Movatterモバイル変換

Bayesianinference for state-space models with R

rbi is anRinterface tolibbi, a library forBayesian inference.

It mainly contains:

• various functions to retrieve and process the results from libbi(which are in NetCDF format)
• abi_model class, to manipulate libbi models
• alibbi wrapper class, to perform Bayesian using libbiinference from within R,

Installation

The easiest way to install the latest stable version ofrbi is viaCRAN:

install.packages("rbi")

Alternatively, the current development version can be installed usingtheremotes package

# install.packages("remotes")library("remotes")install_github("sbfnk/rbi")

Therbi package has only been tested on GNU/Linuxand OS X, but it should mostly work everywhereR works.

If you want to userbi as a wrapper toLibBi then you need a working version ofLibBi. To installLibBi on a Mac, the easiest way is to installHomebrew, followed by (using a command shell,i.e. Terminal or similar):

brew install libbi

On linux, follow theinstructionsprovided with LibBi.

The path tolibbi script can be passed as an argument torbi, otherwise the package tries to find itautomatically using thewhich linux/unix command.

If you just want to process the output fromLibBi,then you do not need to haveLibBi installed.

Getting started

A good starting point is to look at the included demos:

demo(PZ_generate_dataset)## generating a data set from a modeldemo(PZ_PMMH)## particle Markov-chain Metropolis-Hastingsdemo(PZ_SMC2)## SMC^2demo(PZ_filtering)## filtering

For further information, have a look at the introductory vignettefrom the link from therbi CRAN package.

Using coda

LibBi contains theget_traces methodwhich provides an interface tocoda.

Other packages

For higher-level methods to interact withLibBi, have a look atrbi.helpers.