Introduce a new family of transition models that generalise transition models driven by white noise to the Levy process.
Provides support for normal sigma-mean and normal variance-mean mixture processes.
Users can inject the desired driving processes into the selected transition model
Introduce marginalised particle filtering for inferencing the newly-introduced class of models.
Things to do:
Update tutorial notebook for the new class of models and marginalised particle filter
Writing unit tests, still figuring out how to test every possible combination of the driving process and transition model pair without bloat, suggestions are welcome.
Implementation of PDF and log PDF functions, certain Levy distributions have non-trivial PDFs/CDFs, currently looking into inverse FFT methods involving characteristic functions.
Provide support for more models, e.g. Singer
Additional comments:
I believe the current design is a good compromise between speed and compatibility with the existing stone soup framework to avoid introducing any breaking changes. The newly introduced models decouple the deterministic (transition model) and non-deterministic (noise driver) components. That being said, suggestions are welcome and highly appreciated.
@hpritchett-dstl
Still W.I.P.
Features:
Things to do:
Additional comments:
I believe the current design is a good compromise between speed and compatibility with the existing stone soup framework to avoid introducing any breaking changes. The newly introduced models decouple the deterministic (transition model) and non-deterministic (noise driver) components. That being said, suggestions are welcome and highly appreciated.