streamline.utils.evaluation module

streamline.utils.evaluation.class_eval(y_true, y_pred)

Calculates standard classification metrics including: True positives, false positives, true negative, false negatives, standard accuracy, balanced accuracy recall, precision, f1 score, negative predictive value, likelihood ratio positive, and likelihood ratio negative

Parameters:

• y_true – True Labels

• y_pred – Predicted Labels

Returns: list [bac, ac, f1, re, sp, pr, tp, tn, fp, fn, npv, lrp, lrm] ordered list of balanced accuracy, accuracy, F1-score, recall, specificity, precision, true positive, true negatives, false positives, false negatives, negative predictive value, likelihood ratio positive, and likelihood ratio negative