archetypes.FairAA#

class archetypes.FairAA(n_archetypes, *, fairness_const=200, max_iter=300, tol=0.0001, init='uniform', n_init=1, init_params=None, save_init=False, method='nnls', method_params=None, verbose=False, random_state=None)#

Fair Archetype Analysis.

This class implements the Fair Archetype Analysis algorithm, which is a variant of the archetype analysis that incorporates fairness constraints.

Parameters:

n_archetypes: int: The number of archetypes to compute.
fairness_const: float, default=200.0
max_iterint, default=300: Maximum number of iterations of the archetype analysis algorithm for a single run.
tolfloat, default=1e-4: Relative tolerance of two consecutive iterations to declare convergence.
initstr, default=’uniform’: Method used to initialize the archetypes, must be one of the following: ‘uniform’, ‘furthest_sum’, ‘furthest_first’ or ‘aa_plus_plus’. See Initialization Methods.
n_initint, default=1: Number of time the archetype analysis algorithm will be run with different initializations. The final results will be the best output of n_init consecutive runs.
init_paramsdict, default=None: Additional keyword arguments to pass to the initialization method.
save_initbool, default=False: If True, save the initial archetypes in the attribute archetypes_init_,
method: str, default=’pgd’: The optimization method to use for the archetypes and the coefficients, must be one of the following: ‘pgd’, ‘pseudo_pgd’. See Optimization Methods.
method_paramsdict, default=None: Additional arguments to pass to the optimization method. See Optimization Methods.
verbosebool, default=False: Verbosity mode.
random_stateint, RandomState instance or None, default=None: Determines random number generation of coefficients in initialization. Use an int to make the randomness reproducible.

Attributes:

archetypes_: np.ndarray: The computed archetypes. It has shape (n_archetypes, n_features).
n_archetypes_: int: The number of archetypes after fitting.
archetypes_init_np.ndarray: The initial archetypes. It is only available if save_init=True.
coefficients_, A_np.ndarray: The similarity degree of each sample to each archetype. It has shape (n_samples, n_archetypes).
arch_coefficients_, B_np.ndarray: The similarity degree of each archetype to each sample. It has shape (n_archetypes, n_samples).
labels_np.ndarray: The label of each sample. It is the index of the closest archetype. It has shape (n_samples,).
loss_list: The loss at each iteration.
rss_, reconstruction_error_float: The residual sum of squares of the fitted data.

Methods

`fit`(X[, y, Z])	Compute Archetype Analysis.
`fit_transform`(X[, y, Z])	Compute the archetypes and transform X to the archetypal space.
`get_metadata_routing`()	Get metadata routing of this object.
`get_params`([deep])	Get parameters for this estimator.
`set_fit_request`(*[, Z])	Configure whether metadata should be requested to be passed to the `fit` method.
`set_output`(*[, transform])	Set output container.
`set_params`(**params)	Set the parameters of this estimator.
`set_transform_request`(*[, Z])	Configure whether metadata should be requested to be passed to the `transform` method.
`transform`(X, Z)	Transform X to the archetypal space.

fit(X, y=None, Z=None)#

Compute Archetype Analysis.

Parameters:

Xarray-like of shape (n_samples, n_features): Training instances to compute the archetypes. It must be noted that the data will be converted to C ordering, which will cause a memory copy if the given data is not C-contiguous.
yIgnored: Not used, present here for API consistency by convention.
Zarray-like of shape (n_samples, n_sensitive_features): Sensitive features to compute the fairness constraint.

Returns:

selfobject: Fitted estimator.

fit_transform(X, y=None, Z=None, **params)#

Compute the archetypes and transform X to the archetypal space.

Equivalent to fit(X).transform(X).

Parameters:

Xarray-like of shape (n_samples, n_features): New data to transform.
yIgnored: Not used, present here for API consistency by convention.
Zarray-like of shape (n_samples, n_sensitive_features): Sensitive features to compute the fairness constraint.
Returns
——-
Andarray of shape (n_samples, n_archetypes): X transformed in the new space.

get_metadata_routing()#

Get metadata routing of this object.

Please check User Guide on how the routing mechanism works.

Returns:

routingMetadataRequest: A MetadataRequest encapsulating routing information.

get_params(deep=True)#

Get parameters for this estimator.

Parameters:

deepbool, default=True: If True, will return the parameters for this estimator and contained subobjects that are estimators.

Returns:

paramsdict: Parameter names mapped to their values.

set_fit_request(*, Z: bool | None | str = '$UNCHANGED$') → FairAA#

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.
False: metadata is not requested and the meta-estimator will not pass it to fit.
None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.
str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters:

Zstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for Z parameter in fit.

Returns:

selfobject: The updated object.

set_output(*, transform=None)#

Set output container.

See Introducing the set_output API for an example on how to use the API.

Parameters:

transform{“default”, “pandas”, “polars”}, default=None

Configure output of transform and fit_transform.

“default”: Default output format of a transformer
“pandas”: DataFrame output
“polars”: Polars output
None: Transform configuration is unchanged

Added in version 1.4: “polars” option was added.

Returns:

selfestimator instance: Estimator instance.

set_params(**params)#

Set the parameters of this estimator.

The method works on simple estimators as well as on nested objects (such as Pipeline). The latter have parameters of the form <component>__<parameter> so that it’s possible to update each component of a nested object.

Parameters:

**paramsdict: Estimator parameters.

Returns:

selfestimator instance: Estimator instance.

set_transform_request(*, Z: bool | None | str = '$UNCHANGED$') → FairAA#

Configure whether metadata should be requested to be passed to the transform method.

The options for each parameter are:

True: metadata is requested, and passed to transform if provided. The request is ignored if metadata is not provided.
False: metadata is not requested and the meta-estimator will not pass it to transform.
None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.
str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters:

Zstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for Z parameter in transform.

Returns:

selfobject: The updated object.

transform(X, Z)#

Transform X to the archetypal space.

In the new space, each dimension is the distance to the archetypes.

Parameters:

Xarray-like of shape (n_samples, n_features): New data to transform.
Zarray-like of shape (n_samples, n_sensitive_features): Sensitive features to compute the fairness constraint.

Returns:

Andarray of shape (n_samples, n_archetypes): X transformed in the new space.

archetypes.FairAA#

This Page