ETIA.CausalLearning.algorithms.cdt_algorithms package

Submodules

ETIA.CausalLearning.algorithms.cdt_algorithms.SAMAlgorithm module

class SAMAlgorithm(algorithm='sam', verbose=False)

Bases: object

A class that implements the SAM (Structural Agnostic Model) algorithm for causal discovery.

check_parameters()

Validates the parameters for running the SAM algorithm.

prepare_data(data)

Prepares the data for the SAM algorithm by converting it to a pandas DataFrame.

set_parameters(parameters)

Sets the algorithm’s parameters from a provided dictionary, using defaults where necessary.

run(data, parameters, prepare_data=True)

Runs the SAM algorithm on the provided data and parameters, and returns the learned causal structure.

check_parameters()

Validates the parameters required for running the SAM algorithm.

Returns:

True if all parameters are valid, raises ValueError otherwise.

Return type:

bool

Raises:

ValueError – If any parameter is out of the valid range or not of the expected type.

prepare_data(data)

Prepares the data for the SAM algorithm by converting it to a pandas DataFrame if necessary.

Parameters:

data (any) – The dataset to be used. Can be a pandas DataFrame or an object that implements the get_dataset method.

Returns:

data – The prepared dataset as a pandas DataFrame.

Return type:

pandas.DataFrame

set_parameters(parameters)

Sets the parameters for the SAM algorithm, using default values where necessary.

Parameters:

parameters (dict) – A dictionary containing the parameters to set, such as learning rates, lambda values, number of hidden units, etc.

Raises:

ValueError – If any of the parameters are invalid.

run(data, parameters, prepare_data=True)

Runs the SAM algorithm on the provided data and parameters.

Parameters:

• data (any) – The dataset to be used, either as a pandas DataFrame or an object implementing get_dataset.

• parameters (dict) – The parameters to configure the SAM algorithm.

• prepare_data (bool, optional) – If True, the data will be prepared before running the algorithm. Default is True.

Returns:

A tuple containing: - mec_graph : DAGWrapper

The learned causal structure represented as a graph.

• library_resultsdict

  A dictionary containing the resulting graph and additional results.

Return type:

tuple

Raises:

ValueError – If any of the parameters are invalid.