0
$\begingroup$

I am doing a binary classification task and I am calculating the AUC ROC and the Recall as follows :

  recall= metrics.recall_score(y_test,y_pred,pos_label=None,average='macro')
  auc=roc_auc_score(y_test, y_pred)

and I always end up wit the exact same value for both ( recall = auc )! I have no idea why this is happening .. any help ?!

$\endgroup$
3
  • $\begingroup$ They are not the same thing. Please give data set for us to debug. $\endgroup$
    – SmallChess
    Sep 20, 2016 at 7:17
  • $\begingroup$ I have changed the recall calculation to the following and now it gives me different score .. does this make sense ?.... recall= metrics.recall_score(y_test,y_pred) .. it seems that when I calculate for each class ( the macro ) it gives the same as the AUC ! I have also calculate the AUC using the function auc to make sure of the result and it gave me the same as roc_auc_score $\endgroup$
    – Ophilia
    Sep 20, 2016 at 13:03
  • $\begingroup$ Please give out dataset so we can check. $\endgroup$
    – SmallChess
    Sep 20, 2016 at 13:47

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.