poli.objective_repository.penalized_logp_lambo.register.PenalizedLogPLamboBlackBox

class poli.objective_repository.penalized_logp_lambo.register.PenalizedLogPLamboBlackBox(string_representation: Literal['SMILES', 'SELFIES'] = 'SMILES', penalized: bool = True, batch_size: Optional[int] = None, parallelize: bool = False, num_workers: Optional[int] = None, evaluation_budget: int = inf, force_isolation: bool = False)

A black box objective function that returns the penalized logP of a molecule, using the same function that LaMBO [1] uses.

In particular, they adjust the penalized logP using some “magic numbers”, which are the empirical means and standard deviations of the dataset.

__init__(string_representation: Literal['SMILES', 'SELFIES'] = 'SMILES', penalized: bool = True, batch_size: Optional[int] = None, parallelize: bool = False, num_workers: Optional[int] = None, evaluation_budget: int = inf, force_isolation: bool = False)

Initialize the AbstractBlackBox object.

Parameters

• batch_size (int, optional) – The batch size for parallel execution, by default None.

• parallelize (bool, optional) – Flag indicating whether to parallelize the execution, by default False.

• num_workers (int, optional) – The number of workers for parallel execution, by default we use half the available CPUs.

• evaluation_budget (int, optional) – The maximum number of evaluations allowed for the black box function, by default float(“inf”).

Methods

__init__([string_representation, penalized, ...])	Initialize the AbstractBlackBox object.
reset_evaluation_budget()	Resets the evaluation budget by setting the number of evaluations made to 0.
set_observer(observer)	Set the observer object for recording observations during evaluation.
terminate()	Terminate the black box optimization problem.