dtour: a steerable tour de vis through high-dimensional data
For data scientists and analysts exploring high-dimensional data, dtour provides a more flexible and steerable tour interface than existing tools, but the improvement is incremental.
The authors present dtour, a tour interface for high-dimensional data exploration that combines static projection previews, reversible scrubbing along geodesic paths, manual manipulation, and a grand tour in a single progressive interface. It scales to millions of points via GPU acceleration and runs in browsers, demonstrated on text, image, and single-cell data for revealing structure and validating non-linear dimensionality reduction.
Understanding high-dimensional data requires projecting it into lower-dimensional spaces, but any single projection inevitably loses information or introduces distortions. Tours address this limitation through animation of 2D projection sequences, yet existing tools present tradeoffs in the freedom and steerability of projection traversal, providing little to no ability to move between expert-guided paths and unrestrained exploration. We present dtour, a tour interface that combines static projection previews, reversible scrubbing along continuous geodesic projection paths, manual projection manipulation, and a wandering grand tour, all within a single progressive exploration interface. dtour scales to millions of points via GPU-accelerated rendering, runs in any modern browser, and integrates with both Python and JavaScript ecosystems. We demonstrate dtour on text, image, and single-cell data for two usage scenarios: gradually revealing structure in high-dimensional data and validating non-linear dimensionality reduction outputs.