Skip to contents

tar_stan_mcmc() creates targets to run one MCMC per model and separately save summaries draws, and diagnostics.


  data = list(),
  compile = c("original", "copy"),
  quiet = TRUE,
  stdout = NULL,
  stderr = NULL,
  dir = NULL,
  pedantic = FALSE,
  include_paths = NULL,
  cpp_options = list(),
  stanc_options = list(),
  force_recompile = FALSE,
  seed = NULL,
  refresh = NULL,
  init = NULL,
  save_latent_dynamics = FALSE,
  output_dir = NULL,
  output_basename = NULL,
  sig_figs = NULL,
  chains = 4,
  parallel_chains = getOption("mc.cores", 1),
  chain_ids = seq_len(chains),
  threads_per_chain = NULL,
  opencl_ids = NULL,
  iter_warmup = NULL,
  iter_sampling = NULL,
  save_warmup = FALSE,
  thin = NULL,
  max_treedepth = NULL,
  adapt_engaged = TRUE,
  adapt_delta = NULL,
  step_size = NULL,
  metric = NULL,
  metric_file = NULL,
  inv_metric = NULL,
  init_buffer = NULL,
  term_buffer = NULL,
  window = NULL,
  fixed_param = FALSE,
  show_messages = TRUE,
  diagnostics = c("divergences", "treedepth", "ebfmi"),
  variables = NULL,
  variables_fit = NULL,
  inc_warmup = FALSE,
  inc_warmup_fit = FALSE,
  summaries = list(),
  summary_args = list(),
  return_draws = TRUE,
  return_diagnostics = TRUE,
  return_summary = TRUE,
  draws = NULL,
  summary = NULL,
  tidy_eval = targets::tar_option_get("tidy_eval"),
  packages = targets::tar_option_get("packages"),
  library = targets::tar_option_get("library"),
  format = "qs",
  format_df = "fst_tbl",
  repository = targets::tar_option_get("repository"),
  error = targets::tar_option_get("error"),
  memory = targets::tar_option_get("memory"),
  garbage_collection = targets::tar_option_get("garbage_collection"),
  deployment = targets::tar_option_get("deployment"),
  priority = targets::tar_option_get("priority"),
  resources = targets::tar_option_get("resources"),
  storage = targets::tar_option_get("storage"),
  retrieval = targets::tar_option_get("retrieval"),
  cue = targets::tar_option_get("cue"),
  description = targets::tar_option_get("description")



Symbol, base name for the collection of targets. Serves as a prefix for target names.


Character vector of Stan model files. If you supply multiple files, each model will run on the one shared dataset generated by the code in data. If you supply an unnamed vector, fs::path_ext_remove(basename(stan_files)) will be used as target name suffixes. If stan_files is a named vector, the suffixed will come from names(stan_files).


Code to generate the data for the Stan model.


(logical) Do compilation? The default is TRUE. If FALSE compilation can be done later via the $compile() method.


(logical) Should the verbose output from CmdStan during compilation be suppressed? The default is TRUE, but if you encounter an error we recommend trying again with quiet=FALSE to see more of the output.


Character of length 1, file path to write the stdout stream of the model when it runs. Set to NULL to print to the console. Set to R.utils::nullfile() to suppress stdout. Does not apply to messages, warnings, or errors.


Character of length 1, file path to write the stderr stream of the model when it runs. Set to NULL to print to the console. Set to R.utils::nullfile() to suppress stderr. Does not apply to messages, warnings, or errors.


(string) The path to the directory in which to store the CmdStan executable (or .hpp file if using $save_hpp_file()). The default is the same location as the Stan program.


(logical) Should pedantic mode be turned on? The default is FALSE. Pedantic mode attempts to warn you about potential issues in your Stan program beyond syntax errors. For details see the Pedantic mode chapter in the Stan Reference Manual. Note: to do a pedantic check for a model without compiling it or for a model that is already compiled the $check_syntax() method can be used instead.


(character vector) Paths to directories where Stan should look for files specified in #include directives in the Stan program.


(list) Any makefile options to be used when compiling the model (STAN_THREADS, STAN_MPI, STAN_OPENCL, etc.). Anything you would otherwise write in the make/local file. For an example of using threading see the Stan case study Reduce Sum: A Minimal Example.


(list) Any Stan-to-C++ transpiler options to be used when compiling the model. See the Examples section below as well as the stanc chapter of the CmdStan Guide for more details on available options:


(logical) Should the model be recompiled even if was not modified since last compiled. The default is FALSE. Can also be set via a global cmdstanr_force_recompile option.


(positive integer(s)) A seed for the (P)RNG to pass to CmdStan. In the case of multi-chain sampling the single seed will automatically be augmented by the the run (chain) ID so that each chain uses a different seed. The exception is the transformed data block, which defaults to using same seed for all chains so that the same data is generated for all chains if RNG functions are used. The only time seed should be specified as a vector (one element per chain) is if RNG functions are used in transformed data and the goal is to generate different data for each chain.


(non-negative integer) The number of iterations between printed screen updates. If refresh = 0, only error messages will be printed.


(multiple options) The initialization method to use for the variables declared in the parameters block of the Stan program. One of the following:

  • A real number x>0. This initializes all parameters randomly between [-x,x] on the unconstrained parameter space.;

  • The number 0. This initializes all parameters to 0;

  • A character vector of paths (one per chain) to JSON or Rdump files containing initial values for all or some parameters. See write_stan_json() to write R objects to JSON files compatible with CmdStan.

  • A list of lists containing initial values for all or some parameters. For MCMC the list should contain a sublist for each chain. For other model fitting methods there should be just one sublist. The sublists should have named elements corresponding to the parameters for which you are specifying initial values. See Examples.

  • A function that returns a single list with names corresponding to the parameters for which you are specifying initial values. The function can take no arguments or a single argument chain_id. For MCMC, if the function has argument chain_id it will be supplied with the chain id (from 1 to number of chains) when called to generate the initial values. See Examples.


(logical) Should auxiliary diagnostic information about the latent dynamics be written to temporary diagnostic CSV files? This argument replaces CmdStan's diagnostic_file argument and the content written to CSV is controlled by the user's CmdStan installation and not CmdStanR (for some algorithms no content may be written). The default is FALSE, which is appropriate for almost every use case. To save the temporary files created when save_latent_dynamics=TRUE see the $save_latent_dynamics_files() method.


(string) A path to a directory where CmdStan should write its output CSV files. For interactive use this can typically be left at NULL (temporary directory) since CmdStanR makes the CmdStan output (posterior draws and diagnostics) available in R via methods of the fitted model objects. The behavior of output_dir is as follows:

  • If NULL (the default), then the CSV files are written to a temporary directory and only saved permanently if the user calls one of the $save_* methods of the fitted model object (e.g., $save_output_files()). These temporary files are removed when the fitted model object is garbage collected (manually or automatically).

  • If a path, then the files are created in output_dir with names corresponding to the defaults used by $save_output_files().


(string) A string to use as a prefix for the names of the output CSV files of CmdStan. If NULL (the default), the basename of the output CSV files will be comprised from the model name, timestamp, and 5 random characters.


(positive integer) The number of significant figures used when storing the output values. By default, CmdStan represent the output values with 6 significant figures. The upper limit for sig_figs is 18. Increasing this value will result in larger output CSV files and thus an increased usage of disk space.


(positive integer) The number of Markov chains to run. The default is 4.


(positive integer) The maximum number of MCMC chains to run in parallel. If parallel_chains is not specified then the default is to look for the option "mc.cores", which can be set for an entire R session by options(mc.cores=value). If the "mc.cores" option has not been set then the default is 1.


(integer vector) A vector of chain IDs. Must contain as many unique positive integers as the number of chains. If not set, the default chain IDs are used (integers starting from 1).


(positive integer) If the model was compiled with threading support, the number of threads to use in parallelized sections within an MCMC chain (e.g., when using the Stan functions reduce_sum() or map_rect()). This is in contrast with parallel_chains, which specifies the number of chains to run in parallel. The actual number of CPU cores used is parallel_chains*threads_per_chain. For an example of using threading see the Stan case study Reduce Sum: A Minimal Example.


(integer vector of length 2) The platform and device IDs of the OpenCL device to use for fitting. The model must be compiled with cpp_options = list(stan_opencl = TRUE) for this argument to have an effect.


(positive integer) The number of warmup iterations to run per chain. Note: in the CmdStan User's Guide this is referred to as num_warmup.


(positive integer) The number of post-warmup iterations to run per chain. Note: in the CmdStan User's Guide this is referred to as num_samples.


(logical) Should warmup iterations be saved? The default is FALSE.


(positive integer) The period between saved samples. This should typically be left at its default (no thinning) unless memory is a problem.


(positive integer) The maximum allowed tree depth for the NUTS engine. See the Tree Depth section of the CmdStan User's Guide for more details.


(logical) Do warmup adaptation? The default is TRUE. If a precomputed inverse metric is specified via the inv_metric argument (or metric_file) then, if adapt_engaged=TRUE, Stan will use the provided inverse metric just as an initial guess during adaptation. To turn off adaptation when using a precomputed inverse metric set adapt_engaged=FALSE.


(real in (0,1)) The adaptation target acceptance statistic.


(positive real) The initial step size for the discrete approximation to continuous Hamiltonian dynamics. This is further tuned during warmup.


(string) One of "diag_e", "dense_e", or "unit_e", specifying the geometry of the base manifold. See the Euclidean Metric section of the CmdStan User's Guide for more details. To specify a precomputed (inverse) metric, see the inv_metric argument below.


(character vector) The paths to JSON or Rdump files (one per chain) compatible with CmdStan that contain precomputed inverse metrics. The metric_file argument is inherited from CmdStan but is confusing in that the entry in JSON or Rdump file(s) must be named inv_metric, referring to the inverse metric. We recommend instead using CmdStanR's inv_metric argument (see below) to specify an inverse metric directly using a vector or matrix from your R session.


(vector, matrix) A vector (if metric='diag_e') or a matrix (if metric='dense_e') for initializing the inverse metric. This can be used as an alternative to the metric_file argument. A vector is interpreted as a diagonal metric. The inverse metric is usually set to an estimate of the posterior covariance. See the adapt_engaged argument above for details about (and control over) how specifying a precomputed inverse metric interacts with adaptation.


(nonnegative integer) Width of initial fast timestep adaptation interval during warmup.


(nonnegative integer) Width of final fast timestep adaptation interval during warmup.


(nonnegative integer) Initial width of slow timestep/metric adaptation interval.


(logical) When TRUE, call CmdStan with argument "algorithm=fixed_param". The default is FALSE. The fixed parameter sampler generates a new sample without changing the current state of the Markov chain; only generated quantities may change. This can be useful when, for example, trying to generate pseudo-data using the generated quantities block. If the parameters block is empty then using fixed_param=TRUE is mandatory. When fixed_param=TRUE the chains and parallel_chains arguments will be set to 1.


(logical) When TRUE (the default), prints all output during the sampling process, such as iteration numbers and elapsed times. If the output is silenced then the $output() method of the resulting fit object can be used to display the silenced messages.


(character vector) The diagnostics to automatically check and warn about after sampling. Setting this to an empty string "" or NULL can be used to prevent CmdStanR from automatically reading in the sampler diagnostics from CSV if you wish to manually read in the results and validate them yourself, for example using read_cmdstan_csv(). The currently available diagnostics are "divergences", "treedepth", and "ebfmi" (the default is to check all of them).

These diagnostics are also available after fitting. The $sampler_diagnostics() method provides access the diagnostic values for each iteration and the $diagnostic_summary() method provides summaries of the diagnostics and can regenerate the warning messages.

Diagnostics like R-hat and effective sample size are not currently available via the diagnostics argument but can be checked after fitting using the $summary() method.


(character vector) Optionally, the names of the variables (parameters, transformed parameters, and generated quantities) to read in.

  • If NULL (the default) then all variables are included.

  • If an empty string (variables="") then none are included.

  • For non-scalar variables all elements or specific elements can be selected:

    • variables = "theta" selects all elements of theta;

    • variables = c("theta[1]", "theta[3]") selects only the 1st and 3rd elements.


Character vector of variables to include in the big CmdStanFit object returned by the model fit target. The variables argument, by contrast, is for the "draws" target only. The "draws" target can only access the variables in the CmdStanFit target. Control the variables in each with the variables and variables_fit arguments.


(logical) Should warmup draws be included? Defaults to FALSE. Ignored except when used with CmdStanMCMC objects.


Logical of length 1, whether to include warmup draws in the big MCMC object (the target with "mcmc" in the name). inc_warmup must not be TRUE if inc_warmup_fit is FALSE.


Optional list of summary functions passed to ... in posterior::summarize_draws() through $summary() on the CmdStanFit object.


Optional list of summary function arguments passed to .args in posterior::summarize_draws() through $summary() on the CmdStanFit object.


Logical, whether to create a target for posterior draws. Saves posterior::as_draws_df(fit$draws()) to a compressed tibble. Convenient, but duplicates storage.


Logical, whether to create a target for posterior::as_draws_df(fit$sampler_diagnostics()). Saves posterior::as_draws_df(fit$draws()) to a compressed tibble. Convenient, but duplicates storage.


Logical, whether to create a target for fit$summary().


Deprecated on 2022-07-22. Use return_draws instead.


Deprecated on 2022-07-22. Use return_summary instead.


Logical, whether to enable tidy evaluation when interpreting command and pattern. If TRUE, you can use the "bang-bang" operator !! to programmatically insert the values of global objects.


Character vector of packages to load right before the target runs or the output data is reloaded for downstream targets. Use tar_option_set() to set packages globally for all subsequent targets you define.


Character vector of library paths to try when loading packages.


Character of length 1, storage format of the non-data-frame targets such as the Stan data and any CmdStanFit objects. Please choose an all=purpose format such as "qs" or "aws_qs" rather than a file format like "file" or a data frame format like "parquet". For more on storage formats, see the help file of targets::tar_target().


Character of length 1, storage format of the data frame targets such as posterior draws. We recommend efficient data frame formats such as "feather" or "aws_parquet". For more on storage formats, see the help file of targets::tar_target().


Character of length 1, remote repository for target storage. Choices:

Note: if repository is not "local" and format is "file" then the target should create a single output file. That output file is uploaded to the cloud and tracked for changes where it exists in the cloud. The local file is deleted after the target runs.


Character of length 1, what to do if the target stops and throws an error. Options:

  • "stop": the whole pipeline stops and throws an error.

  • "continue": the whole pipeline keeps going.

  • "abridge": any currently running targets keep running, but no new targets launch after that. (Visit to learn how to debug targets using saved workspaces.)

  • "null": The errored target continues and returns NULL. The data hash is deliberately wrong so the target is not up to date for the next run of the pipeline.


Character of length 1, memory strategy. If "persistent", the target stays in memory until the end of the pipeline (unless storage is "worker", in which case targets unloads the value from memory right after storing it in order to avoid sending copious data over a network). If "transient", the target gets unloaded after every new target completes. Either way, the target gets automatically loaded into memory whenever another target needs the value. For cloud-based dynamic files (e.g. format = "file" with repository = "aws"), this memory strategy applies to the temporary local copy of the file: "persistent" means it remains until the end of the pipeline and is then deleted, and "transient" means it gets deleted as soon as possible. The former conserves bandwidth, and the latter conserves local storage.


Logical, whether to run base::gc() just before the target runs.


Character of length 1. If deployment is "main", then the target will run on the central controlling R process. Otherwise, if deployment is "worker" and you set up the pipeline with distributed/parallel computing, then the target runs on a parallel worker. For more on distributed/parallel computing in targets, please visit


Numeric of length 1 between 0 and 1. Controls which targets get deployed first when multiple competing targets are ready simultaneously. Targets with priorities closer to 1 get dispatched earlier (and polled earlier in tar_make_future()).


Object returned by tar_resources() with optional settings for high-performance computing functionality, alternative data storage formats, and other optional capabilities of targets. See tar_resources() for details.


Character of length 1, only relevant to tar_make_clustermq() and tar_make_future(). Must be one of the following values:

  • "main": the target's return value is sent back to the host machine and saved/uploaded locally.

  • "worker": the worker saves/uploads the value.

  • "none": almost never recommended. It is only for niche situations, e.g. the data needs to be loaded explicitly from another language. If you do use it, then the return value of the target is totally ignored when the target ends, but each downstream target still attempts to load the data file (except when retrieval = "none").

    If you select storage = "none", then the return value of the target's command is ignored, and the data is not saved automatically. As with dynamic files (format = "file") it is the responsibility of the user to write to the data store from inside the target.

    The distinguishing feature of storage = "none" (as opposed to format = "file") is that in the general case, downstream targets will automatically try to load the data from the data store as a dependency. As a corollary, storage = "none" is completely unnecessary if format is "file".


Character of length 1, only relevant to tar_make_clustermq() and tar_make_future(). Must be one of the following values:

  • "main": the target's dependencies are loaded on the host machine and sent to the worker before the target runs.

  • "worker": the worker loads the targets dependencies.

  • "none": the dependencies are not loaded at all. This choice is almost never recommended. It is only for niche situations, e.g. the data needs to be loaded explicitly from another language.


An optional object from tar_cue() to customize the rules that decide whether the target is up to date.


Character of length 1, a custom free-form human-readable text description of the target. Descriptions appear as target labels in functions like tar_manifest() and tar_visnetwork(), and they let you select subsets of targets for the names argument of functions like tar_make(). For example, tar_manifest(names = tar_described_as(starts_with("survival model"))) lists all the targets whose descriptions start with the character string "survival model".


tar_stan_mcmc() returns a list of target objects. See the "Target objects" section for background. The target names use the name argument as a prefix, and the individual elements of stan_files appear in the suffixes where applicable. As an example, the specific target objects returned by tar_stan_mcmc(name = x, stan_files = "y.stan", ...)

are as follows.

  • x_file_y: reproducibly track the Stan model file. Returns a character vector with paths to the model file and compiled executable.

  • x_lines_y: read the Stan model file for safe transport to parallel workers. Omitted if compile = "original". Returns a character vector of lines in the model file.

  • x_data: run the R expression in the data argument to produce a Stan dataset for the model. Returns a Stan data list.

  • x_mcmc_y: run MCMC on the model and the dataset. Returns a cmdstanr CmdStanMCMC object with all the results.

  • x_draws_y: extract draws from x_mcmc_y. Omitted if draws = FALSE. Returns a tidy data frame of draws.

  • x_summary_y: extract compact summaries from x_mcmc_y. Returns a tidy data frame of summaries. Omitted if summary = FALSE.

  • x_diagnostics: extract HMC diagnostics from x_mcmc_y. Returns a tidy data frame of HMC diagnostics. Omitted if diagnostics = FALSE.


Most of the arguments are passed to the $compile(), $sample(), and $summary() methods of the CmdStanModel class. If you previously compiled the model in an upstream tar_stan_compile() target, then the model should not recompile.

Target objects

Most stantargets functions are target factories, which means they return target objects or lists of target objects. Target objects represent skippable steps of the analysis pipeline as described at Please read the walkthrough at to understand the role of target objects in analysis pipelines.

For developers, explains target factories (functions like this one which generate targets) and the design specification at details the structure and composition of target objects.


if (Sys.getenv("TAR_LONG_EXAMPLES") == "true") {
targets::tar_dir({ # tar_dir() runs code from a temporary directory.
# Do not use temporary storage for stan files in real projects
# or else your targets will always rerun.
path <- tempfile(pattern = "", fileext = ".stan")
tar_stan_example_file(path = path)
    stan_files = path,
    data = tar_stan_example_data(),
    variables = "beta",
    summaries = list(~quantile(.x, probs = c(0.25, 0.75))),
    stdout = R.utils::nullfile(),
    stderr = R.utils::nullfile()
}, ask = FALSE)