Skip to contents

binwise 0.0.17

  • Cite the published APSA preprint in the research vignette, help page, and case-study metadata using DOI 10.33774/apsa-2026-7wxvk.

binwise 0.0.16

  • Explain the research heatmaps through individual cells, comparisons across groups, and worked percentage-point differences.

binwise 0.0.15

  • Focus the research vignette on the reader workflow, removing manuscript revision history and local development notes while retaining statistical scope.

binwise 0.0.14

  • Add observed, representative-prediction, and difference heatmaps to the research vignette, with cell counts and a step-by-step reading guide.
  • Preserve verified paper heatmap aggregates and their audit provenance.

binwise 0.0.13

  • Add a research case-study vignette on intra-industry trade and sanctions busting, with small real-data aggregate summaries and provenance.
  • Distinguish substantive categorization, in-sample cell agreement, clustered mean intervals, and additional held-out prediction comparisons.
  • Add a local RStudio walkthrough verifying the paper’s category assignments.

binwise 0.0.12

  • Add optional natural cubic spline comparators to bin_interact() and bin_validate() using spline_df. Existing calls retain their behavior.
  • Learn spline knots separately within each validation training fold and retain them for inspection. Predictions reuse fitted spline definitions.
  • Add reference to bin_validate() for direct paired comparisons against a spline or another returned model.
  • Explain spline comparisons in model assessments and documentation.
  • Adopt MIT licensing with Keith A. Preble as copyright holder.

binwise 0.0.11

  • Add getting-started and real-data vignettes, with executed examples, interpretation guidance, related software, and primary research.