In goldingn/greta: Simple and Scalable Statistical Modelling in R

Beetles

Beetles considers dose-response data from an experiment applying carbon disulphide to 8 beetles. The original example compares three different link functions; the logit, probit and complementary log-log. Here, only the code for the logit link is shown. You can implement the other two link functions in greta by changing ilogit to iprobit or icloglog.

See WinBUGS examples volume 2 (pdf) for details.

data

x <- c(1.6907, 1.7242, 1.7552, 1.7842, 1.8113, 1.8369, 1.8610, 1.8839)
n <- c(59, 60, 62, 56, 63, 59, 62, 60)
r <- c(6, 13, 18, 28, 52, 53, 61, 60)
N <- 8

greta code

alpha_star <- normal(0, 32)
beta <- normal(0, 32)
p <- ilogit(alpha_star + beta * (x - mean(x)))
distribution(r) <- binomial(n, p)

alpha <- alpha_star - beta * mean(x)
rhat <- p * n

BUGS/JAGS code

for( i in 1 : N ) {
  r[i] ~ dbin(p[i],n[i])
  logit(p[i]) <- alpha.star + beta * (x[i] - mean(x[]))
  rhat[i] <- n[i] * p[i]
  culmative.r[i] <- culmative(r[i], r[i])
}
alpha <- alpha.star - beta * mean(x[])
beta ~ dnorm(0.0,0.001)
alpha.star ~ dnorm(0.0,0.001)

Stan code

cat(readLines('https://raw.githubusercontent.com/stan-dev/example-models/master/bugs_examples/vol2/beetles/beetles_logit.stan'), sep = '\n')

goldingn/greta documentation built on May 24, 2021, 11 a.m.