fair_seldonian.algorithms package

Submodules

fair_seldonian.algorithms.qsa module

fair_seldonian.algorithms.qsa.QSA(X, Y, T, seldonian_type, init_sol, init_sol1, config=SeldonianConfig(delta=0.05, inequality=<Inequality.HOEFFDING_INEQUALITY: 2>, constraint='TP(1) TP(0) - abs 0.25 TP(1) * -', candidate_ratio=0.4))[source]

This function is used to run the qsa (Quasi-Seldonian Algorithm)

Parameters:

  • X (ndarray) – The features of the dataset

  • Y (ndarray) – The corresponding labels of the dataset

  • T (ndarray) – The corresponding sensitive attributes of the dataset

  • seldonian_type (str) – The mode used in the experiment

  • init_sol (Tensor | None) – The initial theta values for the model

  • init_sol1 (Tensor | None) – The additional initial theta values for the model

  • config (SeldonianConfig) – Algorithm configuration

Returns:

(thetexit

Return type:

tuple[Tensor, Tensor, bool]

a, theta1, passed_safety) tuple

fair_seldonian.algorithms.qsa.safety_test(theta, theta1, safe_data_X, safe_data_Y, safe_data_T, seldonian_type, config=SeldonianConfig(delta=0.05, inequality=<Inequality.HOEFFDING_INEQUALITY: 2>, constraint='TP(1) TP(0) - abs 0.25 TP(1) * -', candidate_ratio=0.4))[source]

This function does the safety test.

Parameters:

  • theta (Tensor) – The optimal theta values for the model

  • theta1 (Tensor) – The additional optimal theta values for the model

  • safe_data_X (ndarray) – The features of the safety dataset

  • safe_data_Y (ndarray) – The corresponding labels of the safety dataset

  • safe_data_T (ndarray) – The corresponding sensitive attributes of the safety dataset

  • seldonian_type (str) – The mode used in the experiment

  • config (SeldonianConfig) – Algorithm configuration

Returns:

Bool value of whether the candidate solution passed safety test or not.

Return type:

bool

fair_seldonian.algorithms.qsa.get_cand_solution(cand_data_X, cand_data_Y, cand_data_T, seldonian_type, init_sol, init_sol1, config=SeldonianConfig(delta=0.05, inequality=<Inequality.HOEFFDING_INEQUALITY: 2>, constraint='TP(1) TP(0) - abs 0.25 TP(1) * -', candidate_ratio=0.4))[source]

This function provides the candidate solution.

Parameters:

  • cand_data_X (ndarray) – The features of the candidate dataset

  • cand_data_Y (ndarray) – The corresponding labels of the candidate dataset

  • cand_data_T (ndarray) – The corresponding sensitive attributes of the candidate dataset

  • seldonian_type (str) – The mode used in the experiment

  • init_sol (Tensor | None) – The initial theta values for the model

  • init_sol1 (Tensor | None) – The additional initial theta values for the model

  • config (SeldonianConfig) – Algorithm configuration

Returns:

The candidate solution (theta, theta1).

Return type:

tuple[Tensor, Tensor]

fair_seldonian.algorithms.qsa.cand_obj(theta, cand_data_X, cand_data_Y, cand_data_T, seldonian_type, config)[source]

Objective function minimized by the optimizer.

Parameters:

  • theta (ndarray) – The theta values for the model

  • cand_data_X (ndarray) – The features of the candidate dataset

  • cand_data_Y (ndarray) – The corresponding labels of the candidate dataset

  • cand_data_T (ndarray) – The corresponding sensitive attributes of the candidate dataset

  • seldonian_type (str) – The mode used in the experiment

  • config (SeldonianConfig) – Algorithm configuration

Returns:

The objective value.

Return type:

float

Module contents