The rankclust R package implements the Rank-Clustered Bradley-Terry-Luce model proposed by Pearce and Erosheva (2025). This Bayesian model for ordinal comparison data of many different types (such as complete or partial rankings, pairwise comparisons, etc.) infers an overall ranking of objects while permitting some objects to be “rank-clustered”, i.e., assigned equal rank in the overall ranking.
To learn more about the novel rank-clustering methodology implemented in this package, check out our 2025 paper in Psychometrika.
For information on how to download the rankclust package from Github, see the Overview tab. After, check out the Tutorial and Reference tabs for details on the package’s use.