Toroidal PCA via density ridges
From MaRDI portal
Abstract: Principal Component Analysis (PCA) is a well-known linear dimension-reduction technique designed for Euclidean data. In a wide spectrum of applied fields, however, it is common to observe multivariate circular data (also known as toroidal data), rendering spurious the use of PCA on it due to the periodicity of its support. This paper introduces Toroidal Ridge PCA (TR-PCA), a novel construction of PCA for bivariate circular data that leverages the concept of density ridges as a flexible first principal component analog. Two reference bivariate circular distributions, the bivariate sine von Mises and the bivariate wrapped Cauchy, are employed as the distributional basis of TR-PCA. Efficient algorithms are presented to compute density ridges for these two models. A complete PCA methodology adapted to toroidal data (including scores, variance decomposition, and resolution of edge cases) is introduced and implemented in the companion R package ridgetorus. The usefulness of TR-PCA is showcased with a novel case study involving the analysis of sea currents on the coast of Santa Barbara.
Recommendations
- Torus principal component analysis with applications to RNA structure
- Principal component analysis and clustering on manifolds
- Nonlinear kernel density principal component analysis with application to climate data
- Dihedral angles principal geodesic analysis using nonlinear statistics
- Statistical methods generalizing principal component analysis to non-Euclidean spaces
Cites work
- A Möbius transformation-induced distribution on the torus
- A multivariate von mises distribution with applications to bioinformatics
- A tractable and interpretable four-parameter family of unimodal distributions on the circle
- Analysis of principal nested spheres
- Another look at principal curves and surfaces
- Asymptotic theory for density ridges
- Bivariate models for dependence of angular observations and a related Markov process
- Dihedral angles principal geodesic analysis using nonlinear statistics
- Locally defined principal curves and surfaces
- Nonparametric ridge estimation
- Principal Curves
- Principal curves of oriented points: theoretical and computational improvements
- Probabilistic model for two dependent circular variables
- Protein Bioinformatics and Mixtures of Bivariate von Mises Distributions for Angular Data
- Recent advances in directional statistics
- Spatial analysis of wave direction data using wrapped Gaussian processes
- Theoretical analysis of nonparametric filament estimation
- Torus principal component analysis with applications to RNA structure
Cited in
(3)
This page was built for publication: Toroidal PCA via density ridges
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q82001)
