Thanks Thushan for the suggestion. Can you please create a Jira with this suggestion?
On Mon, Aug 24, 2015 at 5:33 AM, Thushan Ganegedara <thu...@gmail.com> wrote: > Hi, > > The current prediction logic requires the prediction test file to have 1 > column less than of the training file (i.e. removing the response column). > > But I think it will be more convenient for users if we do not enforce that > rule. Instead, perform the following check in prediction logic. > > > If column_count_predict == column_count_train - 1 > > Follow normal logic > > else if column_count_predict == column_count_train > > Get the response column ID/Index > Remove the corresponding column from the predict file > Warn the user about the operation > Follow normal logic > > Else > > Produce an error > > Because something, test files deliberately include the classes column to > make it easier to test the accuracy > > > -- > Regards, > > Thushan Ganegedara > School of IT > University of Sydney, Australia > -- Thanks & regards, Nirmal Team Lead - WSO2 Machine Learner Associate Technical Lead - Data Technologies Team, WSO2 Inc. Mobile: +94715779733 Blog: http://nirmalfdo.blogspot.com/
_______________________________________________ Dev mailing list Dev@wso2.org http://wso2.org/cgi-bin/mailman/listinfo/dev
