Version 0.2.2

bssdm 0.2.2

What's Changed (TL;DR)

• Add extrapolation plot methods; bundle dataset; update README by @johnbaums in #7

Full Changelog: 0.2.1...0.2.2

---

New features

• Structured result classes
  mess() and exdet() now return MessResult and ExdetResult objects respectively. These are still plain lists under the hood, so existing $-based access (e.g. result$mess, result$exdet) continues to work without change. The S3 class enables the new plot methods below.

• Plot methods for extrapolation outputs
  New plot() S3 methods for both result types (requires tmap, listed under Suggests):

  • plot(mess_result) — MESS surface with diverging colour scale (red = novel, blue = similar); which = "mod" or "mos" for categorical most-dissimilar / most-similar covariate maps with donut chart insets.
  • plot(exdet_result) — ExDet map distinguishing Type 1 novelty (univariate, red), Type 2 novelty (multivariate, blue), and analog conditions (yellow) as separate layers; which = "mic1" or "mic2" for most-influential-covariate maps.
  • Examples in ?mess and ?exdet have been simplified to use these methods.

• New bundled dataset
  hawkweed — cleaned global occurrences of Hieracium pilosella from GBIF (doi:10.15468/dl.q6q6fk), processed with CoordinateCleaner. Used in package examples.

Documentation

• README rewritten from a placeholder stub to a full worked example covering climate data download via geodata, the hawkweed dataset, Range bagging and Climatch SDM workflows, and MESS / ExDet novelty assessment — with rendered figures throughout.

Minor

• .gitignore updated to exclude .DS_Store, README.html, and Rplots.pdf.
• Stale .DS_Store removed from tests/testthat/test_inputs/.

Version 0.2.1

What's Changed

• Remove default n in call to blocks; bump to 0.2.1 by @johnbaums in #6

Full Changelog: 0.2.0...0.2.1

Version 0.2.0

• feat: add similarity analysis (mess, exdet) by @johnbaums in #5

Full Changelog: 0.1.10...0.2.0

Version 0.1.10

Increased raster block number (approx. double) to help with BSC platform memory failures.
Added option debug flag for block writing information.

0.1.9

SDM range bagging evaluation error fix (predictions > 1)

Version 0.1.8

John's fixes:

• Fixed column indexing for rangebag
• Added test for data.frame input for rangebag
• Removed points outside extent of enviro data for rangebag plus documented
• Throw error on invalid input parameters for climatch & rangebag
• Updated climatch docs to indicate d_max is in km

Version 0.1.7

Utilise terra read/write in blocks
Parallel via grouping

Version 0.1.6

Handle single predictor column for climatch

Version 0.1.5

Added matrixStats library

Version 0.1.4

Added raster filename save to predict methods