PyFlux is an open source time series library for Python. The library has a good array of modern time series models, as well as a flexible array of inference options (frequentist and Bayesian) that can be applied to these models. By combining breadth of models with breadth of inference, PyFlux allows for a probabilistic approach to time series modelling.
See some examples and documentation below. PyFlux is still only alpha software; this means you use it at your own risk, that test coverage is still in need of expansion, and also that some modules are still in need of being optimized.
- ARIMA models
- Dynamic Paired Comparison models
- GARCH models
- GAS models
- GAS State Space models
- Gaussian State Space models
- Non-Gaussian State Space models
- VAR models
- Black Box Variational Inference
- Laplace Approximation
- Maximum Likelihood and Penalized Maximum Likelihood
pip install pyflux
Supported on Python 2.7 and 3.5.
- PyData San Francisco 2016 - August 2016 - a tour of time series (and predicting NFL games)
- PyData London Meetup - June 2016 - an introduction to the library in its early stages
PyFlux is still alpha software so results should be treated with care, but citations are very welcome:
Ross Taylor. 2016. PyFlux: An open source time series library for Python http://www.pyflux.com