Release v0.7.0
Pre-releaseThis is the release accompanying the article:
F. v. d. Meulen, M. Schauer: Continuous-discrete smoothing of diffusions, 2017. arxiv:1712.03807.
The article describes a novel smoothing algorithm for discretely observed diffusions. From the abstract:
Suppose X is a multivariate diffusion process that is observed discretely in time. At each observation time, a linear transformation of the state of the process is observed with additive noise. The smoothing problem consists of recovering the path of the process, consistent with the observations. We derive a novel Markov Chain Monte Carlo algorithm to sample from the exact smoothing distribution. Key to this is an extension of the linear guided proposals introduced in Schauer et. al (2017). We illustrate the efficiciency of our method on both the Lorenz system and a partially observed integrated diffusion model.
The code supplementing the article is in the folder supplements/smoothing
. It makes heavy use of the new GuidedBridge
, which provides a new unifying interface proposal processes for conditional diffusion processes under different observation scheme and for different auxilliary processes.
Further changes
- Refactored
Gaussian
- Updated to
StaticArrays 0.6.3
- Notebook illustrating how to do parameter inference for multivariate stochastic differential equations with this package with nice 3d-illustrations: https://github.com/mschauer/Bridge.jl/blob/master/example/3DExamples.ipynb
- Plot recipes for
Plots
(and convenience functions forMakie
in the folderextra
) - Thinning sampler for inhomogenous Poisson process
- Interoperability changes to allow for calling solvers in Bridge from within the
JuliaDiffEq
common interface viaBridgeDiffEq.jl
(https://github.com/JuliaDiffEq/BridgeDiffEq.jl) - Exact sampling of Ornstein-Uhlenbeck processes
- Add an example for Levy driven SDEs.