ETIA.CausalLearning.CDHPO package

Subpackages

Submodules

ETIA.CausalLearning.CDHPO.CDHPOBase module

class CDHPOBase(configurations, dataset)

Bases: object

A base class for Causal Discovery Hyperparameter Optimization (CDHPO) algorithms. This class defines the basic structure for implementing hyperparameter optimization algorithms.

run()

Runs the hyperparameter optimization process. This method should be overridden by subclasses.

run_new()

Re-runs the optimization process with new or updated configurations. This method should be overridden by subclasses.

find_best_config(algorithms)

Finds the best configuration from a list of algorithms. This method should be overridden by subclasses.

save_progress(path)

Saves the current progress of the optimization process.

load_progress(path)

Loads the progress of a previously run optimization from the specified path.

run()

Runs the hyperparameter optimization process. This method should be overridden by subclasses to provide the specific optimization algorithm.

Returns:

The optimal configuration after the hyperparameter optimization process.

Return type:

object

run_new()

Re-runs the optimization process, typically with new or updated configurations. This method should be overridden by subclasses to define the behavior for restarting or continuing optimization.

find_best_config(algorithms)

Finds the best configuration among the given algorithms.

Parameters:

algorithms (list) – A list of algorithms to evaluate and choose the best configuration from.

Returns:

The best configuration determined from the list of algorithms.

Return type:

object

save_progress(path)

Saves the progress of the optimization process to a file.

Parameters:

path (str) – The file path where the progress should be saved.

load_progress(path)

Loads the progress of a previously run optimization from a file.

Parameters:

path (str) – The file path from where the progress should be loaded.