This is the beginnings of a python package for computing features of a time series using algorithms from across the scientific time-series analysis literature.
The current version is a template for such a package using partials, and is compatible with the nitime data format for time series. The goal is for useful time-series algorithms to be ported across to native python implementations, for use in open source scientific applications (including those relating to signal processing in neuroscience).
The associated (and highly-developed) hctsa Matlab repository is based on the highly comparative time-series analysis framework, which has an associated website here.