v2.1.2

Added function to compute various entropy measures.

v2.1.1

Updates:

• Paper written for JOSS

v2.1.0

Updates:

• Deprecated function ews_compute() now removed. All computation of EWS must be done through the TimeSeries class and associated methods.
• Tutorials that used deprecated functions removed
• Small fix to apply_classifier_inc() to make sure that classifier is applied to all time values in the range provided

v2.0.1

Major changes include:

• A TimeSeries class : In version 2, users put their data into a TimeSeries class, and compute EWS by applying associated methods to this class. In older versions, users would use the stand-alone function ews_compute(), however this became impractical as I wanted add more features to the package. Now, when a new feature is desired (e.g. a new type of EWS), we can add a new method to the TimeSeries class. This class also allows us to store local variables associated with the data without having to pass them to each function.
• Methods to apply a deep learning classifier to make predictions : apply_classifier() and apply_classifier_inc().
• A method to visualise EWS with an interactive figure in Plotly: make_plotly()
• Added tutorials and documentation

First official release

Python package for computing, analysing and visualising early warning signals (EWS) in time series data. Includes a novel approach to characterise bifurcations using Spectral EWS.

ewstools: first release

Now contains tools for bootstrapping and eigenvalue reconstruction.

Pre-release version of ewstools

Python package for computing, analysing and visualising early warning signals (EWS) in time-series data. Includes a novel approach to characterise bifurcations using EWS.