Inspects a configuration before a long run and reports: total task count,
grid dimensions, metrics and their needs vs the fitter's capabilities
(surfacing mismatch warnings before a 10-hour run), whether mirai daemons
are set, and (for BrmsFitter) the number of distinct models to compile.
When run automatically at the top of run_simulation(), it prints a
condensed one-line summary.
Arguments
- config
A
SimulationConfig.- pilot
Logical; if TRUE, run a single pilot task and extrapolate its wall-clock time to an estimate of the total study time (default FALSE). The estimate is printed and returned in the
pilot_secondsandestimated_total_secondselements of the returned list. A pilot that cannot execute (e.g. a Stan fitter with no CmdStan installed) yields NA estimates rather than failing the preflight. Note that for Stan fitters a pilot compiles the model, so it is not instant.- condensed
Logical; if TRUE, print a single one-line summary instead of the full report (used by run_simulation).