MC simulation

Detailed Description

Run and manage MC simulations.

This group contains triqs::mc_tools::mc_generic, the driver that ties together moves, measurements and the random generator. A typical lifecycle is:

1. construct an mc_generic<S> with a chosen engine name and seed,
2. register moves with add_move, measurements with add_measure, and any auxiliary callables with add_measure_aux,
3. run warmup (equilibrate the chain, optionally calibrate moves), then accumulate (perform measurements) — or the combined warmup_and_accumulate / run,
4. call collect_results to reduce per-rank estimators across an mpi::communicator.

The driver organises work into cycles of MC steps separated by measurement passes, and checks for user-requested termination and for exceptions on configurable intervals so that ranks can stop in a coordinated way even with asymmetric workloads.

Classes
class	triqs::mc_tools::mc_generic< MCSignType >
	Generic Monte Carlo class. More...