pub fn run_sweep_parallel(
lattice: &Lattice,
realizations: &mut [Realization],
n_replicas: usize,
n_temps: usize,
config: &SimConfig,
interrupted: &AtomicBool,
on_sweep: &(dyn Fn() + Sync),
) -> Result<SweepResult, String>Expand description
Run the sweep loop in parallel over multiple disorder realizations.
Each realization is processed by run_sweep_loop, then results are
averaged via SweepResult::aggregate. For a single realization the
call is made directly, skipping rayon thread-pool overhead.