run_mcmc: Posterior sampling using differential evolution
In jeff324/derp: Differential Evolution in R

powder Runs power posterior sampling using differential evolution markov chain monte carlo

run_mcmc(model, pars = NULL, data, sampler = "de", num_samples = NULL,
  num_chains = NULL, migration_start = NULL, migration_end = NULL,
  migration_freq = NULL, randomize_phi = TRUE, update = 100,
  init_theta = NULL, init_phi = NULL, return_as_mcmc = TRUE,
  parallel_backend = "none", n_cores = NULL, benchmark = FALSE)

`model`	A `function`. Defines a Bayesian model. See example below.
`pars`	A `list`. Specifies the names of parameters along with their initialization bounds and blocks. See example below.
`data`	A `list`. Specifies the data to be modeled. See example.
`sampler`	A `character`. Specifies the name of the sampler to be used. `de` is the only built-in sampler.
`num_samples`	A `numeric`. The total number of samples to collect.
`num_chains`	A `numeric`. The number of chains to run.
`migration_freq`	A `numeric`. Number of iterations to wait between each migration step. If `NULL`, migration is not done.
`randomize_phi`	A `logical`. Should the correlational structure between level-1 and level-2 parameters be ignored.
`update`	A `numeric`. Specifies the number of iterations before printing the current iteration number to the console.
`init_theta`	A list where each element contains a named vector of parameter initial start values for each subject.
`init_phi`	A named vector parameter initial start values
`parallel_backend`	A character vector either 'MPI', 'doParallel', or 'none' indicating backend for parallelization. Default is none.
`n_cores`	A `numeric`. Number of cores when parallel_backend is specified.
`benchmark`	A `logical`. Produces compute times benchmark purposes. Oftentimes, ignoring this correlation will lead to better sampling.
`migration.start`	A `numeric`. Iteration to start migrating. This should be after chains are burned in. If `NULL`, migration is not done.
`migration.end`	A `numeric`. Iteration to stop migrating. Migration should stop well before sampling is finished. If `NULL`, migration is not done.

## Not run: 
library(derp)
#### Simulate data
data = list(response = rnorm(100,2,1))
#### Model parameters
pars = list(
'mu' =    list('init'=c(1,10)),
'sd' =    list('init'=c(.1,5))
)
#### Define model
model = function() {
    normal_lpdf(mu, 0, 3)
    gamma_lpdf(sd, 1, 1)
    normal_lpdf(response, mu, sd)
}
#### Run the sampler
samples = run_mcmc(model, pars, data,
                  migration_start = 500, migration_end = 700,
                  migration_freq = 10, num_samples=3000)
mcmc_snip = snip(samples,burnin=1000,thin=1)
summary(mcmc_snip)

## End(Not run)