fitodeMCMC {fitode}

Fit ordinary differential equations model using MCMC

Description

This function fits ordinary differential equations models to a uni- or multi-variate time series by MCMC using the Metropolis-Hastings update rule. It searches through the parameter space on link scales, which can provide more efficient posterior sampling.

Usage

fitodeMCMC(
  model,
  data,
  start,
  tcol = "times",
  proposal.vcov,
  prior = list(),
  chains = 1,
  iter = 2000,
  burnin = iter/2,
  thin = 1,
  refresh = max(iter/10, 1),
  prior.only = FALSE,
  link,
  fixed = list(),
  solver.opts = list(method = "rk4"),
  solver = ode,
  ...
)

Arguments

model

ode model

data

data frame with time column and observation column

start

named vector of starting parameter values

tcol

time column

proposal.vcov

variance-covariance matrix of a multivariate normal proposal distribution

prior

list of formulas specifying prior distributions

chains

(numeric) number of chains

iter

(numeric) number of iterations per chain

burnin

(numeric) number of burnin interations

thin

(numeric) thining interval between consecutive observations

refresh

(numeric) refresh interval

prior.only

(logical) sample from prior distribution only?

link

named vector or list of link functions for model parameters

fixed

named vector or list of model parameters to fix and their values

solver.opts

options for ode integration. See ode

solver

ode solver

...

additional arguments (unused)

Value

An object of class “fitodeMCMC” as described in fitodeMCMC-class.