add_samples_adaptive: Add samples adaptively

In NErler/simvalidator: Validation of Statistical Models via Simulation

Add samples adaptively

Description

Add samples adaptively

Usage

add_samples_adaptive(
  fitted_model,
  extra_iter = NULL,
  minsize = 500L,
  step = 200L,
  subset = NULL,
  cutoff = 1.2,
  prop = 0.8,
  gr_max = 1.5,
  max_try = 5L
)

Arguments

fitted_model	a object of class 'JointAI'
extra_iter	number of iterations that should be added to the model if the Gelman-Rubin criterion is too large
minsize	the minimum number of iterations to be considered
step	the step size in which iterations are omitted as burn-in
subset	subset of parameters on which the Gelman-Rubin criterion should be evaluated. Follows the logic used in JointAI
cutoff	the cut-off used for the Gelman Rubin criterion
prop	proportion of parameters that need to be below the cutoff
gr_max	maximum allowed value for the Gelman-Rubin criterion
max_try	maximum number of runs of JointAI::add_samples()

NErler/simvalidator documentation built on May 17, 2022, 7:54 a.m.