Release notes

Release notes#

[Unreleased]#

[0.12.0] - 2026-08-17#

Changed#

  • Breaking: check grid shape suitability for chosen strategy and level selection. Unsuitable grids will be rejected to prevent distorted results.

  • Breaking: new richness analysis (former a4) implementation that’s better in line with the original one from Palmer & White.

  • Breaking: new spatial analysis (former a5) implementation that’s better in line with the original one from Palmer & White. Previous one resulted only in single ratio for the entire grid for any given level.

  • Breaking: force-merge quadrats in the same transect for transect-centric strategies. This is to prevent behaviour where in some analyses extent grew significantly in one direction while stagnating in the other.

  • Breaking: replace species abundance distribution part of a8 with cumulative distribution function plot

  • Breaking: rename distances from diagonal, centroid and max1d to euclid-diagonal, euclid-centroid and chebyshev respectively

  • Breaking: remove taskfile table [locality]. Keys that it previously contained were moved into top-level table. This both simplifies API for analysis config creation and reduces room for error in taskfiles (putting keys under wrong table) for people new to TOML. In addition name and sides keys formerly under [locality] were renamed to locality_name and quadrat_sides respectively.

  • Breaking: shift AOPK and Biolib coordinates to (0, 0) during conversion.

  • Breaking: drop join-analyses subcommand from misc.

  • Breaking: drop analysis 6. Various ratios were simply more expressions of quadrat (dis)similarity. Users are encouraged to use Jaccard dissimilarity analysis instead.

  • Breaking: –drop-nas preprocess CLI option renamed to –drop-null

  • Breaking: Any previous output may slightly differ regardless of used seed. Changes include reordered rows, overview aux formatting, or different order of accumulations.

  • Breaking: filter commutative duplicates in dissimilarity analysis.

  • Breaking: drop include_tiny, exclude_families, include_families, exclude_phases, and locality_name config keys from taskfile. Replaced with more flexible include-enc/exclude-enc keys.

  • Breaking: rename striped transect merging strategy to nested transects.

  • Breaking: rename repeated transect merging strategy to contiguous transects.

  • Breaking: drop numerical names of analyses (a1-a8), for proper names see documentation.

  • Breaking: change plotting backend. This means appearance of every figure produced by seal plot has been changed.

  • Breaking: drop auxiliary data of richness analysis.

  • Breaking: round CLI results to 4 decimals.

  • Replace terse and too-concise README with full-fledged documentation

  • Rename secondary analysis results from additional data to auxiliary data.

  • Allow dumping in-memory taskfiles into files.

  • Unify output figure size.

  • Add auxiliary data with number of resulting NaNs to dissimilarity analysis.

  • Support distance binning in dissimilarity analysis.

Added#

  • Add difference of encountered species for quadrat pairs in turnover analysis (former a3).

  • Add subcommand to quantify and plot auxiliary data results from species–area relationship analysis (former a2). seal misc adjust-grid subcommand was added to help adjust grid for suitability criteria conformance.

  • Create an API to enable use as a library (analysis, preprocess and misc)

  • Support analyzing arbitrary columns with with categorical_col and quantity_col config keys

  • Support dumping grids transformed for requested levels via seal misc dump-transformed

  • Support Sørensen and Bray–Curtis dissimilarity in dissimilarity analysis (a7)

  • Output modified taskfile in seal misc adjust-grid

  • Add typo-similar preprocess check to detect potential typos based on Jaro-Winkler string similarity (threshold 0.95).

Removed#

  • Breaking: remove all strategies except nested-quadrats, repeated-transect-merging (renamed to contiguous-transects) and striped transect merging (renamed to nested-transects). They were pretty much just experiments and not really useful or plainly wrong.

  • Breaking: remove estimation of expected number of species from abundance analysis (former a8). Recent literature shows previously used Preston’s formula is based on wrong assumptions, such as employing bins of exponentially increasing size resulting in hump-shaped pattern that is not congruent with the shape of untransformed distribution, and should not be used.

  • Breaking: remove subcommands from misc and checks from preprocess that required quadrat_id column with overly specific format.

  • Remove Jupyter notebook for lack of interest. Target audience indicates strong preference for R interface instead.

Fixed#

  • Fix windows compatibility by enforcing utf-8 encoding.

  • Fix binning in abundance analysis (former a8) to include number of individuals equal to the rightmost interval boundary.

  • Display description and help string whitespaces correctly.

  • Fix per-quadrat stats computation in overview’s (former a1) auxiliary data.

  • Prevent iterative division of coordinates in RTM (newly CT) strategy, which resulted in coordinates having lower values than they should in higher levels.

  • Fix incorrect quadrat side and coordinate recalculation when discarding transects or zones

  • Fix centroid and quadrat adjustment for every strategy.

  • Handle empty intermediate results in analyses with binned distance.

  • Do not try to save (and crash) preprocessed dataset if --output argument not specified.

[0.11.0] - 2024-04-26#

Changed#

  • Breaking: change taskfile format to TOML

  • Breaking: make quadrat_type column in quadrat lists optional. The default behavior of quadrat_types parameter in taskfile is now not to filter anything. Use quadrat_type = ["normal"] to retain old behavior.

  • Do not make use of quadrat_id column in input data thus making it optional

  • Make optional every column except species, coord_x and coord_y

  • Make quadrat list optional

  • Always shift AOPK and Biolib coordinates by the same offset

  • Don’t include dummy columns while converting data

  • Display medians as numeric value in analysis 5 plots

  • Display points in analysis 5 with light border (smaller than outliers)

  • Add more spinner emoji

  • Make colormap size to be more in line with size of heatmap in overview (former a1)

  • Improve post-level adjustment error reporting

  • Improve ticks spacing in SAR (former a2) graph

  • Improve quadrat list type-checking

Added#

  • Add barplots depicting (species+individuals)/quadrat for overview (former a1)

  • Cover equitability of abundance, species abundance distribution and theoretical abundance data by newly added abundance analysis (former a8)

  • Select graphs format using output_format in taskfile

  • Add quadrat list example for CZ KFME grid usable with AOPK or Biolib data. This is a complete grid, unlike the previously removed example.

Removed#

  • Breaking: remove --png CLI option. Use output_format in taskfile.

  • Delete unnecessary AOPK quadrat list example as quadrat lists were made optional.

Fixed#

  • Properly seed RNG for reproducible plots

  • Biolib conversion now converts x coordinates in proper order. They were switched pair-wise due to a bug. E.g. 0<->1, 2<->3…

  • Don’t overwrite analysis results when computing multiple analysis of the same type (with possibly different analysis parameters)

  • Actually use error-type config key

[0.10.0] - 2024-02-07#

Added#

  • Initial release