SMC: SMC sampler

In sgolchi/BPD: Bayesian Particle Discovery

Description

SMC sampler

Usage

SMC(N = 1000, G, y, cnew, snew, p0 = 0.5, mu_var = 0.5,
  back_mean = "B polynomial")

Arguments

N	Monte Carlo sample size
G	grid over mass range
y	vector of observations
cnew	matrix of signal normalizing constants - dimensions are (size of grid G)by(number of categories/channels,41)
snew	matrix of signal widths, same dimensions as cnew
p0	prior probability of mass(Higgs)=0
mu_var	prior variance of signal strength parameter
back_mean	parametric form used to specify the mean of background logGP prior

Value

A list of sequence of posterior samples for model parameters, sequence of acceptance rates for all parameters (for monitoring), sequence of effective sample sizes (ESS) and sequence of temperatures

Examples

load('data/Higgs_data')
G = H_data$grid
y = H_data$y
cnew = H_data$cnew
snew = H_data$snew
post = SMC(N = 1e03, G = G, y = y, cnew = cnew, snew = snew, p0 = .5, mu_var = .5, back_mean = 'B polynomial')
save(post, file = 'data/SMC_full_poly')

sgolchi/BPD documentation built on May 19, 2019, 3:05 a.m.