rhosa
Higher-Order Spectral Analysis
Higher-order spectra or polyspectra of time series, such as bispectrum and bicoherence, have been investigated in abundant literature and applied to problems of signal detection in a wide range of fields. This package aims to provide a simple API to estimate and analyze them. The current implementation is based on Brillinger and Irizarry (1998)
rhosa: Higher-Order Spectral Analysis in R
This package aims to provide functions to estimate higher-order spectra or polyspectra of multivariate time series, such as bispectrum and bicoherence (Brillinger and Irizarry 1998). They are useful for e.g. detecting nonlinear interaction between stationary time series driven by periodic signals (Abe et al. 2024).
Installation
You can install the released version of rhosa from CRAN with:
install.packages("rhosa")
Alternatively, the development version from GitHub with remotes:
# install.packages("remotes")
remotes::install_github("tabe/rhosa")
Acknowledgement
The author thanks Alessandro E. P. Villa for his generous support to this project.
References
Abe, Takeshi, Yoshiyuki Asai, Alessandra Lintas, and Alessandro E. P. Villa. 2024. “Detection of Quadratic Phase Coupling by Cross-Bicoherence and Spectral Granger Causality in Bifrequencies Interactions.” Scientific Reports 14 (1): 8521. https://doi.org/10.1038/s41598-024-59004-8.
Brillinger, D. R., and R. A. Irizarry. 1998. “An Investigation of the Second- and Higher-Order Spectra of Music.” Signal Processing 65 (2): 161–79. https://doi.org/10.1016/S0165-1684(97)00217-X.