shinystan (version 2.1.0)

as.shinystan: Create and test shinystan objects

Description

Create and test shinystan objects

Usage

as.shinystan(X, ...)
is.shinystan(object)

Value

For as.shinystan an object of class shinystan that can be used with launch_shinystan. For is.shinystan a logical value indicating whether the tested object is a shinystan object.

Details

If X is a stanfit object then no additional arguments should be specified in ... (they are taken automatically from the stanfit object). If X is a stanreg object the argument ppd (logical) can be specified indicating whether to draw from the posterior predictive distribution before launching ShinyStan. The default is TRUE, although for large objects it can be wise to set it to FALSE as drawing from the posterior predictive distribution can be time consuming. If X is not a stanfit or stanreg object then the following arguments can be specified but are not required: [object Object],[object Object],[object Object],[object Object]

See Also

launch_shinystan, launch_shinystan_demo

Examples

#################
### Example 1 ###
#################

# If X is a mcmc.list, 3D array or list of 2D chains then just do:
X_sso <- as.shinystan(X, ...) # replace ... with optional arguments or omit it

# You can also do the above if X is a stanfit object although it is not
# necessary since launch_shinystan accepts stanfit objects. 


##############################################
### Example 2: if X is a list of 2D chains ###
##############################################

# Generate some fake data
chain1 <- cbind(beta1 = rnorm(100), beta2 = rnorm(100), sigma = rexp(100))
chain2 <- cbind(beta1 = rnorm(100), beta2 = rnorm(100), sigma = rexp(100))
X <- list(chain1, chain2)
X_sso <- as.shinystan(X)

# We can also specify some or all of the optional arguments
# note: in order to use param_dims we need to rename 'beta1' and 'beta2'
# to 'beta[1]' and 'beta[2]'
colnames(chain1) <- colnames(chain2) <- c(paste0("beta[",1:2,"]"), "sigma")
X_sso <- as.shinystan(X, param_dims = list(beta = 2, sigma = 0),
                         model_name = "Example",
                         burnin = 0)
launch_shinystan(my_shinystan)