Use Youden index to determine cut-off for classification

sensivity_specifity_cutoff.py

import numpy as np
from sklearn.metrics import roc_curve

def sensivity_specifity_cutoff(y_true, y_score):
    '''Find data-driven cut-off for classification
    
    Cut-off is determied using Youden's index defined as sensitivity + specificity - 1.
    
    Parameters
    ----------
    
    y_true : array, shape = [n_samples]
        True binary labels.
        
    y_score : array, shape = [n_samples]
        Target scores, can either be probability estimates of the positive class,
        confidence values, or non-thresholded measure of decisions (as returned by
        “decision_function” on some classifiers).
        
    References
    ----------
    
    Ewald, B. (2006). Post hoc choice of cut points introduced bias to diagnostic research.
    Journal of clinical epidemiology, 59(8), 798-801.
    
    Steyerberg, E.W., Van Calster, B., & Pencina, M.J. (2011). Performance measures for
    prediction models and markers: evaluation of predictions and classifications.
    Revista Espanola de Cardiologia (English Edition), 64(9), 788-794.
    
    Jiménez-Valverde, A., & Lobo, J.M. (2007). Threshold criteria for conversion of probability
    of species presence to either–or presence–absence. Acta oecologica, 31(3), 361-369.
    '''
    fpr, tpr, thresholds = roc_curve(y_true, y_score)
    idx = np.argmax(tpr - fpr)
    return thresholds[idx]