Compute Configuration Fingerprint — compute_config

Computes a cryptographic hash of the simulation configuration. The fingerprint uniquely identifies a simulation configuration for caching and deduplication purposes.

Usage

compute_config_fingerprint(config)

Arguments

config: An S7 SimulationConfig object or a StudySpec created by new_study_spec().

Value

A character string containing the SHA256 hash of the study design.

Details

The fingerprint excludes runtime policy settings (B4):

result_path: Output location doesn't affect simulation identity
checkpoint_every / checkpoint_format: checkpoint cadence/format is runtime optimization
retain: retention is runtime policy; fingerprint exclusion does not make every retention change legal on resume (a compatible resume may narrow retention; widening is rejected once completed outcomes lack the requested artifacts)
max_errors: error tolerance is runtime policy

Examples

if (FALSE) { # \dontrun{
config <- simulation_config(
  data_grid = data.frame(n = 100),
  fit_grid = data.frame(model = "baseline"),
  data_generator = my_data_gen,
  n_replicates = 10L,
  seed = 42L
)

fingerprint <- compute_config_fingerprint(config)
# Use fingerprint for caching or deduplication
} # }