xai_compare.abstract.explainer

Classes

Explainer(model, X_train, y_train[, y_pred, ...])

An abstract base class for creating explainers that can interpret the predictions made by machine learning models.

class xai_compare.abstract.explainer.Explainer(model, X_train: DataFrame, y_train: DataFrame | Series | ndarray, y_pred: DataFrame | Series | ndarray | None = None, mode: str = 'regression')

An abstract base class for creating explainers that can interpret the predictions made by machine learning models.

Attributes:

model:

A machine learning model whose predictions are to be interpreted.

X_train (pd.DataFrame):

Training data used to fit the model.

y_train (Union[pd.DataFrame, pd.Series, np.ndarray]):

Training labels or targets.

y_pred (Union[pd.DataFrame, pd.Series, np.ndarray, None], optional):

Predicted values. Defaults to None.

mode (str):

The mode of the explainer, which could be ‘regression’ or ‘classification’ from config.py.

Methods:

explain_global(x_data: pd.DataFrame) -> pd.DataFrame:

Abstract method to compute global explanations.

explain_local(x_data: pd.DataFrame) -> pd.DataFrame:

Abstract method to compute local explanations.

abstract explain_global(x_data: DataFrame) → DataFrame

Generates a global explanation of the model predictions over the entire dataset.

Attributes:

x_data (pd.DataFrame):

Dataset for which the global explanation is required.

Returns:

pd.DataFrame:

A DataFrame containing the global explanation results.

abstract explain_local(x_data: DataFrame) → DataFrame

Generates a local explanation of the model predictions for individual samples.

Attributes:

x_data (pd.DataFrame):

Dataset for which local explanations are required.

Returns:

pd.DataFrame:

A DataFrame containing the local explanation results for each sample.