Core Concepts
The simplex projection method enables lossless visualization of 4D compositional data on a 2D canvas.
Abstract
The article introduces the simplex projection method for visualizing high-dimensional data. It addresses challenges in representing multi-dimensional data on a two-dimensional canvas, focusing specifically on 4D compositional data. The method preserves geometrical and topological properties while accurately representing the data. The proof of concept involves projecting single points, sets of points, and continuous probability density functions onto lower-dimensional facets. Recursive marginal approximations are possible through interpolation, ensuring information preservation in finite dimensions.
Introduction:
Visualizing high-dimensional data poses challenges.
Simplex projection method introduced for lossless visualization of 4D compositional data.
Preliminaries:
Barycentric coordinates used to represent points in simplices.
Convex hulls defined for subsets of vertices in simplices.
Simplex Projection:
Lossless visualization technique for preserving compositional data structure.
Bijection mapping from higher-dimensional simplex to lower-order facets demonstrated.
Related Work:
Comparison of visualization techniques like parallel coordinates, stacked plots, and simplex plots.
Advantages and limitations of each technique discussed.
Conclusion:
Simplex projection overcomes limitations of existing techniques.
Mathematical proof supports information preservation in higher dimensions.