LoginSign Up
← Back to all sparks
D

dubicube

ANALYTICS
Velocity0.0

R package dubicube by b-cubed-eu — release notes from GitHub.

github.com/b-cubed-eu/dubicube ↗

dubicube grew from a bootstrap helper into the uncertainty layer other B-Cubed packages call.

biodiversitydata cubesbootstrappinguncertaintydata qualityr package
◆Current state
dubicube supplies bootstrapping and confidence-interval machinery for biodiversity data cubes in the B-Cubed project. The 0.10–0.12 series added the things a library needs to be depended on rather than copied: automatic detection of group-specific versus whole-cube bootstrapping, an optional boot backend, and then a second capability area in 0.12.0 with data quality diagnostics and cube filtering. The sibling indicator package b3gbi now delegates its confidence intervals here.
◆Where it's heading
Release notes are terse — usually one line and an issue number — but the direction is legible in what gets automated. Decisions the caller used to make explicitly are being inferred: resampling scope in 0.10.0, the no-bias option in 0.11.0, and process_cube_args threaded through filter_cube() so the filtering path matches cube processing. The diagnostics work in 0.12.x is the newer line, and 0.12.2's rename of the heatmap option to rule suggests that surface is still settling.
◆Prediction
The diagnostics and filtering additions have needed a follow-up fix in each of the two releases since they landed, so the next release is most likely more consolidation there rather than a new capability area.

◆Recent moves

  1. 1mo ago

    Package build fixes

    A single-line release fixing package build errors, with no user-visible change described.

    View source ↗
  2. 3mo ago

    Filter vignette documentation and rule-function fix

    Documents process_cube_args in the filter function vignette and fixes a rule-function bug surfaced by updating the vignette's example cube. Both items are scoped to the tutorial material rather than the API.

    View source ↗
  3. 3mo ago

    process_cube_args in filter_cube(); heatmap option renamed to rule

    Adds a process_cube_args argument to filter_cube() so filtering can pass through the same cube-processing options as the rest of the pipeline, and renames the heatmap option in plot.cube_diagnostics() to rule. Tutorial content was revised after internal review.

    View source ↗
  4. 4mo ago

    Data quality diagnostics and cube filtering

    ⚡ SPARK

    Opens a second capability area alongside bootstrapping: data quality diagnostics and data filtering for cubes. The one-line note is terse, but the two releases that follow reference filter_cube() and plot.cube_diagnostics() directly, confirming both landed as user-facing surface.

    View source ↗
  5. 5mo ago

    Zenodo grant ID and metadata fixes

    Metadata and archival housekeeping: a grant ID for Zenodo integration, a ROR URL fix, a publisher field, and a lighter pkgdown build.

    View source ↗
  6. 6mo ago

    No-bias bootstrap option automated

    Automates the no-bias option in the bootstrap functionality, continuing the pattern of inferring a setting the caller previously had to choose, and improves the figures in the whole-cube versus group-specific tutorial.

    View source ↗