

• fix an issue with subsetting the uncertainty matrix to correctly align with the rest of the data (thanks to Giacomo Galli for reporting the bug)
• replace all not a basic ASCII character in documentation

RRatepol 1.2.2

• add more information about additional resources and workshops
• use {neotoma2} from CRAN
• fix an issue with a case, where the age difference between samples is zero. The zeros are replaced by arbitrary 0.1.

RRatepol 1.2.1

Uncertainty sampling

• fix an issue with an error message for a large number of iterations in sampling age-sequence

RRatepol 1.2.0

Name uniformity

• In order to make all (exported) functions easier to use (use verbs, snake_style,…), rename all functions and their argument. For back-compatibility, the original functions and arguments are kept but flagged as deprecated. More details here.
• main exported function name change:
  • rename fc_estimate_RoC() to estimate_roc() (#23)
  • rename fc_plot_RoC_sequence() to plot_roc() (#24)
  • rename fc_detect_peak_points() to detect_peak_points() (#25)
  • rename all internal functions (#26)
• Update README (#27)

RRatepol 1.1.0

neotoma -> neotoma2

• update the vignette to use neotoma2 package (neotoma was deprecated)

utility functions Internal checks

• change util_* functions to use RUtilpol package

Function references

• remove several functions references by flagging them as internal

Bug fixes

• correctly not transform to proportions when selected by a user.

Other

• within select() replace .data$foo with "foo" (suggested in tidyselect)
• replace 1:length(... with seq_along

RRatepol 1.0.0

General

• package has been re-factored (most internal processes have been updated)
• several arguments of functions have been changed and marked with lifecycle
• add several more checks while data sourcing

Dissimilarity calculation

• use vegan package to calculate all dissimilarities except of standardised euclidean.

parallel computation

• use the pbapply package
• data for iteration is prepared separately
• added progress bar

example data

• names for example columns as been adjusted to avoid using .

RRatepol 0.6.1

Dissimilarity calculation

• use vegan package to calculate Gower and Bray-Curtis (new) dissimilarity

Univariate RoC estimation

• fix issues when using only single-taxon dataset

Code cosmetics changes

• TRUE/FALSE written as full
• all arguments are written for each function
• spelling fixes

RRatepol 0.6.0

Overhaul change of the estimation of RoC

• RoC score is now returned in units selected by user (time_standardisation)
• Method of sample selection from bin is now optional (bin_selection)
• RoC is estimated only for subsequent bins (only_subsequent == TRUE)
• Bins are now created form the beginning of the core (instead of 0)

Console outputs

• General overhaul of console outputs presented to user while running RRatepol
• more information and warning messages are presented with each run

doParallel

have been curated by you during the last years packages doSNOW and snow are replaced by doParallel

• progress bar for randomisation is currently not present

Other

• Added various checks for correct argument selection
• all hyphens are replaced with dashes (–) in the whole package
• examples are wrapped into dontrun

RRatepol 0.5.6

README

• Figures are now saved in the new folder (man folder) with the new names.
• cosmetics changes in the code (spaces, new lines, etc)

Vignette

• stop using pre-saved data
• decrease the i_multiplier to 0.5 to speed up the vignette building
• change the treads arguments in fc_estimate_RoC to FALSE so the vignette can be build in machines without multiple cores.
• rename r chunks to remove empty spaces in names

Other

• remove RRatepol namespace inside package for internal functions
• fix couple of typos in function descriptions
• add cran-comments

RRatepol 0.5.5

• sample.id or sample_id can be used in as the sample identification
• RRatepol::fc_plot_RoC_sequence: fix typo in argument age_threshold
• citation(package = “RRatepol”) now show correct citation
• update the Figure in README file