Distributions
Every row links to a page containing the matched HOBBS, Stan, and JAGS model definitions from the package test suite, plus the complete R test file.
| Distribution | HOBBS sampling statement | Source |
|---|---|---|
| Normal | x(i) ~ dnorm(mean, sd); |
test source |
| Standard normal | x(i) ~ normal01(); |
test source |
| Normal with SD 1 | x(i) ~ normal_sd1(mean); |
test source |
| Uniform | x(i) ~ dunif(min, max); |
test source |
| Exponential | x(i) ~ dexp(rate); |
test source |
| Gamma | x(i) ~ dgamma(shape, rate); |
test source |
| Inverse gamma | x(i) ~ dinvgamma(shape, rate); |
test source |
| Beta | x(i) ~ dbeta(a, b); |
test source |
| Cauchy | x(i) ~ dcauchy(location, scale); |
test source |
| Student t | x(i) ~ dt(df, location, scale); |
test source |
| Chi-square | x(i) ~ dchisq(df); |
test source |
| Lognormal | x(i) ~ dlnorm(meanlog, sdlog); |
test source |
| Logistic | x(i) ~ dlogis(location, scale); |
test source |
| Laplace | x(i) ~ dlaplace(location, scale); |
test source |
| Weibull | x(i) ~ dweibull(shape, scale); |
test source |
| Pareto | x(i) ~ dpareto(xmin, alpha); |
test source |
| Half-normal | x(i) ~ dhalfnorm(sd); |
test source |
| Half-Cauchy | x(i) ~ dhalfcauchy(scale); |
test source |
| Bernoulli | y(i) ~ dbern(prob); |
test source |
| Bernoulli logit | y(i) ~ bernoulli_logit(eta); |
test source |
| Bernoulli probit | y(i) ~ bernoulli_probit(eta); |
test source |
| Bernoulli cloglog | y(i) ~ bernoulli_cloglog(eta); |
test source |
| Binomial | y(i) ~ dbinom(size, prob); |
test source |
| Binomial logit | y(i) ~ binomial_logit(size, eta); |
test source |
| Poisson | y(i) ~ dpois(lambda); |
test source |
| Poisson log | y(i) ~ poisson_log(eta); |
test source |
| Negative binomial | y(i) ~ dnbinom(size, prob); |
test source |
| Negative binomial log-mean | y(i) ~ dnbinom_log(eta, size); |
test source |
| Bivariate normal | v(1:2) ~ dbvn(mean2, Sigma2); |
test source |
| Multivariate normal | x(1:k) ~ dmvn(mean, Sigma); |
test source |
| Wishart | W(1:k2) ~ dwish(scale, df, k); |
test source |
| Inverse Wishart | W(1:k2) ~ dinvwish(scale, df, k); |
test source |
| LKJ correlation, 2D | R(1:4) ~ dlkjcorr2(eta); |
test source |
Run all distribution tests
From the HOBBS package root:
devtools::test(filter = "dist-")The reference-comparison tests are skipped when rstan or rjags is unavailable. The HOBBS integration tests also require Cargo, Rust, and a C compiler.