Ewstools

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

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

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.

Now contains tools for bootstrapping and eigenvalue reconstruction.

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