gelman.diag.dmc: Gelman and Rubin Convergence Diagnostic

In TasCL/ggdmc: Dynamic Model of Choice with Parallel Computation, and C++ Capabilities

Description

gelman.diag.dmc calls coda gelman.diag to get R hats for one or a list of subjects. It can calculate at the data or hyper level.

Usage

gelman.diag.dmc(x, hyper = FALSE, digits = 2, start = 1,
  autoburnin = FALSE, transform = TRUE, end = NA, ...)

Arguments

x	a DMC sample
hyper	a switch to extract hyper attribute and calculate it
digits	print out how many digits
start	start iteration
autoburnin	turn on auto burnin
transform	turn on transform
end	end iteraton
...	arguments passing to coda gelman.diag.

Examples

m1 <- model.dmc(
     p.map     = list(a="1",v="F",z="1",d="1",sz="1",sv="1",t0="1",st0="1"),
     match.map = list(M=list(s1="r1",s2="r2")),
     factors   = list(S=c("s1","s2"),F=c("f1","f2")),
     constants = c(st0=0,d=0),
     responses = c("r1","r2"),
     type      = "rd")

pop.mean  <- c(a=1.15, v.f1=1.25, v.f2=1.85, z=.55,  sz=.15, sv=.32, t0=.25)
pop.scale <- c(a=.10,  v.f1=.8,   v.f2=.5,   z=0.1,  sz=.05, sv=.05, t0=.05)
pop.prior <- prior.p.dmc(
  dists = rep("tnorm", length(pop.mean)),
  p1    = pop.mean,
  p2    = pop.scale,
  lower = c(0,-5, -5, 0, 0,   0, 0),
  upper = c(5, 7,  7, 1, 0.5, 2, 2))

dat  <- h.simulate.dmc(m1, nsim=30, ns=8, p.prior=pop.prior)
mdi1 <- data.model.dmc(dat, m1)
ps   <- attr(dat,  "parameters")
### FIT RANDOM EFFECTS
p.prior <- prior.p.dmc(
  dists = c("tnorm","tnorm","tnorm","tnorm","tnorm", "tnorm", "tnorm"),
  p1=pop.mean,
  p2=pop.scale*5,
  lower=c(0,-5, -5, 0, 0, 0, 0),
  upper=c(5, 7,  7, 2, 2, 2, 2))

mu.prior <- prior.p.dmc(
  dists = c("tnorm","tnorm","tnorm","tnorm","tnorm", "tnorm", "tnorm"),
  p1=pop.mean,
  p2=pop.scale*5,
  lower=c(0,-5, -5, 0, 0, 0, 0),
  upper=c(5, 7,  7, 2, 2, 2, 2))

sigma.prior <- prior.p.dmc(
  dists = rep("beta", length(p.prior)),
  p1=c(a=1, v.f1=1,v.f2 = 1, z=1, sz=1, sv=1, t0=1),p2=c(1,1,1,1,1,1,1),
  upper=c(2,2,2,2,2, 2, 2))

pp.prior <- list(mu.prior, sigma.prior)

hsamples0 <- h.samples.dmc(nmc=10, p.prior=p.prior, pp.prior=pp.prior,
  data=mdi1, thin=1)
hsamples0 <- h.run.dmc(hsamples0)
gelman.diag.dmc(hsamples0, hyper=TRUE)
gelman.diag.dmc(hsamples0)

TasCL/ggdmc documentation built on May 9, 2019, 4:19 p.m.