GapFinder Rocks

This is an interactive example of the GapFinder approach to exploring large distance matrices that have a hierarchical structure to their observations. Applied to a small data sample representing forty geological observations, scalable strategies are presented that accurately visualise pairwise dissimilarities (or distances) between observations from a hierarchical taxonomy, in ways that use the most effective visual channel of the human visual system. \nLineage: This work was motivated by a need to explore distances between DNA sequences in relation to the Linnean taxonomy of biological classification and builds on the GapFinder work of Xin-Yi Chua et al. (2022). The interface implements the following R packages and their dependencies: rmarkdown, shiny, plotly, ggplot2, data.table, ggbeeswarm, ggdendro, scico, philentropy.