fatf.utils.data.datasets.load_health_records

fatf.utils.data.datasets.load_health_records() → Dict[str, numpy.ndarray]

Loads in a fake health records dataset.

The dataset contains a mixture of categorical and numerical columns generated with faker. The data array is a structured numpy array with the following columns and types: 'name' (string), 'email' (string), 'age' (integer), 'weight' (integer), 'gender' (string), 'zipcode' (string), 'diagnosis' (string) and 'dob' (string) – date of birth. The target variable is binary (numerical) and encodes a medical treatment has succeeded: 1 is 'success' and 0 is 'fail'.

Returns

dataDict[string, numpy.ndarray]

A dictionary with the dataset and its metadata. See fatf.utils.data.datasets.load_data for the data format.

Examples using fatf.utils.data.datasets.load_health_records

Measuring Fairness of a Predictive Model – Disparate Impact

Measuring Fairness of a Prediction – Counterfactual Fairness

Measuring Fairness of a Data Set