Mechanism

This module contains various surrogate models used in the illumination library.

Descriptor Class

class illumination.mechanism.Descriptor(config)

Bases: object

A strategy class for calculating the descriptor vector of a molecule.

static rescale(feature: List[float], range: List[float]) → List[float]

Rescales the feature to the unit range.

Surrogate Class

class illumination.mechanism.Surrogate(config)

Bases: object

A factory class for creating instances of various surrogate functions based on the provided configuration. The supported surrogate function types are “Smiles” and “Fingerprint”. Note that Fingerprint acquisition functions need further specifications by passing on the config file.

Methods: __new__: Static method for creating and returning an instance of a specific fitness function based on the configuration.

Acquisition Class

class illumination.mechanism.Acquisition(config)

Bases: object

A factory class for creating instances of various acquisition functions based on the provided configuration. Requires an Archive instance to be passed with the configuration file. The supported acquisition function types are “Mean”, “UCB”, “EI”, and “logEI”.

Methods: __new__: Static method for creating and returning an instance of a specific acquisition function based on the configuration.