shapiq ¶

get_tabpfn_explainer ¶

get_tabpfn_explainer(
    model: Union[TabPFNRegressor, TabPFNClassifier],
    data: Union[DataFrame, ndarray],
    labels: Union[DataFrame, ndarray],
    index: str = "k-SII",
    max_order: int = 2,
    class_index: Optional[int] = None,
    **kwargs
)

Get a TabPFNExplainer from shapiq.

This function returns the TabPFN explainer from the shapiq[1]_ library. The explainer uses a remove-and-recontextualize paradigm of model explanation[2][3] to explain the predictions of a TabPFN model. See shapiq.TabPFNExplainer documentation for more information regarding the explainer object.

Parameters:

Name	Type	Description	Default
`model`	`TabPFNRegressor or TabPFNClassifier`	The TabPFN model to explain.	required
`data`	`DataFrame or ndarray`	The background data to use for the explainer.	required
`labels`	`DataFrame or ndarray`	The labels for the background data.	required
`index`	`str`	The index to use for the explanation. See shapiq documentation for more information and an up-to-date list of available indices. Defaults to "k-SII" and "SV" (Shapley Values like SHAP) with `max_order=1`.	`'k-SII'`
`max_order`	`int`	The maximum order of interactions to consider. Defaults to 2.	`2`
`class_index`	`int`	The class index of the model to explain. If not provided, the class index will be set to 1 per default for classification models. This argument is ignored for regression models. Defaults to None.	`None`
`**kwargs`		Additional keyword arguments to pass to the explainer.	`{}`

Returns:

Type	Description
	shapiq.TabPFNExplainer: The TabPFN explainer.

References

.. [1] shapiq repository: https://github.com/mmschlk/shapiq .. [2] Muschalik, M., Baniecki, H., Fumagalli, F., Kolpaczki, P., Hammer, B., Hüllermeier, E. (2024). shapiq: Shapley Interactions for Machine Learning. In: The Thirty-eight Conference on Neural Information Processing Systems Datasets and Benchmarks Track. url: https://openreview.net/forum?id=knxGmi6SJi .. [3] Rundel, D., Kobialka, J., von Crailsheim, C., Feurer, M., Nagler, T., Rügamer, D. (2024). Interpretable Machine Learning for TabPFN. In: Longo, L., Lapuschkin, S., Seifert, C. (eds) Explainable Artificial Intelligence. xAI 2024. Communications in Computer and Information Science, vol 2154. Springer, Cham. https://doi.org/10.1007/978-3-031-63797-1_23

get_tabpfn_imputation_explainer ¶

get_tabpfn_imputation_explainer(
    model: Union[TabPFNRegressor, TabPFNClassifier],
    data: Union[DataFrame, ndarray],
    index: str = "k-SII",
    max_order: int = 2,
    imputer: str = "marginal",
    class_index: Optional[int] = None,
    **kwargs
)

Gets a TabularExplainer from shapiq with using imputation.

This function returns the TabularExplainer from the shapiq[1][2] library. The explainer uses an imputation-based paradigm of feature removal for the explanations similar to SHAP[3]_. See shapiq.TabularExplainer documentation for more information regarding the explainer object.

Parameters:

Name	Type	Description	Default
`model`	`TabPFNRegressor or TabPFNClassifier`	The TabPFN model to explain.	required
`data`	`DataFrame or ndarray`	The background data to use for the explainer.	required
`index`	`str`	The index to use for the explanation. See shapiq documentation for more information and an up-to-date list of available indices. Defaults to "k-SII" and "SV" (Shapley Values like SHAP) with `max_order=1`.	`'k-SII'`
`max_order`	`int`	The maximum order of interactions to consider. Defaults to 2.	`2`
`imputer`	`str`	The imputation method to use. See `shapiq.TabularExplainer` documentation for more information and an up-to-date list of available imputation methods.	`'marginal'`
`class_index`	`int`	The class index of the model to explain. If not provided, the class index will be set to 1 per default for classification models. This argument is ignored for regression models. Defaults to None.	`None`
`**kwargs`		Additional keyword arguments to pass to the explainer.	`{}`

Returns:

Type	Description
	shapiq.TabularExplainer: The TabularExplainer.

References

.. [1] shapiq repository: https://github.com/mmschlk/shapiq .. [2] Muschalik, M., Baniecki, H., Fumagalli, F., Kolpaczki, P., Hammer, B., Hüllermeier, E. (2024). shapiq: Shapley Interactions for Machine Learning. In: The Thirty-eight Conference on Neural Information Processing Systems Datasets and Benchmarks Track. url: https://openreview.net/forum?id=knxGmi6SJi .. [3] Lundberg, S. M., & Lee, S. I. (2017). A Unified Approach to Interpreting Model Predictions. Advances in Neural Information Processing Systems 30 (pp. 4765--4774).