Home

Summary

Stochastic calculus and univariate and multivariate stochastic processes/Markov processes in continuous time.

The key objects introduced are the abstract type ContinuousTimeProcess{T} parametrised by the state space of the path, for example T == Float64 and various structs suptyping it, for example Wiener{Float64} for a real Brownian motion. These play roughly a similar role as types subtyping Distribution in the Distributions.jl package.

Secondly, the struct

struct SamplePath{T}
    tt::Vector{Float64}
    yy::Vector{T}
end

serves as container for sample path returned by direct and approximate samplers (sample, euler, ...). tt is the vector of the grid points of the simulation and yy the corresponding vector of states.

Help is available at the REPL:

help?> Bridge.ContinuousTimeProcess
  ContinuousTimeProcess{T}

  Types inheriting from the abstract type ContinuousTimeProcess{T}
  characterize the properties of a T-valued stochastic process, play a similar
  role as distribution types like Exponential in the package Distributions.

Pre-defined processes defined are Wiener, WienerBridge, Gamma, LinPro (linear diffusion/generalized Ornstein-Uhlenbeck) and others.

Features

  • Define and simulate diffusion processes in one or more dimension
  • Continuous and discrete likelihood using Girsanovs theorem and transition densities
  • Monte Carlo sample diffusion bridges, diffusion processes conditioned to hit a point v at a prescribed time T
  • Brownian motion in one and more dimensions
  • Ornstein-Uhlenbeck processes
  • Bessel processes
  • Gamma processes
  • Basic stochastic calculus functionality (Ito integral, quadratic variation)
  • Euler-Scheme and implicit methods (Runge-Kutta)

The layout/api was originally written to be compatible with Simon Danisch's package FixedSizeArrays.jl. It was refactored to be compatible with StaticArrays.jl by Dan Getz.

The example programs in the example/directory have additional dependencies: ConjugatePriors and a plotting library.