Input and output variables of a Stan program

The $variables() method of a CmdStanModel object returns a list, each element representing a Stan model block: data, parameters, transformed_parameters and generated_quantities. The information describes the program the executable was built from, captured when the model object was created.

The standalone function variables_stan_file() returns the same list for a Stan program as it is now, without creating a model object or compiling.

Each element in the returned object contains a list of variables, with each variable represented as a list with information on its scalar type (real, int or complex) and number of dimensions. A tuple's type is instead a list with one such entry per tuple element, nested for nested tuples, and its dimensions count only the array dimensions around the tuple.

The number of dimensions reported is the number of indexing dimensions in the declared Stan variable, equivalently the number of indices needed to access one scalar element. This means a scalar has 0 dimensions, a vector or one-dimensional array has 1, and a matrix or two-dimensional array has 2. Array dimensions are added to any vector or matrix dimensions, so array[J] matrix[N, K] has 3 dimensions. See Examples.

transformed data is not included, as variables in that block are not part of the model's input or output.

Usage

variables()

variables_stan_file(stan_file, include_paths = NULL)

Arguments

stan_file

(string) The path to a Stan program.

include_paths

(character vector) Paths to directories where Stan should look for files specified in #include directives in the Stan program. Relative paths are resolved against the working directory when the model object is created and stored as absolute paths, so subsequent changes to the working directory do not affect them. When the program contains #include directives and no paths are given, the program's own directory is used.

Value

A list with information on input and output variables for each of the Stan model blocks.

See also

write_stan_json() for writing data for CmdStan.

Other CmdStanModel methods: model-method-build_info, model-method-check_syntax, model-method-cmdstan_defaults, model-method-diagnose, model-method-expose_functions, model-method-format, model-method-generate-quantities, model-method-is_current, model-method-laplace, model-method-model-info, model-method-optimize, model-method-pathfinder, model-method-sample, model-method-sample_mpi, model-method-variational

Examples

# \dontrun{
stan_file <- write_stan_file("
data {
  int N;
  array[2, 3] int y;
}
parameters {
  real alpha;
  vector[N] beta;
  array[2] matrix[3, 4] theta;
}
")

mod <- cmdstan_model(stan_file)

vars <- mod$variables()
str(vars)
#> List of 5
#>  $ parameters            :List of 3
#>   ..$ alpha:List of 2
#>   .. ..$ type      : chr "real"
#>   .. ..$ dimensions: int 0
#>   ..$ beta :List of 2
#>   .. ..$ type      : chr "real"
#>   .. ..$ dimensions: int 1
#>   ..$ theta:List of 2
#>   .. ..$ type      : chr "real"
#>   .. ..$ dimensions: int 3
#>  $ included_files        : list()
#>  $ data                  :List of 2
#>   ..$ N:List of 2
#>   .. ..$ type      : chr "int"
#>   .. ..$ dimensions: int 0
#>   ..$ y:List of 2
#>   .. ..$ type      : chr "int"
#>   .. ..$ dimensions: int 2
#>  $ transformed_parameters: Named list()
#>  $ generated_quantities  : Named list()
# }