cem_plan#
- hybrid_aif.planning.cem_plan(
- ctx,
- key,
- *,
- num_candidates=64,
- cem_iters=5,
- elite_frac=0.125,
- init_mean=None,
- init_std=None,
- min_std=None,
- use_bound='midpoint',
- metrics=True,
Run CEM against
ctx; see the module docstring for the algorithm.init_mean(default: the box midpoint) is the warm start, e.g. the previous plan shifted one step.init_std/min_stddefault toINIT_STD_FRACTION/MIN_STD_FRACTIONof the box width; explicit values are in action units, scalars or(action_dim,)arrays.metrics=Falseskips the term tree without changing the posterior.