How to generate a ROC curve for a dataset?

Question

[Honey]

I would like to generate a ROC curve for my dataset containing two classes(A,B) The data-set has around 16000 columns and 60 rows. And the last column Class represents the class of the variables.

Now do I need to iterate the whole code 10 times(since i m doing 10 cross validation) to generate 10 different confusion matrix, and based on that I can draw ROC curve on to check the performance of training and test data??

[e.rempel]

I think this question should be asked on StackOverflow.

Answer 1

[moxu]

You can generate 10 ROC curves with 10 AUC scores. AUC scores can be compared easily.