RStan — R interface to Stan

Description

RStan is the R interface to the Stan C++ package. RStan provides

• full Bayesian inference using the No-U-Turn sampler (NUTS), a variant of Hamiltonian Monte Carlo (HMC)

• approximate Bayesian inference using automatic differentiation variational inference (ADVI)

• penalized maximum likelihood estimation using L-BFGS optimization

For more information about Stan visit http://mc-stan.org/.

Details

For more information on Stan and its modeling language, see the Stan Modeling Language User's Guide and Reference Manual available at http://mc-stan.org/.

Author(s)

References

Stan Development Team Stan Modeling Language User's Guide and Reference Manual. http://mc-stan.org/.

See Also

The stan function for details on fitting models and stanfit for information on the fitted model objects.

Several related R packages are also available from the Stan Development Team: loo (loo-package) offers model comparison on estimated out-of-sample predictive performance, shinystan (shinystan-package) provides the ShinyStan GUI for exploring fitted Bayesian models, and rstanarm is an appendage to rstan providing an R formula interface for Bayesian regression modeling.

Examples

## Not run:  

stanmodelcode <- "
data {
  int<lower=0> N;
  real y[N];
} 

parameters {
  real mu;
} 

model {
  target += normal_lpdf(mu | 0, 10);
  target += normal_lpdf(y  | mu, 1);
} 
"

y <- rnorm(20) 
dat <- list(N = 20, y = y); 
fit <- stan(model_code = stanmodelcode, model_name = "example", 
            data = dat, iter = 2012, chains = 3, sample_file = 'norm.csv',
            verbose = TRUE) 
print(fit)
traceplot(fit)

# extract samples 
e <- extract(fit, permuted = TRUE) # return a list of arrays 
mu <- e$mu 

m <- extract(fit, permuted = FALSE, inc_warmup = FALSE) # return an array 
print(dimnames(m))

# using as.array directly on stanfit objects 
m2 <- as.array(fit)


## End(Not run)