Github user avulanov commented on the pull request: https://github.com/apache/spark/pull/1155#issuecomment-48590095 @mengxr I've addressed your comments. I've also added weighted true positive and false positive and confusion matrix. Regarding raw scores, as you asked in one of the previous messages, now I think it is possible to incorporate them, if they will be between 0 and 1. There will be default parameter in constructor: raw: Boolean = false. The question is how the interface of this class will change. Probably, adding receiver-operating characteristic ROC: Array[Double, Double] will be enough. I don't think adding all kinds of precision: Array[Double, Double] etc will be beneficial. What do you think?
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---