meld.remd.adaptor.EqualAcceptanceAdaptor

class meld.remd.adaptor.EqualAcceptanceAdaptor(n_replicas, adaptation_policy, min_acc_prob=0.1)

Bases: _AcceptanceCounter

Adaptor based on making acceptance rates uniform

__init__(n_replicas, adaptation_policy, min_acc_prob=0.1)

Initialize EqualAcceptanceAdaptor

Parameters

• n_replicas (int) – number of replicas

• adaptation_policy (AdaptationPolicy) – policy to use to decide when to adapt

• min_acc_prob (float) – floor on the acceptance probability used for adaptation

Methods

__init__(n_replicas, adaptation_policy[, ...])	Initialize EqualAcceptanceAdaptor
adapt(previous_lambdas, step)	Compute new optimal values of lambda.
get_acceptance_probabilities()	Get acceptance probabilities
reset()	Forget about any previous updates.
update(i, accepted)	Update statistics

Attributes

accept_probs
t_lens

adapt(previous_lambdas, step)

Compute new optimal values of lambda.

Parameters

• previous_lambdas (List[float]) – the previous lambda values

• step (int) – the current simulation step

Return type

List[float]

Returns

the new, optimized lambda values

reset()

Forget about any previous updates.

Resets all internal counters and statistics to zero.

Return type

None

update(i, accepted)

Update statistics

Parameters

• i (int) – lower replica index of pair with attempted swap

• accepted (bool) – True if swap was successful

Return type

None