NEWS

stantargets 0.1.3

To align with https://github.com/ropensci/targets/issues/1244 and https://github.com/ropensci/targets/pull/1262, switch the hashing functions from digest::digest() to secretbase::siphash13().

Remove temporary files generated CmdStan.
Add the new description arguments of tar_target() (`targets >= 1.5.1.9001).
Append model file information to the target descriptions using tar_map() (tarchetypes >= 0.7.12.9001).

Add a new transform argument to the tar_stan_*_rep_draws() functions (except mle) to support simulation-based calibration (SBC) (#31, @mike-lawrence, @martinmodrak, @dmi3kno).

Implement resilient rep-specific seeds that do not change if the batching structure changes.

Update the signatures of most user-side functions to accommodate changes from cmdstanr 0.2.0 to 0.5.0 (https://github.com/ropensci/targets/discussions/884, @stuvet).
Add new arguments variables_fit and inc_warmup_fit to be passed to the CmdStanFit target in tar_stan_mcmc() instead of variables and inc_warmup (#56, @yizhang-yiz). Same for the variables argument of tar_stan_gq(), tar_stan_mle(), and tar_stan_vb().

Support the repository argument for targets >= 0.11.0.
Append a new .dataset_id column to target outputs to aid in model comparisons across the same datasets.

Skip tests if CmdStan is not installed (@sakrejda).
Use custom generate_data() function in the docs, as opposed to tar_stan_example_data() directly (@sakrejda).
Add the pedantic argument for compilation (@sakrejda).
Reduce dependencies on some rlang functions like sym() (@sakrejda).
Change trn() to if_any() (@mattwarkentin, @sakrejda, @tjmahr).
Add @sakrejda and @mattwarkentin as reviewers in the DESCRIPTION.
Talk about the R package and system dependencies of stantargets in the README (@mattwarkentin).
Throw an error earlier if the Stan file does not exist (@sakrejda, @mattwarkentin)
Use @format roxygen2 tag for data generation (@mattwarkentin).
Use @family go cross-reference functions (@mattwarkentin).
Elaborate on the roles and return values of specific targets generated by target factories (@mattwarkentin).
Undergo rOpenSci peer review and transition to rOpenSci.
Link to an example project.

Return the executable file after the Stan source file in model compilation targets.
Replace the log argument with stdout and stderr (#23).
Switch meaning of %||% and %|||% to conform to historical precedent.

Join on data to summary output using .join_data in the Stan data (#18).
Pre-compile models for testing and add an environment variable to skip tests that always force recompilation (#19).
Load packages for any target computing summaries.