bioregion 1.5.0.9000
This is a list of changes made in the development/GitHub version of the package between bioregion 1.5.0 (CRAN release 2026-10-10) and the next CRAN release.
- Multiple bug fixes following recent name changes in bioregionalization_metrics(): fixed
print()outputs offind_optimal_n(), fixed thecluster_infoofhclu_diana(), re-implemented the inclusion of the output offind_optimal_n()inton_clustargument ofcut_tree(),hclu_hierarclust()andhclu_diana()(documented but broke in recent updates), fixedprint()andstr()on the output ofsite_species_metrics(), which did not list thesilhouettemetric and, forstr(), theprop_endemicsmetric), fixedprint()on the output ofhclu_hierarclust(). - New
plot()for all clustering outputs (?plot.bioregion.clusters), with three types of plots (plot_type):-
"bars", for all algorithms: one bar per cluster, made of one square per site, and, if bipartite, one circle per species. -
"tree", forhclu_hierarclust(),hclu_diana()andcut_tree()(the default for these functions), and for nested partitions such as the levels ofnetclu_infomap(). The clusters are shown as colored bars under the tree (one bar per partition) instead of rectangles. -
"cluster_tree": the tree drawn down to the clusters, with the bar of each cluster at the tip of its branch (the default for hierarchicalnetclu_infomap()). - cool customizations for all three plots, check out the visualization vignette
- arguments that do not apply to the chosen plot (e.g.,
collapsefor bars) are ignored with a warning.
-
-
Breaking change: in
map_bioregions(), a numericpartition_indexnow gives the number of the partition (1for the first partition) instead of its column inbioregionalization$clusters(where the first partition was2). Code using numbers must subtract 1 from them; names are not affected. A message recalls this change the first time numbers are used in a session. -
bioregion_colors()now ranks clusters of the same size by name, with numbers sorted as numbers (2.1,2.2, …,2.10), instead of by their order of appearance, and itscolorselement lists clusters in the same order. - With
map_as_output = TRUE,map_bioregions()now adds the colors ofbioregion_colors()to its output when the bioregionalization has them: one column per partition, named after the partition with the suffix_color(e.g.,K_3_color), to reuse the colors in custom maps. -
map_bioregions()now uses the default colors ofbioregion_colors()when the bioregionalization has no colors (instead of the default colors ofsf), so that colors match among the package visualization functions when default parameters are used.
bioregion 1.5.0
This is a list of changes made in the development/GitHub version of the package between bioregion 1.4.0 (CRAN release 2026-03-29) and bioregion 1.5.0 (CRAN release 2026-10-10).
- The tree evaluation function is now implemented in C++ to be faster across all methods in
hclu_hierarclust(). - The iterative hierarchical consensus tree is now available on its own with the new function
ihct(). Inhclu_hierarclust(), the default ofoptimal_tree_methodis now called"ihct“” which now runs 5 to 7 times faster. The former name"iterative_consensus_tree"is still accepted and gives the same result (PR #38). - Integrated fixes for the upcoming release of dbscan 1.2.7 (PR #37).
- Minor fix to the map_bioregions() documentation (#35).
- Improved get_pairwise_membership efficiency (PR #32).
- Metric names have been changed to snake_case in
site_species_metrics(),bioregion_metrics()andbioregionalization_metrics(). The package is being progressively standardized on snake_case metric names. Old metric names are still accepted as inputs but are converted to snake_case in outputs. -
find_optimal_n()has been updated.bioregionalizationsargument changed toevaluation_df. -
bioregionalization_metrics()has been updated. Theanosimmetric (along with the associated p-value) is now computed withvegan. Theeval_metricargument changed toeval_metricsandnettocomat.col_sites, andcol_speciesarguments are now deprecated. - Relaxed matching controls in
site_species_metrics()andbioregion_metrics().
bioregion 1.4.0
This is a list of changes made between bioregion 1.3.0 (CRAN release 2026-01-23) and bioregion 1.4.0 (CRAN release 2026-03-29).
-
bioregion_metrics()has been updated. It now supports multiple bioregionpartitions and spatial coherence can be computed using either the size orthe area of the largest contiguous component. Thecol_bioregionargumentis now deprecated. -
map_bioregion()has been reformatted:bioregionalizationargument changed topartition_index,clusterstobioregionalization,geometrytomapandwrite_clusterstomap_as_output. -
map_bioregion()can now handleterrainputs (terraadded inSuggests). -
clustersargument changed tobioregionalizationinsite_species_subset(). - Added
seedargument tohclu_hierarclust()(issue #10). - Fixed invalid input, control and output in
nhclu_affprop()(issue #14).
bioregion 1.3.0
This is a list of changes made between bioregion 1.2.0 (CRAN release 2025-01-31) and bioregion 1.3.0 (CRAN release 2026-01-23).
Function changes
- Added the
inputs$data_typefield to all clustering outputs to explicitly track whether original co-occurrence data were occurrence-based or abundance-based. This field is automatically determined based on the algorithm type and the similarity/dissimilarity metric used. - Added the
inputs$node_typefield to all clustering outputs to explicitly indicate whether the clustering includes only sites or both sites and species. This changes include the hiddennode_typeattributes. -
site_species_metrics()has been thoroughly reformatted and now provides: species-per-bioregions metrics, species-in-bioregionalization metrics, site-per-chorotypes metrics, site-in-chorological classification metrics, site-per-bioregions metrics, and site-in-bioregionalization metrics, all of which are rigorously defined in the corresponding vignette. -
site_species_subset()has been simplified taking advantage oninputs$node_type. - Renamed the class
bioregion.pairwise.metrictobioregion.pairwise. - Added a
verboseargument to all talkative functions allowing users to control the display of progress messages. -
bioregionalization_metrics()now returns adata.frame(columnspartition,n_bioregions, then one column per metric) instead of alistwith anevaluation_dfelement. Code that usedx$evaluation_dfshould now usexdirectly.
New features
- Added export of the function
exportGDF()with documentation and tests. - Added
bioregion_colors()to provide consistent bioregion color palettes for use across multiple visualizations (maps, networks, graphs, etc.). - Updated
map_bioregions()to handle bioregion colors. - Added a generic function
summary()for a clearer display of results. - Added
bind_pairwise()to combine pairwise (dis)similarity objects. - Added
as_bioregion_pairwise()to replace and improve uponbetapart_to_bioregion(), which is now deprecated. - Added a comparison with other R packages for computing dissimilarity metrics in tutorial 3 (Pairwise similarity/dissimilarity metrics).
Bug fixes
- Fixed an tibble input problem with
net_to_mat()(issue #11). - Fixed a problem in
inputs$pairwise_metricwhen numericindexin all clustering outputs. - Modified the
keep_trialsargument inhclu_hierarclust()and fixed a potential issue with randomized matrix storage.
bioregion 1.2.0
This is a list of changes made between bioregion 1.1.1 (CRAN release 2024-04-19) and bioregion 1.2.0 (CRAN release 2025-01-31).
Added affinity propagation algorithm (
nhclu_affprop()).Added a new method in
hclu_hierarclust()to construct a consensus tree called Iterative Hierarchical Consensus Tree (IHCT). This resolves issues related to the order of sites in the distance matrix and builds a consensus hierarchical tree with meaningful topology.Made many changes to functions related to
hclu_hierarclust()due to this major update.Updated generic functions to provide
plotandprintmethods forhclu_diana().Added
site_species_metrics()to the package and workflow.Added
bioregion_metrics()to the package and workflow.Renamed
subset_node()tosite_species_subset().Added indices
Cztosite_species_metrics().-
Updated
install_binaries():- Archive
bin.zipnow stored on GitHub and backed up on NextCloud. - Added Infomap version 2.8.0.
- Added argument
download_onlyto execute only the download step.
- Archive
Added
check_installargument tonetclu_infomap(),netclu_louvain(), andnetclu_oslom().Added
betapart_to_bioregion()to the package.Added
compare_bioregionalizations()to the package.Added
bioregionalization_metrics()to the package.Updated documentation, vignettes, and tests.
Modified the way seeds are generated for
nhclu_clara()andnhclu_clarans().
bioregion 1.1.1
This is a list of changes made between bioregion 1.1.0 (CRAN release 2024-03-19) and bioregion 1.1.1 (CRAN release 2024-04-19).
- Added hierarchy support for Louvain (C++).
- Added
seedargument to stochastic algorithms (except Louvain C++). - Added
cut_weightargument tonetclu_*functions. - Changed value for sites without clusters from
0toNA. - Updated automated tests (code coverage > 60%).
- Standardized controls, inputs, and outputs.
- Fixed a bug in
find_optimal_n()for cases where partition metrics did not vary.
bioregion 1.1.0
This is a list of changes made between bioregion 1.0.0 (CRAN release 2023-04-15) and bioregion 1.1.0 (CRAN release 2024-03-19).
- Added the
resolutionparameter to the igraph Louvain implementation. - Added options to
mat_to_net()to exclude diagonal and lower triangular matrices usinginclude_diagandinclude_lower. - Added a function to extract a subset of nodes (sites or species) from
bioregion.clustersobjects containing both types. - Added a generic function to maintain attributes of
bioregion.pairwiseobjects and track the number of sites and species. - Added new functions:
nhclu_clara()andnhclu_clarans(). - Edited vignettes to document new functions.
- Modified controls for
bioregion.pairwiseobjects. - Added the
include_formulaargument tosimilarity_dissimilarity_conversion()to (not) select formula metrics. - Allowed negative values in
similarity()with the Euclidean metric.