fairlearn.reductions.ExponentiatedGradient#

class fairlearn.reductions.ExponentiatedGradient(estimator, constraints, *, objective=None, eps=0.01, max_iter=50, nu=None, eta0=2.0, run_linprog_step=True, sample_weight_name='sample_weight')[source]#

An Estimator which implements the exponentiated gradient reduction.

Added in version 0.3.0.

The exponentiated gradient algorithm is described in detail by Agarwal et al.[1].

Read more in Passing pipelines to mitigation techniques.

Changed in version 0.3.0: Was a function before, not a class

Changed in version 0.4.6: Requires 0-1 labels for classification problems

Parameters:

  • estimator (estimator) – An estimator implementing methods fit(X, y, sample_weight) and predict(X), where X is the matrix of features, y is the vector of labels (binary classification) or continuous values (regression), and sample_weight is a vector of weights. In binary classification labels y and predictions returned by predict(X) are either 0 or 1. In regression values y and predictions are continuous.

  • constraints (Moment) – The fairness constraints expressed as a Moment.

  • objective (Moment | None) – The objective expressed as a Moment. The default is ErrorRate() for binary classification and MeanLoss(...) for regression.

  • eps (float) –

    Allowed fairness constraint violation; the solution is guaranteed to have the error within 2*best_gap of the best error under constraint eps; the constraint violation is at most 2*(eps+best_gap).

    Changed in version 0.5.0: eps is now only responsible for setting the L1 norm bound in the optimization

  • max_iter (int) –

    Maximum number of iterations

    Added in version 0.5.0: Used to be T

  • nu (float | None) – Convergence threshold for the duality gap, corresponding to a conservative automatic setting based on the statistical uncertainty in measuring classification error

  • eta0 (float) –

    Initial setting of the learning rate

    Added in version 0.5.0: Used to be eta_mul

  • run_linprog_step (bool) –

    if True each step of exponentiated gradient is followed by the saddle point optimization over the convex hull of classifiers returned so far; default True

    Added in version 0.5.0.

  • sample_weight_name (str) –

    Name of the argument to estimator.fit() which supplies the sample weights (defaults to sample_weight)

    Added in version 0.5.0.

fit(X, y, **kwargs)[source]#

Return a fair classifier under specified fairness constraints.

Parameters:
get_metadata_routing()[source]#

Get metadata routing of this object.

Please check User Guide on how the routing mechanism works.

Returns:

routing (MetadataRequest) – A MetadataRequest encapsulating routing information.

get_params(deep=True)[source]#

Get parameters for this estimator.

Parameters:

deep (bool, default=True) – If True, will return the parameters for this estimator and contained subobjects that are estimators.

Returns:

params (dict) – Parameter names mapped to their values.

predict(X, random_state=None)[source]#

Provide predictions for the given input data.

Predictions are randomized, i.e., repeatedly calling predict with the same feature data may yield different output. This non-deterministic behavior is intended and stems from the nature of the exponentiated gradient algorithm.

Parameters:

  • X (numpy.ndarray or pandas.DataFrame) – Feature data

  • random_state (int or RandomState instance, default=None) – Controls random numbers used for randomized predictions. Pass an int for reproducible output across multiple function calls.

Return type:

Scalar or vector

Returns:

The prediction. If X represents the data for a single example the result will be a scalar. Otherwise the result will be a vector

Notes

A fitted ExponentiatedGradient has an attribute predictors_, an array of predictors, and an attribute weights_, an array of non-negative floats of the same length. The prediction on each data point in X is obtained by first picking a random predictor according to the probabilities in weights_ and then applying it. Different predictors can be chosen on different data points.

set_params(**params)[source]#

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:

**params (dict) – Estimator parameters.

Returns:

self (estimator instance) – Estimator instance.

set_predict_request(*, random_state='$UNCHANGED$')[source]#

Configure whether metadata should be requested to be passed to the predict 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 predict 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 predict.

  • 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:

random_state (Union[bool, None, str]) – Metadata routing for random_state parameter in predict.

Return type:

ExponentiatedGradient

Returns:

self (object) – The updated object.