fusion_TA_exp_4: Time-adapting Fusion

In rchan26/exp4Tempering: Hierarchical and Sequential Monte Carlo Fusion for exp(-(x^4)/2)

Description

Time-adapting Monte Carlo Fusion for sub-posteriors of the form exp(-((x^4)*beta)/2)

Usage

fusion_TA_exp_4(N, mean = 0, time, C, samples_to_fuse, betas,
  sample_weights, level = 1, acceptance_rate = F, timed = F)

Arguments

N	number of samples
mean	mean value
time	time T for time-adapting fusion algorithm
C	number of sub-posteriors to combine
samples_to_fuse	list of length C, where samples_to_fuse[c] containg the samples for the c-th sub-posterior
betas	vector of length C, where betas[c] is the beta for c-th sub-posterior
sample_weights	defaults to sqrt(betas) if missing: vector of length C, where sample_weights[c] is the sample wieght for the c-th sub-posterior
level	defaults to 1, used in hierarchical and sequential Monte Carlo Fusion
acceptance_rate	boolean value: defaults to F, determines whether or not to return acceptance rates
timed	boolen value: defaults to T, determines whether or not to return the time elapsed to run

Value

samples: fusion samples

iters_rho: number of iterations from the first accept/reject step (rho)

iters_Q: number of iterations from the second (diffusion) accept/reject step (Q)

time: run-time of fusion sampler

Examples

input_samples <- base_rejection_sampler_exp_4(beta = 1/2, nsamples = 100000, proposal_mean = 0, proposal_sd = 1, dominating_M = 1.3)
test1_standard <- fusion_diff_exp_4(N = 10000, time = 1, C = 2, samples_to_fuse = input_samples, betas = rep(1/2, 2), timed = T)

rchan26/exp4Tempering documentation built on June 20, 2019, 10:07 p.m.