Skip to contents

Find a good Step-Size for a given MCMC Algorithm

tune_step_size.Rd

Given a closure MCMC(p,N,eps), where p is the initial Markov-chain position, N a sample-size, and eps a step-size, this function finds a good value for eps.

Usage

tune_step_size(
  MCMC,
  parMCMC = attr(MCMC, "init"),
  target_acceptance = 0.25,
  iter.max = 6,
  h = 1e-04
)

Arguments

MCMC

a Markov chain Monte Carlo closure (function)

parMCMC

initial position of the Markov chain, has to be initialized with mcmc_init.

target_acceptance

a scalar value for the desired acceptance rate, some algorithms are most efficient with 20% to 30% acceptance, some work well with a very high acceptance.

iter.max

maximum number of iterations until the function has to return.

h

initial guess for the MCMC step size

Value

optimal step size

Details

It will take 100 sample points repeatedly, until an acceptance of target_acceptance is reached (defaults to 25%). The step-size is decreased if acceptance is very low and increased when it is too high.

This function will do at most

Examples

# \donttest{
m <- model_from_tsv(uqsa_example("AKAP79"))
rwm <- high_level_metropolis(m) # "random walk", metropolis algorithm
#> The parameters are given in log10-scale, so the simulator will do the reverse transformation: 10^p.
p <- rwm %@% "init"             # a valid starting point
h <- tune_step_size(rwm,p)
#> acceptance rate: 0.12, step-size: 0.0001;
#> acceptance rate: 0.24, step-size: 3.74512e-05;
N <- 200
smallSample <- rwm(rwm %@% "init",N,h)
print(h)
#> [1] 3.745124e-05
plot(
  smallSample %@% "logLikelihood",
  type="l",
  main=sprintf("step size: %g",h),
  xlab="iterations",
  ylab="log-likelihood"
)

# }