Purpose

This package provides some minimal working examples for implementing the particle Metropolis-Hastings (PMH) algorithm for parameter inference in nonlinear state space models. The package accompanies a tutorial:

Dahlin, J. & Schön, T. B. "Getting started with particle Metropolis-Hastings for inference in nonlinear dynamical models." pre-print, arXiv:1511.01707, 2015.

Currently available at:

http://arxiv.org/pdf/1511.01707

Usage

The main functions of the package are the five examples connected to the paper:

• ** example1_lgss() ** Demostrates the particle filter for estimating the

state in a linear Gaussian state space model.

• ** example2_lgss() ** Demostrates PMH for estimating the parameters in a

linear Gaussian state space model.

• ** example3_sv() ** Demostrates PMH for estimating the parameters in a

stochastic volatility model.

• ** example4_sv() ** Demostrates PMH for estimating the parameters in a

stochastic volatility model using a tailored proposal distribution.

• ** example5_sv() ** Demostrates PMH for estimating the parameters in a

stochastic volatility model using a reparameterised model.

Simple example

The examples can be executed by e.g.

example2_lgss()

which will recreate on of the plots in the aforementioned tutorial. See the reference for more background and details.

How do I get it?

The package is available through CRAN and can be installed by

install.packages("pmhtutorial")