PyFTS Versions Save

• Type hints
• New methods
• Performance improvements

• Several improvements and bugfixes;
• Performance optimization on fuzzyfication process using indexes;
• New multivariate and nonstationary methods;
• Hyperparameter optimization;
• Distributed methods for scalability;
• FTS with Fuzzy Cognitve Maps.

Several improvements and bugfixes, and the 1st Sphinx generated documentation.

• Complete refactoring for PEP 8 and PEP257 compliance
• pyPI compliance and support pip
• Seasonal, Multivariate, Nonstationary and Ensemble FTS models, partitioners and fuzzy sets.
• Unification of .forecast*() methods under .predict() method
• Generalization of .train() methods under .fit() method
• Support for distributed fit and forecast using dispy tool
• Improvements on IFTS, PWFTS and HOFTS methods.
• Better suport for probabilistic forecasting
• Jupyter notebooks with usage examples
• Sample datasets (Enrollments, AirPassengers, Sunspots, TAIEX, S&P 500, NASDAQ, SONDA, INMET, BEST)
• Tested with Python 3.6

Export for DOI generation at Zenobo tool