compute_distance_metropolis_to_condbern: The Metropolis sampler for Conditional Bernoulli distribution

In nianqiaoju/agents: Toolkit for inference in agent-based modelss

Description

compute an upper bound for the 2-Wasserstein distance between Metropolis sampler in metropolis_condbern.R and the Conditional-Bernoulli distribution

Usage

compute_distance_metropolis_to_condbern(
  alpha,
  sum_x,
  num_mcmc,
  num_repeat,
  distance = "tv"
)

Arguments

alpha	vector of probabilities for the Poisson-Binomial distribution
sum_x	sum of all entries
num_mcmc	number of MCMC iterations to run before coupling
num_repeat	number of repetitions to get an estimate of the upper bound
distance	can be 'tv' or 'wasserstein'

Value

a vector of distance upper bounds for different number of iterations

nianqiaoju/agents documentation built on Feb. 19, 2021, 12:18 a.m.