NEWS.md

tabula 3.1.1.9000

tabula 3.1.1

Internals

• Use palette functions from khroma.

tabula 3.1.0

New classes and methods

• Add she() to plot SHE analysis.
• Add profiles() to plot diversity profiles.
• Add test_shannon() and test_simpson() to compare the Shannon and Simpson diversities in two samples.
• Add index_squares() to compute Alroy (2018) richness estimator.

Enhancements

• plot_heatmap() gained a new argument to use a fixed aspect ratio.
• Improve color gradient in plot legends.

Bugfixes & changes

• Fix ACE species richness estimator.
• Deprecate test_diversity().

tabula 3.0.1

Enhancements

• Omit axes labels where they would abut or overlap previously drawn labels.

tabula 3.0.0

Bugfixes & changes

• Fix handling of missing values in heterogeneity index (#27).

Breaking changes

• Use graphics instead of ggplot2 for plotting.* The internal use of ggplot2 was badly interoperable or composable. This also reduces hard dependencies.

Internals

• Use tinytest and tinysnapshot instead of testthat and vdiffr.

tabula 2.0.0

New classes and methods

• Add plot_diceleraas() to display a Dice-Lerass diagram.
• Add matrigraph() to display a heatmap highlighting the deviations from independence.
• Add seriograph() to display a Ford diagram highlighting the relationships between rows and columns.

Bugfixes & changes

• EPPM argument of plot_ford() is defunct (use seriograph() instead).

Breaking changes

• Seriation methods are no longer reexported from kairos.

tabula 1.8.0

Seriation methods are now reexported from kairos and will be removed in a future release.

New classes and methods

• Add index_boone(): Boone heterogeneity index.
• Add index_baxter(): Baxter rarefaction index.
• Add the cantabria dataset.

Bugfixes & changes

• Remove all previously deprecated methods.

tabula 1.7.0

Seriation methods are now reexported from kairos and will be removed in a future release.

New classes and methods

• Add RarefactionIndex: S4 class that represent rarefied species richness.
• Add bootstrap(), jackknife() and simulate() methods to perform bootstrap and jackknife resampling and to measure diversity in simulated assemblages.
• Add heterogeneity(), evenness(), richness() and composition() methods for matrix and data.frame.
• Add simulate() to simulate observations from a multinomial distribution.
• Add autoplot() and plot() methods for RarefactionIndex objects.

Bugfixes & changes

• rarefaction() now returns a RarefactionIndex object.
• Deprecate bootstrap_*(), jackknife_*() and simulate_*().
• Deprecate index_heterogeneity(), index_evenness(), index_richness() and index_composition().

Breaking changes

• turnover() no longer accept multiple methods as argument.
• Fix bootstrap() method for DiversityIndex objects: resample with replacement instead of simulating observations from a multinomial distribution.
• No longer use classes from arkhe: all methods are now defined for matrix and data.frame.
• Deprecate seriation methods: seriate_*() and permute() now belong to kairos.

tabula 1.6.1

Bugfixes & changes

• Remove all previously deprecated methods.

Breaking changes

• Remove dating methods: *_mcd(), *_event(), *_accumulation(), plot_date() now belong to kairos.
• Remove frequency increment test: test_fit() and plot_time() now belong to kairos.

tabula 1.6.0

New classes and methods

• Add DateEvent: S4 class to store the event and accumulation times of archaeological assemblages.
• Add CompositionIndex: S4 class that represent an asymptotic species richness.
• Add IncrementTest: S4 class that represent a Frequency Increment Test results.
• Add eppm() and pvi() to calculate independence statistics.
• Add predict_event() and predict_accumulation to estimate the event and accumulation dates of an assemblage.
• Add bootstrap_*() and jackknife_*() to perform bootstrap and jackknife resampling.
• Add simulate_evenness() and simulate_richness() to measure diversity in simulated assemblages.
• seriate_average() replaces seriate_correspondance().
• seriate_rank() replaces seriate_reciprocal().

Bugfixes & changes

• Remove all previously deprecated methods.
• Deprecate seriate_correspondance(), seriate_reciprocal() and refine_dates(),
• index_composition() now returns a CompositionIndex object.
• similarity() now returns a dist object.
• date_mcd() now returns an DateMCD object.
• test_fit() now returns an IncrementTest object.
• DiversityIndex lost bootstrap() and jackknife() slots.
• Rename BootCA to RefineCA.
• Rename DateModel to DateEvent.
• RefineCA and DateEvent now inherit from dimensio::CA.

Enhancements

• date_mcd(), date_event() and plot_time() gained a new dates argument.

Internals

• Imports arkhe >= 0.3.0
• Re-export methods from arkhe.
• Imports dimensio.
• Use datasets from folio.
• Remove pbapply from suggested packages.
• Rewrite plot_ford() and plot_bertin().

tabula 1.5.1

Bugfixes & changes

• CRAN pacakge check warning (r-devel) has been fixed (use a stringsAsFactors = FALSE default).

tabula 1.5.0

Bugfixes & changes

• CRAN package check error with long doubles disabled has been fixed (tested with R-hub debian-gcc-devel-nold).
• CRAN package check notes have been fixed.

Internals

• Depend on arkhe.

tabula 1.4.0

• Published in the Journal of Open Source Software.

New classes and methods

• DiversityIndex, HeterogeneityIndex, EvennessIndex and RichnessIndex: S4 classes that represent diversity index.
• index_heterogeneity() replaces diversity().
• index_evenness() replaces evenness().
• index_richness() replaces richness().
• index_composition() allows to estimate asymptotic species richness.
• plot_diversity() produces a diversity vs sample size graph and allow to compare estimates with simulated assemblages.
• Add replacement methods for the *Matrix classes.

Bugfixes & changes

• Deprecate diversity(), evenness() and richness().

Internals

• Display progress bars only if interactive() is TRUE and pbapply is installed.

tabula 1.3.0

New classes and methods

• Matrix S4 class is now the superclass of all matrix-like classes.
• AbundanceMatrix virtual S4 class is defined as the superclass of CountMatrix, FrequencyMatrix and IncidenceMatrix.
• SpaceTime S4 class represents space-time informations.
• as_*() coerce a matrix or data.frame to a CountMatrix, FrequencyMatrix, IncidenceMatrix, OccurrenceMatrix or SimilarityMatrix.
• date_event() replaces dateEvent().
• date_mcd() allows Mean Ceramic Date estimation.
• get_dates() and set_dates<- allow to extract and replace chronological informations in AbundanceMatrix objects.
• plot_bertin() and plot_ford() replace plotBar().
• plot_date() replaces plotDate().
• plot_date() gained a method for AbundanceMatrix objects.
• plot_heatmap() replaces plotMatrix().
• plot_rank() replaces plotRank().
• plot_spot() replaces plotSpot().
• plot_time() produces an abundance vs. time graph.
• refine_dates() and refine_seriation() replace refine().
• seriate_reciprocal() and seriate_correspondance() replace seriate().
• test_diversity() allows Shannon diversity test.
• test_fit() produces a Frequency Increment Test.

Bugfixes & changes

• CountMatrix, FrequencyMatrix and IncidenceMatrix now also contain the SpaceTime class.
• Deprecate plotBar(), plotMatrix(), plotRank(), plotSpot(), refine(), seriate().
• Remove dateEvent().
• Empty rows/columns are removed prior to CA seriation to avoid error in svd().

Enhancements

• Add the Merzbach ceramics dataset.
• The plot_date() method for DateModel objects now allows to display an activity or a tempo plot.

Internals

• Reduce required R version to 3.2.
• Error handling has been revised and error messages have been harmonized.
• Refer to ggplot2 functions using :: (stop importing the entire package).
• Use vdiffr to test graphical output.
• Replace FactoMinerR::CA() with ca::ca() (this avoids having to install all {FactoMineR} dependencies when only one function is used).
• Remove dplyr from the imported packages, move magrittr to suggested packages.

Experimental

• When a Matrix object is first created, an identifier (UUID v4) is generated with generate_uuid(). This ID is preserved when coercing to another class. This makes it possible to identify objects representing the same initial data and associate them with the results of specific computations.
• get_coordinates() and set_coordinates<- allow to extract and replace spatial informations in AbundanceMatrix objects.
• get_features() allows to convert an AbundanceMatrix object to a data.frame. It is intended for compatibility with the sf package.

tabula 1.2.0

New classes and methods

• The function dateEvent() allows to compute chronological models as described in Bellanger and Husi (2006).
• DateModel this S4 class stores the results of dateEvent().
• SimilarityMatrix this S4 class represents a (dis)similarity matrix.
• plotDate() method for DateModel objects.
• plotSpot() methods for SimilarityMatrix and OccurrenceMatrix objects.
• [ operators for several classes.

Bugfixes & changes

• OccurrenceMatrix now stores the number of times each pair of taxa occurs together in at least one sample.
• similarity() now returns an object of class SimilarityMatrix.
• plotBar() no longer add confidence interval by default.
• Remove useless accessors.

Enhancements

• similarity() gained a new estimator: binomial co-occurrence assessment method (similarity between types).
• seriate() gained a new argument to pass a BootCA object.

Internals

• Add an optional progress bars with {pbapply} in long running functions.

tabula 1.1.0

Bugfixes & changes

• similarity() now returns a symmetric matrix.

Enhancements

• richness() gained new estimators:
• For abundance data: Chao1, bias-corrected Chao1, improved Chao1 and Abundance-based Coverage Estimator (ACE).
• For replicated incidence data: Chao2, bias-corrected Chao2, improved Chao2 and Incidence-based Coverage Estimator (ICE).

Internals

• Add references in the Description field of the DESCRIPTION file.
• Split the documentation for alpha-diversity measures.
• Split the documentation for beta-diversity measures.

tabula 1.0.0

• Initial version on CRAN.

New classes and methods

• BootCA this S4 class stores partial bootstrap CA-based seriation results.
• [[ operators acting on PermutationOrder and BootCA to extract parts.

Bugfixes & changes

• refine() method for CountMatrix now use stats::rmultinorm() for partial bootstrap CA.

Enhancements

• Add the Zuni and Mississippi ceramics datasets.
• similarity() gained a new estimator: the Brainerd-Robinson coefficient of similarity.

Internals

• Add a vignette for matrix seriation.

tabula 0.9.0

• Beta release.