When doing exploratory analysis of time series, frequency domain methods, that is, statistical methods that display information about the temporal correlations of one or more time series in terms of frequencies, can be used to infer physical mechanisms for underlying process dynamics in e.g. geophysical time series, medical time series, etc.). The multitaper method, which leverages Slepian functions to estimate power spectral densities, coherences, and so forth, is implemented here for application to univariate, multivariate, and higher-dimensional (e.g. space-time) processes.
See documentation below.
Please install using
Pkg> add Multitaper
This package runs on julia v 1.4.2 and above.
If you make use of Multitaper.jl, please cite the following paper: 
Python users may wish to make use of the python wrapper for this package
. A demo for its useage
is available
.
We also highly recommend the multitaper python package by G. A. Prieto, available .
We welcome input of any kind via issues or by pull requests. Support requests can be directed to haley@anl.gov.