Skip to contents

popmaps2 0.1.0

  • Prepared the first public alpha release of popmaps2. The package is installable, tested, documented, and suitable for validation work, but the API may still change before CRAN submission and the software manuscript.

  • Added a manual full-platform R-CMD-check workflow for release-readiness checks across Ubuntu, macOS, Windows, and R-devel.

  • Added a bundled start-here example script, early coordinate/raster/ancestry validation, clearer input errors, and optional progress messages for popmaps() and tune_popmaps().

  • Added a quick-start vignette and shortened the README so first-time users can start with the minimal workflow and then move into focused workflow articles.

  • Seeded package from the USGS POPMAPS 1.03 release.

  • Renamed the package to popmaps2.

  • Added package metadata for GitHub-based development.

  • Split core dependencies from optional legacy modeling/plotting dependencies.

  • Added clearer errors when optional packages are required.

  • Fixed default argument matching in legacy plotting functions.

  • Added README and roadmap documentation for modernization and public release.

  • Added terra-first input preparation and validation for raster and location inputs.

  • Allowed descriptive location column names while preserving the legacy column-order contract.

  • Updated anc_extract() to use terra cell lookup and clearer coordinate validation.

  • Removed the retired rgeos package from metadata and updated popmap_viz() boundary drawing.

  • Added validate_popmaps_baseline() for repeatable comparison against a frozen POPMAPS 1.03 reference result.

  • Added an optimized geographic-distance engine for popmaps(surface = "G") that precomputes cell-site and empirical-site distances while preserving POPMAPS 1.03 outputs on validation cases.

  • Added popmaps_rast() and write_popmaps() to convert popmaps() output to terra rasters and export GeoTIFF layers.

  • Added tools/validate-aslo-local.R for local, larger-scale ASLO validation without committing raw project data.

  • Added tune_popmaps() for fast leave-one-site-out tuning of geographic-distance POPMAPS parameters with fold-level diagnostics and summary metrics.

  • Added vectorized empirical_pt_dist tuning plus suggest_tuning_grid() and adaptive_tune_popmaps() for data-adaptive parameter exploration.

  • Added half_distance_km and ten_pct_distance_km tuning summaries plus spatial-block validation for more biologically interpretable parameter selection.

  • Added diagnose_tuning() to summarize tuning strength, near-best parameter support, and parameter effects.

  • Added tools/validate-example-tuning.R for repeatable local tuning validation across empirical example datasets kept outside the package repository.

  • Added tools/summarize-example-tuning.R to create empirical tuning summary tables, a markdown report, and diagnostic plots.

  • Added repeated spatial-block validation with reproducible rotated spatial partitions and repeat-level uncertainty summaries.

  • Documented the scientific contract for comparing geographic-distance (surface = "G") and SDM suitability-weighted least-cost (surface = "C") surfaces.

  • Added prepare_popmaps_surface() to declare candidate raster semantics, including suitability, conductance, and resistance-to-conductance transformations.

  • Added internal least-cost distance helpers that mirror POPMAPS 1.03 gdistance conductance distances on small validation rasters.

  • Extended tune_popmaps(surface = "C") to score suitability-, conductance-, and resistance-weighted least-cost surfaces with the modern distance helper.

  • Added compare_popmaps_surfaces() for matched predictive comparison of user-supplied candidate surfaces.

  • Added suggest_surface_tuning_grid() and surface-specific defaults in compare_popmaps_surfaces() so geographic and least-cost surfaces can be tuned on their own distance scales.

  • Added plot_surface_comparison() and write_surface_comparison_report() to turn candidate-surface comparisons into diagnostic figures, CSVs, and Markdown reports.

  • Added tools/compare-example-surfaces.R for local empirical G-vs-SDM surface comparisons, summary tables, reports, and plots across uncommitted example datasets.

  • Updated tools/compare-example-surfaces.R so each completed species/validation comparison writes the standard surface-comparison report artifacts.

  • Replaced the full popmaps(surface = "C") least-cost interpolation engine with the internal distance helper, removing the gdistance requirement from the main ancestry-surface workflow while preserving the original suitability-as-conductance default.

  • Added shared resource configuration for local validation/reporting scripts, including cross-platform processor detection, conservative thread defaults for system libraries, terra memory settings, and run-summary reporting.

  • Added tools/benchmark-aslo-local.R for repeatable small-, medium-, and full-scale ASLO runtime and memory checks.

  • Added input converter helpers: surface_from_points(), locs_from_sf(), surfaces_from_raster_stack(), surface_from_eems(), and surface_from_feems().

  • Added a surface-comparison vignette and a bundled-data example script that uses surface_from_points() plus write_surface_comparison_report().

  • Added a pkgdown scaffold, clearer vignette installation notes, and runnable examples for surface input converter help pages.

  • Added a manual pkgdown deployment workflow, reduced routine R-CMD-check triggers to pull requests/manual runs, cleaned the README, and annotated local validation scripts.

  • Corrected default popmaps() raster-cell coordinates to use true raster cell centers, with legacy_compat = TRUE retained for POPMAPS 1.03 reproducibility.

  • Updated default surface = "C" interpolation to select candidate empirical sites by least-cost distance, while legacy geographic prefiltering remains available only through legacy_compat = TRUE.

  • Added unit-neutral half_distance and ten_pct_distance tuning summaries while retaining the older _km columns for compatibility.

  • Changed optional conductance rescaling to divide by maximum conductance so zero-valued barriers remain zero instead of turning the minimum positive conductance into a barrier.

  • Added plot_popmaps() and write_popmaps_plot() as the modern POPMAPS visualization layer for ancestry-probability, hard-boundary, and ancestry-axis map outputs, including a manuscript-style preset inspired by Massatti and Winkler (2022).