Skip to content

dioph/periodicity

Repository files navigation

Periodicity

Useful tools for periodicity analysis in time series data.

PyPI version Downloads

Documentation: https://periodicity.readthedocs.io

Currently includes:

  • Auto-Correlation Function (and other general timeseries utilities!)
  • Spectral methods:
    • Lomb-Scargle periodogram
    • Bayesian Lomb-Scargle with linear Trend (soon™)
  • Time-frequency methods:
    • Wavelet Transform
    • Hilbert-Huang Transform
    • Composite Spectrum
  • Phase-folding methods:
    • String Length
    • Phase Dispersion Minimization
    • Analysis of Variance (soon™)
  • Decomposition methods:
    • Empirical Mode Decomposition
    • Local Mean Decomposition
    • Variational Mode Decomposition (soon™)
  • Gaussian Processes:
    • george implementation
    • celerite2 implementation
    • celerite2.theano implementation

Installation

The latest version is available to download via PyPI: pip install periodicity.

Alternatively, you can build the current development version from source by cloning this repo (git clone https://github.com/dioph/periodicity.git) and running pip install ./periodicity.

Development

If you're interested in contributing to periodicity, install pipenv and you can setup everything you need with pipenv install --dev.

To automatically test the project (and also check formatting, coverage, etc.), simply run tox within the project's directory.