Acquisition

This module contains various acquisition functions used in the illumination library.

Posterior Mean

class illumination.functions.acquisition.Posterior_Mean(config)

Bases: BO_Acquisition

A strategy class for the posterior mean of a list of molecules.

calculate_acquisition_value(molecules) → None

Updates the acquisition value for a list of molecules based on their predicted fitness.

Args:

molecules: A list of Molecule objects to evaluate.

Returns:

None

Upper Confidence Bound

class illumination.functions.acquisition.Upper_Confidence_Bound(config)

Bases: BO_Acquisition

A strategy class for the upper confidence bound of a list of molecules.

calculate_acquisition_value(molecules) → None

Updates the acquisition value for a list of molecules based on their predicted fitness and uncertainty.

Args:

molecules: A list of Molecule objects to evaluate.

Returns:

None

Expected Improvement

class illumination.functions.acquisition.Expected_Improvement(config)

Bases: BO_Acquisition

A strategy class for the expected improvement of a list of molecules.

calculate_acquisition_value(molecules) → None

Updates the acquisition value for a list of molecules based on their expected improvement.

Args:

molecules: A list of Molecule objects to evaluate.

Returns:

None

Log Expected Improvement

class illumination.functions.acquisition.Log_Expected_Improvement(config)

Bases: BO_Acquisition

A strategy class for the numerically stable logarithm of the expected improvement of a list of molecules.

calculate_acquisition_value(molecules) → None

Updates the acquisition value for a list of molecules based on the logarithm of their expected improvement.

Args:

molecules: A list of Molecule objects to evaluate.

Returns:

None