qNegIntegratedPosteriorVariance¶
- class baybe.acquisition.acqfs.qNegIntegratedPosteriorVariance[source]¶
Bases:
AcquisitionFunction
Monte Carlo based negative integrated posterior variance.
This is typically used for active learning as it is a measure for global model uncertainty.
Public methods
__init__
([sampling_n_points, ...])Method generated by attrs for class qNegIntegratedPosteriorVariance.
evaluate
(candidates, surrogate, searchspace, ...)Get the acquisition values for the given candidates.
from_dict
(dictionary)Create an object from its dictionary representation.
from_json
(string)Create an object from its JSON representation.
get_integration_points
(searchspace)Sample points from a search space for integration purposes.
to_botorch
(surrogate, searchspace, ...[, ...])Create the botorch-ready representation of the function.
to_dict
()Create an object's dictionary representation.
to_json
()Create an object's JSON representation.
Public attributes and properties
Number of data points sampled for integrating the posterior.
Fraction of data sampled for integrating the posterior.
Sampling strategy used for integrating the posterior.
An alternative name for type resolution.
Whether this acquisition function can handle models with multiple outputs.
- __init__(sampling_n_points: int | None = None, sampling_fraction=NOTHING, sampling_method=DiscreteSamplingMethod.Random)¶
Method generated by attrs for class qNegIntegratedPosteriorVariance.
For details on the parameters, see Public attributes and properties.
- evaluate(candidates: DataFrame, surrogate: SurrogateProtocol, searchspace: SearchSpace, objective: Objective, measurements: DataFrame, pending_experiments: DataFrame | None = None, *, jointly: bool = False)¶
Get the acquisition values for the given candidates.
- Parameters:
candidates (
DataFrame
) – The candidate points in experimental representation. For details, seebaybe.surrogates.base.Surrogate.posterior()
.surrogate (
SurrogateProtocol
) – The surrogate model to use for the acquisition function.searchspace (
SearchSpace
) – The search space. Seebaybe.recommenders.base.RecommenderProtocol.recommend()
.objective (
Objective
) – The objective. Seebaybe.recommenders.base.RecommenderProtocol.recommend()
.measurements (
DataFrame
) – Available experimentation data. Seebaybe.recommenders.base.RecommenderProtocol.recommend()
.pending_experiments (
Optional
[DataFrame
]) – Optional pending experiments. Seebaybe.recommenders.base.RecommenderProtocol.recommend()
.jointly (
bool
) – IfFalse
, the acquisition values are computed independently for each candidate. IfTrue
, a single joint acquisition value is computed for the entire candidate set.
- Return type:
- Returns:
Depending on the joint mode, either a single batch acquisition value or a series of individual acquisition values.
- get_integration_points(searchspace: SearchSpace)[source]¶
Sample points from a search space for integration purposes.
Sampling of the discrete part can be controlled via ‘sampling_method’, but sampling of the continuous part will always be random.
- Parameters:
searchspace (
SearchSpace
) – The searchspace from which to sample integration points.- Return type:
- Returns:
The sampled data points.
- Raises:
ValueError – If the search space is purely continuous and ‘sampling_n_points’ was not provided.
- to_botorch(surrogate: SurrogateProtocol, searchspace: SearchSpace, objective: Objective, measurements: DataFrame, pending_experiments: DataFrame | None = None)¶
Create the botorch-ready representation of the function.
The required structure of measurements is specified in
baybe.recommenders.base.RecommenderProtocol.recommend()
.- Return type:
- to_json()¶
Create an object’s JSON representation.
- Return type:
- Returns:
The JSON representation as a string.
-
sampling_fraction:
float
|None
¶ Fraction of data sampled for integrating the posterior.
Cannot be used if sampling_n_points is not None.
-
sampling_method:
DiscreteSamplingMethod
¶ Sampling strategy used for integrating the posterior.