Sample from the target distribution using MCMC

Usage

run_mcmc(
  data,
  is_missing = FALSE,
  allow_relatedness = TRUE,
  thin = 1,
  burnin = 10000,
  samples_per_chain = 1000,
  verbose = TRUE,
  use_message = FALSE,
  eps_pos_alpha = 1,
  eps_pos_beta = 1,
  eps_neg_alpha = 1,
  eps_neg_beta = 1,
  r_alpha = 1,
  r_beta = 1,
  mean_coi_shape = 0.1,
  mean_coi_scale = 10,
  max_eps_pos = 2,
  max_eps_neg = 2,
  max_coi = 40,
  num_chains = 1,
  num_cores = 1,
  pt_chains = 1,
  pt_grad = 1,
  pt_num_threads = 1,
  adapt_temp = TRUE,
  pre_adapt_steps = 25,
  temp_adapt_steps = 25,
  max_initialization_tries = 10000
)

Arguments

data: Data to be used in MCMC, as generated by the load_*_data functions
is_missing: Boolean matrix indicating whether the observation should be treated as missing data and ignored. Number of rows equals the number of loci, number of columns equals the number samples. Alternatively, the user may pass in FALSE if no data should be considered missing.
allow_relatedness: Bool indicating whether or not to allow relatedness within host
thin: Positive Integer. How often to sample from mcmc, 1 means do not thin
burnin: Positive Integer. Number of MCMC samples to discard as burnin
samples_per_chain: Positive Integer. Number of samples to take after burnin
verbose: Logical indicating if progress is printed
use_message: Logical indicating if progress is printed using message or print
eps_pos_alpha: Positive Numeric. Alpha parameter in Beta distribution for eps_pos prior
eps_pos_beta: Positive Numeric. Beta parameter in Beta distribution for eps_pos prior
eps_neg_alpha: Positive Numeric. Alpha parameter in Beta distribution for eps_neg prior
eps_neg_beta: Positive Numeric. Beta parameter in Beta distribution for eps_neg prior
r_alpha: Positive Numeric. Alpha parameter in Beta distribution for relatedness prior
r_beta: Positive Numeric. Beta parameter in Beta distribution for relatedness prior
mean_coi_shape: shape parameter for gamma hyperprior on mean COI
mean_coi_scale: scale parameter for gamma hyperprior on mean COI
max_eps_pos: Numeric. Maximum allowed value for eps_pos
max_eps_neg: Numeric. Maximum allowed value for eps_neg
max_coi: Positive Numeric. Maximum allowed complexity of infection
num_chains: Total number of chains to run, possibly simultaneously
num_cores: Total OMP parallel threads to use to run chains. num_cores * pt_num_threads should not exceed the number of cores available on your system.
pt_chains: Total number of chains to run with parallel tempering or a vector containing the temperatures that should be used for parallel tempering.
pt_grad: Power to raise parallel tempering chains to. A value of 1 results in evenly distributed temperatures between [0,1], below 1 will bias towards 1 and above 1 will bias towards 0. Only used if pt_chains is a single value (i.e. not a vector).
pt_num_threads: Total number of OMP parallel threads to be used to process parallel tempered chains num_cores * pt_num_threads should not exceed the number of cores available on your system.
adapt_temp: Logical indicating whether or not to adapt the parallel tempering temperatures. If TRUE, the temperatures will be adapted during the burnin period, starting after pre_adapt_steps steps. The adaptation will occur every temp_adapt_steps steps until burnin is complete. The range of temperatures will remain the same as specified by pt_chains.
pre_adapt_steps: Number of steps to take before starting to adapt the parallel tempering temperatures. Only used if adapt_temp is TRUE.
temp_adapt_steps: Number of steps to take between temperature adaptation steps. Only used if adapt_temp is TRUE.
max_initialization_tries: Number of times to try to initialize the chain before giving up