slv.inversion.sweep.Sweep.run#

Sweep.run(results_dir, n_jobs=1, swept_params=None, resume=True, config_ids=None)[source]#

Run the sweep.

Parameters:

  • results_dir (str | Path) – Directory for results CSV and grid JSON.

  • n_jobs (int) – Number of parallel workers (0 = write grid only, no runs).

  • swept_params (list[str] | None) – Config fields to capture as columns. None = all scientific fields.

  • resume (bool) – Skip configs already in results CSV.

  • config_ids (list[str] | None) – Optional list of config IDs to run. If None, run all configs. Use this to run a random sample of the full sweep.

Return type:

SweepResults wrapping the output CSV.