jameshay218/mcmcJH: Generic random walk MCMC algorithm for parameter inference
Version 0.0.0.9000

The mcmcJH package contains a set of functions that allow the user to find the posterior distribution of any generic model-generating parameter set. For example, the package was developed with the intention of finding the best set of parameters and corresponding credible intervals of a mathematical model that describes boosting and waning of adaptive immunity. The user provides a model generating function (ie. takes a set of parameters and time points and returns a matrix of the resulting trajectory), a .csv file containing information on the parameters to be fitted (eg. priors), and a .csv file containing parameters for the MCMC algorithm itself (eg. number of iterations). A Metropolis-within-Gibbs algorithm (I think...) is then used to explore the multivariate posterior; returning to the user MCMC density and iteration plots, as well as the MCMC chains themselves if necessary.

Getting started

Package details

Maintainer
LicenseNone whatsoever - good luck!
Version0.0.0.9000
Package repositoryView on GitHub
Installation Install the latest version of this package by entering the following in R:
install.packages("devtools")
library(devtools)
install_github("jameshay218/mcmcJH")
jameshay218/mcmcJH documentation built on May 17, 2017, 9:38 a.m.