StationaryRandomFields.jl
This Julia package is designed to provide toolkits for handling signals sampled from stationary Gaussian random fields that follow a specific stationary power spectrum in the Fourier domain. The package offers:
- Abstract types and methods for multi-dimensional discrete signals.
- Abstract multi-dimensional power spectrum types, which can be easily scaled, stretched, and rotated.
- toolkits to sample signals from input (or generated) Gaussian random fields in the Fourier space.
The package is currently in the significant stages of development. We are planning to implement:
- Chain rules to make the relevant methods compatible with Julia's standard automatic differentiation packages.
- Custom kernels for KernelFunctions.jl, enabling the integration of stationary Gaussian random fields characterized by power spectra into the rich ecosystem of the Julia Gaussian Processing packages.
Disclaimer: The package may have significant changes in near future given the early stage of the development.
Installation
The package is available on the Julia standard repository. It can be installed by
using Pkg
Pkg.add("StationaryRandomFields")Usage
To be written soon.
Acknowledgements
The development & developers of this package has been finantially supported by the following programs.
- AST-2107681, National Science Foundation, USA: v0.1.0 - present
- AST-2034306, National Science Foundation, USA: v0.1.0 - present
- OMA-2029670, National Science Foundation, USA: v0.1.0 - present
- AST-1950348, National Science Foundation, USA: v0.1.0 - present
- AST-1935980, National Science Foundation, USA: v0.1.0 - present